| Component | Translated | Unfinished | Unfinished words | Unfinished characters | Untranslated | Checks | Suggestions | Comments | |
|---|---|---|---|---|---|---|---|---|---|
| kit-json | 0 | 0 | 0 | 0 | 19 | 0 | 0 | ||
|
|
|||||||||
| Glossary The Kit | 0 | 0 | 0 | 0 | 0 | 0 | 0 | ||
|
|
|||||||||
Overview
| Project website | kit.exposingtheinvisible.org |
|---|
5 days ago
String statistics
| Strings percent | Hosted strings | Words percent | Hosted words | Characters percent | Hosted characters | |
|---|---|---|---|---|---|---|
| Total | 114 | 277,862 | 1,996,987 | |||
| Translated | 100% | 114 | 100% | 277,862 | 100% | 1,996,987 |
| Needs editing | 0% | 0 | 0% | 0 | 0% | 0 |
| Read-only | 0% | 0 | 0% | 0 | 0% | 0 |
| Failing checks | 16% | 19 | 61% | 170,861 | 61% | 1,236,811 |
| Strings with suggestions | 0% | 0 | 0% | 0 | 0% | 0 |
| Untranslated strings | 0% | 0 | 0% | 0 | 0% | 0 |
Quick numbers
and previous 30 days
Trends of last 30 days
—
Hosted words
+100%
—
Hosted strings
+100%
—
Translated
+100%
—
Contributors
—
 None
Resource updated |
|
|
None
String added in the repository |
vulnerable-sources
vulnerable-sources
|
|
None
String added in the repository |
Navigating Sensitive Stories: A Guide to Ethical Engagement with Vulnerable Sources
=============================== By Stavros Malichudis  ```cik-in-short ``` **In short:** Drawing on over a decade of investigative reporting experience, this practical guide by Stavros Malichudis in collaboration with Tactical Tech offers concrete steps in working with vulnerable sources. --- ## Introduction How should journalists and researchers approach sources who have experienced trauma? How can they do so in a way that avoids unintentionally re-traumatising them? And how can we build trust, and make sure that individuals willing to share important information with us will not be targeted or face other negative consequences? As an investigative reporter and editor for over a decade, I have worked closely with vulnerable sources in the field. I have spoken to [shipwreck survivors](https://exposingtheinvisible.org/en/films/pylos-shipwreck/); victims of violent state collective expulsions known as ‘pushbacks’; exploited [land workers](https://wearesolomon.com/mag/focus-area/migration/greek-strawberries-made-in-bangladesh/) and [call center agents](https://insidestory.gr/article/teleperformance-part2); unaccompanied minors who lived on the street or had been detained for up to 48 hours in secret ‘prisons,’ which were passenger ferries running between Italy and Greece; residents who lost their neighbors and their houses in a massive fire outside of Athens in the summer of 2018; locals whose loved ones were among [the 64 people who died in 2020 in a care home in Crete, Greece](https://www.investigate-europe.eu/en/posts/death-in-paradise-how-64-care-home-residents-died-in-crete); mothers in Greece and Germany whose sons [were killed by neo-Nazis](https://wearesolomon.com/mag/format/interview/their-sons-were-murdered-by-neo-nazis-now-they-vow-to-keep-their-memory-alive/); and administration and industry insiders who became whistleblowers. This guide draws on these experiences to address ethical concerns that span the entire investigative project lifestyle, from initial design, to implementation, and finally to engaging audiences in the post-publication phase. ## Pre-reporting phase ### 1. Understand which sources are vulnerable First, we need to have a clear understanding of the many diverse roles and identities that a ‘vulnerable source’ can belong to. A ‘vulnerable source’ could be, but is not limited to: * An undocumented worker or a person on the move (migrant); * A person who lost a loved one in an accident; * A person who lost their property in a natural disaster; and * A whistleblower on the brink of exposing misconduct or mistreatment within a company, a state agency, or an international organisation Focusing on vulnerable voices should be the standard practice for *any* story striving to illuminate underreported issues and hold power accountable. Vulnerable sources can enlighten perspectives, for both the researcher and the reader, by offering the opportunity to perceive a situation as thoroughly and holistically as possible. Including vulnerable sources in your research isn't just about doing the right thing; it's also about better research. While many stories highlight the usual prominent voices, tapping into those often overlooked can uncover fresh, underexplored perspectives. It's a powerful way to make your work more inclusive and insightful. But for these individuals, speaking to a researcher carries significant risk. Depending on their status, they could face job loss, retaliation or community rejection, risk of arrest, and more. Given the stakes, researchers serving the public interest with their investigations have an ethical duty to protect the sources of their work. The rest of this guide presents a curated list of best practices to help researchers incorporate the voices of vulnerable sources while minimising potential harm. ### 2. Do your research before approaching vulnerable sources Before establishing contact with an individual or a community, it’s vital that some time is dedicated to pre-reporting. Ask yourself: What do you know about the person you are willing to approach or the community they are part of? Do you have a basic understanding of the community’s manners or customs? A basic overview of the individual’s story? Dedicate some time to answer these questions and gather some basic, initial information. This can be done through searching online, making phone calls to experts, and asking colleagues. A good first impression can save both parties from awkward situations. It can also help create a first layer of trust; people appreciate it when others have made the effort to learn things about them. #### For communities: Research the communities you are approaching, and treat them with respect by adhering to their customs. Demonstrate that you have done your due diligence and establish that this is a source-researcher relationship you are interested in building. For instance, in some cultures men do not greet women with a handshake. In other cultures, for example, refusing to accept a drink that someone offers can be seen as lack of respect. These cultural circumstances can be difficult to navigate at times, as you balance your specific risk assessment and safety concerns with cultural practices in certain contexts (such as the fear of being poisoned by a potential adversary). Above all, try not to take cultural differences for granted — it’s vital to adapt to and enter each context with as much care and awareness as possible. #### For individuals: Again, research as much as possible beforehand. Demonstrate that you took the time to learn about whoever you are approaching. If that person carries with them a traumatic experience, thorough research before meeting them can help avoid re-traumatisation. A journalist's due diligence involves gathering as much context as possible to understand what is at stake for the source. When I first worked on the [story of the care home](https://www.investigate-europe.eu/en/posts/death-in-paradise-how-64-care-home-residents-died-in-crete) in Crete, Greece, where 64 people died in a single year, some of the relatives who lost loved ones had already talked to the media. Having read their interviews, when I met with them I was able to ask more targeted questions to get the most insightful information. Having an idea of the chain of events, I also had the opportunity to treat traumatic memories with more care. Doing as much research as possible can help you show sources that you took the time to learn them, stick to the facts when it comes to traumatic experiences, and respectfully acknowledge the pain people share. While a long-standing saying in journalism says that there are no stupid questions, perhaps the only stupid question that could be posed to a vulnerable source is one that will unnecessarily make them re-experience their trauma. ### 3. Plan your initial approach Often, the best way to approach sources who might be hesitant to talk to you is through people who can vouch for your integrity, your credibility, and your ethical standards. These could be mutual acquaintances, like an NGO that you have worked with before. Or, perhaps you know members of the same community who can introduce you to your potential source. If the individual has a public social media profile, you can search for possible common connections. This allows for a smoother approach, as well as a better informed pre-reporting phase. ### 4. Recognise the power balance between you and your sources Researchers should clearly understand the power dynamics at play when working with vulnerable sources. A vulnerable source inherently holds less power than a reporter. This imbalance can stem from a source's level of wealth or income, legal status, social standing, age, race, ethnicity, nationality, religion, or other factors. Researchers need to take power imbalances in mind when designing an investigation. They must have in mind the insecure position of their source, their ethical duty to protect them and treat them with care, and the possible challenges that may arise during the research phase, and beyond. ## Research phase ### 1. Explain yourself The success of a research project involving vulnerable sources often comes down to how you explain yourself, the process you follow, the outcomes of the research project, and any possible implications such as short or long-term risks posed to your sources in this case. Vulnerable sources are entitled to clear explanations of the above, as early on as possible in the process. Researchers should keep in mind that their sources may not be familiar with their research or its possible implications. Researchers need to provide this context themselves and explain each aspect of a research project as needed. As early as possible in the process, researchers must seek agreement with sources on these major aspects: #### The nature and level of involvement: What is it exactly that you want to gain from your sources? Will this require a single interview or a series of meetings? Should they also expect phone calls from you? Do you expect them to be available for a set period of time, e.g., for the duration of a trial? Is your editor, a photographer, or a fact-checker going to call them later on? #### The duration and timing of the engagement: Very often, vulnerable sources contribute to time-sensitive research projects. For instance, when investigating how governments and organisations used funds provided by the EU for the accommodation of asylum seekers, I met a family who stayed in one of these housing units in Athens. The NGO responsible for them was receiving millions of euros from the EU, but photos the family captured showed that conditions inside their apartment were horrific. They wanted to expose the conditions, but it was best for them to do so after they had left the apartment, due to a fear of being identified. Similar conditions may occur when a company whistleblower wishes to denounce internal processes. It may be better to publish any material only *after vulnerable sources* have left these environments. Time of publication should be agreed upon early on to avoid frustration on both sides. No surprises! #### The format: When you interview a vulnerable source, will you quote them in full, or will you rephrase or paraphrase their words? Before recording a source’s voice or likeness through photos or videos, always obtain informed consent, taking care to explain how this material will be used. When recording their voice, for example, state from the outset whether it is strictly for your own note-taking or for inclusion in the publication of the story. #### Publication details: Where will the story be published, and in what context? In what languages and formats will the outcome be available? #### Fact-checking: Information provided by vulnerable sources, in fact by *any* sources, must be corroborated independently. This can be tiring or disappointing for people who share their trauma with you. Consider the [Pylos shipwreck](https://exposingtheinvisible.org/en/films/pylos-shipwreck/), the largest in the Mediterranean in recent years, which claimed the lives of over 600 men, women, and children. Some survivors said they had stayed in the water for an hour before a Greek Coast Guard vessel came to their rescue. Others said it was closer to half an hour. While probing such details can feel insensitive given the immense suffering, these discrepancies still require meticulous fact-checking. For sources unfamiliar with journalistic processes, it's crucial to explain that independent verification is paramount to upholding the integrity and validity of any investigation. This transparent communication also means being clear that some information, no matter how personally significant, may not be included in the final report if it cannot be independently verified. Clarifying these aspects in advance can help you manage expectations, maintain trust, and mitigate frustrations. ### 2. Let sources know they are also in control All human relations come with power dynamics. Usually, journalists have power over sources, as they ultimately decide what to include in a piece or not. But with sources, especially vulnerable sources, it’s important to know that they are also in control of their own story. One of the first steps I follow when working with vulnerable sources is to make clear to them that they have the option to decide if specific information shared with me is on the record, off the record, or on background. In other words, if they don’t want to answer a specific question, they don’t have to. They should also have the opportunity to pause an interview, if at any time they feel the need to do so. As researchers, we should show patience. Give them the time and space they need, and don’t pressure them. Resist the urge to fill pauses, or periods of silence. If a discussion touches on tragic events, don’t just keep pushing for answers. Take a break when the person seems distressed or simply wants to stop. ```cik-note ``` >**Some terms to explain to your sources:** > >* **On the record:** Unless explicitly agreed otherwise, any conversation with a journalist is considered ‘on the record’. This means everything a source says can be quoted directly, attributed by name, and published. It's the highest level of transparency and accountability, allowing readers to see precisely who said what. > >* **Off the record:** When something is ‘off the record,’ the journalist agrees that the information provided cannot be published or attributed to the source. Journalists and sources must mutually agree that a conversation is ‘off the record’ before any information is shared. > >* **On background:** When a conversation is ‘on background,’ journalists can publish and quote the information provided by a source, but cannot identify the source by name or otherwise attribute that information to the source using other identifiable information. Researchers can describe the source as accurately as possible, for example as ‘a source familiar with the incident’ or ‘an employee on the field’, without revealing the source’s identity. > ### 3. Give it time In 2019, during a months-long investigative project, I was tracking unaccompanied migrant children in Greece. These kids faced unimaginable challenges; some had experienced homelessness, others had been detained in police stations for months. As I worked to build trust and gather information, my editor offered a brilliant piece of advice: “just hang out with them.” No recording, no note-taking, just hanging out. Do what they do. Spend an afternoon with them, then another, without the pressure of collecting material for the written piece I had in mind. This approach was illuminating. It allowed me to visualise and truly understand the humanity of my potential sources first. It fostered a more natural, less rushed relationship, letting things unfold organically. Investigations, especially with vulnerable sources, demand time. That fellowship taught me a crucial lesson about building trust. When sources sense genuine care for their story, they often become your strongest advocates, introducing you to others and vouching for your credibility. It's a powerful ripple effect that extends far beyond any single interview. ### 4. Explain the possible implications Letting vulnerable sources know they’re also in control is related to your responsibility to inform them about the possible implications that may arise for them by speaking to you. For instance, a few years back I worked on an investigation exposing labor exploitation and racism in a call-center giant in Greece. One of the main sources, an ex-employee of the company, insisted on speaking on the record. Although it’s always better for the validity of an investigation that sources speak on the record, it was my duty back then to inform the person that, given the practices in that specific sector, speaking publicly against the industry’s major company could mean that employers in his sector would no longer want him at their company. The source took a few days to think on this. He decided to still do it, as he saw his professional future in an entirely unrelated field. It was his choice in the end, but it was my responsibility to provide context that he wasn’t necessarily aware of. In other cases, when a source shares potentially identifying or incriminating information, it is imperative that a researcher might ask: “Are you sure you want to share this?” - This reminds the source of the risks and empowers them to decide what to share and withhold. ### 5. Don’t create expectations “Why should I speak to you?” - This is a question I often get from residents of refugee camps across Greece. - “You do your story, what will I get?” Vulnerable sources might be in dire need of financial aid or of state documents that could make their life move forward, to name just two examples. You might be asking them to talk to you at a time when they feel their life is falling apart. Researchers should make sure not to exploit the hopes of vulnerable sources. Simply put, don’t offer things in return for information, and don’t promise what you cannot deliver. It is imperative to explain that the actual impact of your story, especially for the individual’s own story, might be limited. But you should also explain that your research will contribute to holding power to account, and may contribute to a wider impact beyond this one story. While you may not be in a position to help them get certain documents, you can help them in other ways, like to let them know which NGOs offer legal support or to tell them what you know about labor inspections. ### 6. Ensure physical safety when meeting in person Are sources talking to you at risk if someone sees them with a reporter or a researcher? Would source protection be compromised? Physical safety during meetings demands careful planning. Be sure to consider: * **Awareness of Surroundings:** Ensure you are not being followed. Have a good understanding of the area you are in. * **Discreet Meeting Locations:** Avoid places where you could be easily seen. If meeting at a café, choose a seat with a direct view of the entrance. Avoid outdoor seating where you might be visible to individuals you cannot see. * **Location Privacy:** If using taxis or ride-sharing apps to meet a vulnerable source, avoid mentioning or inputting the exact meeting address directly into the app. You can ride to a walking distance address, and get on foot from there. * **Mobile Device Protocols:** When meeting sensitive sources, consider not carrying your mobile phone or disabling GPS to protect their safety, even if you perceive no personal risk. * **Inform your source:** Share with your source the different steps you take to ensure their protection. This can help build trust among the two parties, but it can also lead to valuable feedback on the specifics. ### 7. Ensure digital safety In-person meetings are not the only form of communication in need of security. Securing digital communication is also paramount. This includes: * **Password Protection:** Keep your phone password-protected. * **Auto-Delete Messages:** Utilise features that automatically delete messages. Most applications (like Signal, WhatsApp) have this feature. * **Encrypted Communication:** Avoid standard phone lines and SMS. Prioritise encrypted messaging apps like Signal, which is favored by researchers and academics due to its strong security. * **End-to-End Encryption:** Opt for services designed with privacy, where end-to-end encryption is standard, ensuring only you and your source can read messages. ```cik-note ``` >**Digital Security Resources** > >For fundamentals on digital security, physical security and wellbeing as well as risk assessment, you can start with articles and guides such as: > >* Security in a Box guides and tutorials: [https://securityinabox.org/](https://securityinabox.org/) >* Safety First guide: [https://kit.exposingtheinvisible.org/en/safety.html](https://kit.exposingtheinvisible.org/en/safety.html) >* Risk Assessment guide: [https://exposingtheinvisible.org/en/articles/risk-assessment-mindset](https://exposingtheinvisible.org/en/articles/risk-assessment-mindset) >* Holistic Security manual: [https://holistic-security.tacticaltech.org/](https://holistic-security.tacticaltech.org/) > ### 8. Handle documents securely For sensitive investigations, physical transfer of documents can sometimes be the safest method, such as through mail or direct delivery. Physical transfer doesn’t leave metadata. Edits can’t be tracked. This was reportedly how Edward Snowden provided documents to reporters, according to ["Snowden's Box: Trust in the Age of Surveillance"](https://www.versobooks.com/products/852-snowden-s-box?srsltid=AfmBOop4rkUd1yEAj2yNWs2yL5PTqlrs3MMKtHk7bmUuARbO0H0-Ckhi) by Jessica Bruder and Dale Maharidge, two experienced journalists who worked behind the scenes of the Snowden story. As the summary of the book notes, "The biggest national security leak of the digital era was launched via a remarkably analog network, the US Postal Service." If a source provides you with a document, you have to make sure that you properly anonymise it in case you wish to publish it together with your investigation or share it with others outside your fully trusted team. Make sure you really anonymise it. Don’t just cover content with blacked out text boxes that can be removed from the file! Image edits with specific tools can be tracked, as can metadata of the files, so one needs to very carefully work on this to ensure full source protection. ```cik-note ``` >**Metadata** refers to information that describes the properties of a file, be it an image, a document, a sound recording, a map, etc. For example the contents of an image are the visible elements in it, while the date when the image was taken, the location and device information constitute its metadata. In communications – when sending a message or a file - metadata includes details about who sent it, when, where from, and to whom, among others. See more here: [https://ssd.eff.org/module/why-metadata-matters](https://ssd.eff.org/module/why-metadata-matters "https\://ssd.eff.org/module/why-metadata-matters"). > Of course, file and data anonymisation has its drawbacks. Often it means altering evidence, so you will need to make sure you safely preserve original, unaltered versions of files for any possible situations that might arise, including, for example, court procedures. In 2022, after months of investigation, Greece's National Transparency Authority (NTA) released a redacted report addressing allegations of illegal pushbacks of asylum seekers. The report's conclusion: no such pushbacks had occurred. However, researchers quickly discovered a critical flaw: basic digital design tools were all that was needed to remove the blacked-out sections intended to protect the anonymity of the report's sources, thereby compromising the identities of the individuals who had informed the research. ```cik-tip ``` > See more tips and examples about data anonymisation in these articles: > >* “How Data Journalists Can Use Anonymization to Protect Privacy”, GIJN: [https://gijn.org/stories/how-data-journalists-can-use-anonymization-to-protect-privacy/](https://gijn.org/stories/how-data-journalists-can-use-anonymization-to-protect-privacy/) > >* “Privacy and data leaks: How to decide what to report”: [https://datajournalism.com/read/longreads/privacy-and-data-leaks](https://datajournalism.com/read/longreads/privacy-and-data-leaks) > >* “Why does The New York Times use anonymous sources?”, The New York Times, [https://www.nytimes.com/article/why-new-york-times-anonymous-sources.html](https://www.nytimes.com/article/why-new-york-times-anonymous-sources.html) > >* “Here are 12 principles journalists should follow to make sure they’re protecting their sources”, NiemanLab: [https://www.niemanlab.org/2019/01/here-are-12-principles-journalists-should-follow-to-make-sure-theyre-protecting-their-sources/](https://www.niemanlab.org/2019/01/here-are-12-principles-journalists-should-follow-to-make-sure-theyre-protecting-their-sources/) > >* “Investigative Journalism: How to Develop and Manage your Sources”, Al Jazeera Journalism Review: [https://institute.aljazeera.net/en/ajr/article/investigative-journalism-how-develop-and-manage-your-sources](https://institute.aljazeera.net/en/ajr/article/investigative-journalism-how-develop-and-manage-your-sources) > >* “Media Defense Guide for Investigative Journalists”, Media Defence: [https://gijn.org/stories/legal-help-for-journalists/](https://gijn.org/stories/legal-help-for-journalists/) (see more resources from Media Defense here: [https://www.mediadefence.org/resource-hub/](https://www.mediadefence.org/resource-hub/) > >* “The Perugia Principles for Journalists Working with Whistleblowers in the Digital Age”: [https://whistleblowingnetwork.org/WIN/media/pdfs/Journalism-Sources-INT-Blueprint-2018-Perugia-Principles-for-Journalists.pdf](https://whistleblowingnetwork.org/WIN/media/pdfs/Journalism-Sources-INT-Blueprint-2018-Perugia-Principles-for-Journalists.pdf) > ### 9. Show empathy in interviews When interviewing vulnerable sources, traditional journalistic approaches — like quickly shifting between topics — can inadvertently inflict further harm. Instead, a more deliberate and structured interview approach is crucial. Before a sensitive interview, it's vital to carefully plan the progression of your questions. The goal is to guide the source through difficult topics efficiently, ensuring they only have to recount painful experiences a single time. For example, a parent discussing the loss of a child should ideally only need to recall the specifics of their child's death once during the conversation. ### 10. Keep showing up Vulnerable sources often go to extraordinary lengths to provide information for investigative projects. Yet, once the research phase concludes, contact frequently ceases. This abrupt disengagement can leave sources feeling exploited, deepening a sense that they were merely a means to an end for a story, especially when their personal realities remain unchanged. Imagine a journalist who maintains daily contact with a source for weeks, only to vanish once the story is filed. This sudden absence can breed resentment and a profound sense of betrayal. Consider the devastating [Mati fire](https://greekcitytimes.com/2025/03/13/mati-fire-guilty/), which claimed 104 lives in the outskirts of Athens in the summer of 2018. Both national and international media swarmed the area. Some reporters and crews stayed and covered the tragedy for up to three months. However, as new events unfolded in Greece and elsewhere, the journalistic gaze shifted, and many moved on, never to return. My former colleague, Tasos Telloglou of Inside Story, chose a different path. He continued to visit Mati long after the initial media frenzy subsided. A year later, when we began working together on follow-up stories about the fire's aftermath, I - despite being new to the community - was granted immediate access and full trust. They didn’t know me; this was solely because I was collaborating with him. Tasos had earned that trust by consistently making the drive to Mati when no one else did. His dedication was an illuminating example of how genuine care for a community's story and its ongoing realities, in stark contrast to extractive "parachute journalism," can build enduring trust and the whole different set of investigative possibilities this can bring. ## Post-publication phase Researchers and investigative reporters tend to jump from project to project. But for the vulnerable sources that contributed to an investigation, their everyday lives continue. Even reporters who carefully follow specific steps during the pre-reporting and the reporting phase of investigative projects may overlook the importance of the post-publication phase. It is also worth keeping in mind that while veteran researchers might not feel like the days following publication are special, sources who are unfamiliar with the investigative process may feel differently. They can feel pressure, might think they have been exposed, or even regret having talked to someone. After publishing a story, keep the following steps in mind. ### 1. Share the story Is the story finally out? Send it to your sources. Don’t let them find it via social media, or hear about it from other acquaintances. Even worse, don’t let them find out the story has been published only days later. Thank them once more for their contribution to the investigation and explain to them the story’s significance. ### 2. Ask for feedback Don’t just send your sources the story and leave it at that. Ask them what they think of it. While you may not be in a position to make major changes after publication, it’s important to learn how they viewed their contribution to the story. Are their concerns properly framed in it? Do they feel that your portrayal of their reality was just? ### 3. Remain accessible Journalists and researchers often disappear after publication, which can make vulnerable sources hesitant to trust not just them, but all journalists in the future. Be reachable to your sources, follow up on their wellbeing, and do not vanish. Working with people on the move in Greece for several years now, I often have sources of stories asking me why other journalists stopped replying to messages after they had opened their houses to them, had coffee with them, or shared childhood photos with them. Sources who have experienced these ‘parachute journalists’, will find it more difficult to trust researchers in the future and will feel that they were taken advantage of or used. <hr class="thick"> #### About the author **Stavros Malichudis** is a reporter and editor. He has worked for the Agence France-Presse and inside story, and has participated in cross-border investigations with Lighthouse Reports and Investigate Europe. He’s member of Reporters United. His reports have been published in European media. He was shortlisted for the European Press Prize ‘21 and won the IJ4EU Impact Award ‘22. In 2019 he was selected as a fellow for BIRN's Balkan Fellowship for Journalistic Excellence (BFJE). He has been trained in data journalism at Columbia University in New York on a fellowship. <hr class="thick"> *Published in June 2025*
Navigating Sensitive Stories: A Guide to Ethical Engagement with Vulnerable Sources
=============================== By Stavros Malichudis  ```cik-in-short ``` **In short:** Drawing on over a decade of investigative reporting experience, this practical guide by Stavros Malichudis in collaboration with Tactical Tech offers concrete steps in working with vulnerable sources. --- ## Introduction How should journalists and researchers approach sources who have experienced trauma? How can they do so in a way that avoids unintentionally re-traumatising them? And how can we build trust, and make sure that individuals willing to share important information with us will not be targeted or face other negative consequences? As an investigative reporter and editor for over a decade, I have worked closely with vulnerable sources in the field. I have spoken to [shipwreck survivors](https://exposingtheinvisible.org/en/films/pylos-shipwreck/); victims of violent state collective expulsions known as ‘pushbacks’; exploited [land workers](https://wearesolomon.com/mag/focus-area/migration/greek-strawberries-made-in-bangladesh/) and [call center agents](https://insidestory.gr/article/teleperformance-part2); unaccompanied minors who lived on the street or had been detained for up to 48 hours in secret ‘prisons,’ which were passenger ferries running between Italy and Greece; residents who lost their neighbors and their houses in a massive fire outside of Athens in the summer of 2018; locals whose loved ones were among [the 64 people who died in 2020 in a care home in Crete, Greece](https://www.investigate-europe.eu/en/posts/death-in-paradise-how-64-care-home-residents-died-in-crete); mothers in Greece and Germany whose sons [were killed by neo-Nazis](https://wearesolomon.com/mag/format/interview/their-sons-were-murdered-by-neo-nazis-now-they-vow-to-keep-their-memory-alive/); and administration and industry insiders who became whistleblowers. This guide draws on these experiences to address ethical concerns that span the entire investigative project lifestyle, from initial design, to implementation, and finally to engaging audiences in the post-publication phase. ## Pre-reporting phase ### 1. Understand which sources are vulnerable First, we need to have a clear understanding of the many diverse roles and identities that a ‘vulnerable source’ can belong to. A ‘vulnerable source’ could be, but is not limited to: * An undocumented worker or a person on the move (migrant); * A person who lost a loved one in an accident; * A person who lost their property in a natural disaster; and * A whistleblower on the brink of exposing misconduct or mistreatment within a company, a state agency, or an international organisation Focusing on vulnerable voices should be the standard practice for *any* story striving to illuminate underreported issues and hold power accountable. Vulnerable sources can enlighten perspectives, for both the researcher and the reader, by offering the opportunity to perceive a situation as thoroughly and holistically as possible. Including vulnerable sources in your research isn't just about doing the right thing; it's also about better research. While many stories highlight the usual prominent voices, tapping into those often overlooked can uncover fresh, underexplored perspectives. It's a powerful way to make your work more inclusive and insightful. But for these individuals, speaking to a researcher carries significant risk. Depending on their status, they could face job loss, retaliation or community rejection, risk of arrest, and more. Given the stakes, researchers serving the public interest with their investigations have an ethical duty to protect the sources of their work. The rest of this guide presents a curated list of best practices to help researchers incorporate the voices of vulnerable sources while minimising potential harm. ### 2. Do your research before approaching vulnerable sources Before establishing contact with an individual or a community, it’s vital that some time is dedicated to pre-reporting. Ask yourself: What do you know about the person you are willing to approach or the community they are part of? Do you have a basic understanding of the community’s manners or customs? A basic overview of the individual’s story? Dedicate some time to answer these questions and gather some basic, initial information. This can be done through searching online, making phone calls to experts, and asking colleagues. A good first impression can save both parties from awkward situations. It can also help create a first layer of trust; people appreciate it when others have made the effort to learn things about them. #### For communities: Research the communities you are approaching, and treat them with respect by adhering to their customs. Demonstrate that you have done your due diligence and establish that this is a source-researcher relationship you are interested in building. For instance, in some cultures men do not greet women with a handshake. In other cultures, for example, refusing to accept a drink that someone offers can be seen as lack of respect. These cultural circumstances can be difficult to navigate at times, as you balance your specific risk assessment and safety concerns with cultural practices in certain contexts (such as the fear of being poisoned by a potential adversary). Above all, try not to take cultural differences for granted — it’s vital to adapt to and enter each context with as much care and awareness as possible. #### For individuals: Again, research as much as possible beforehand. Demonstrate that you took the time to learn about whoever you are approaching. If that person carries with them a traumatic experience, thorough research before meeting them can help avoid re-traumatisation. A journalist's due diligence involves gathering as much context as possible to understand what is at stake for the source. When I first worked on the [story of the care home](https://www.investigate-europe.eu/en/posts/death-in-paradise-how-64-care-home-residents-died-in-crete) in Crete, Greece, where 64 people died in a single year, some of the relatives who lost loved ones had already talked to the media. Having read their interviews, when I met with them I was able to ask more targeted questions to get the most insightful information. Having an idea of the chain of events, I also had the opportunity to treat traumatic memories with more care. Doing as much research as possible can help you show sources that you took the time to learn them, stick to the facts when it comes to traumatic experiences, and respectfully acknowledge the pain people share. While a long-standing saying in journalism says that there are no stupid questions, perhaps the only stupid question that could be posed to a vulnerable source is one that will unnecessarily make them re-experience their trauma. ### 3. Plan your initial approach Often, the best way to approach sources who might be hesitant to talk to you is through people who can vouch for your integrity, your credibility, and your ethical standards. These could be mutual acquaintances, like an NGO that you have worked with before. Or, perhaps you know members of the same community who can introduce you to your potential source. If the individual has a public social media profile, you can search for possible common connections. This allows for a smoother approach, as well as a better informed pre-reporting phase. ### 4. Recognise the power balance between you and your sources Researchers should clearly understand the power dynamics at play when working with vulnerable sources. A vulnerable source inherently holds less power than a reporter. This imbalance can stem from a source's level of wealth or income, legal status, social standing, age, race, ethnicity, nationality, religion, or other factors. Researchers need to take power imbalances in mind when designing an investigation. They must have in mind the insecure position of their source, their ethical duty to protect them and treat them with care, and the possible challenges that may arise during the research phase, and beyond. ## Research phase ### 1. Explain yourself The success of a research project involving vulnerable sources often comes down to how you explain yourself, the process you follow, the outcomes of the research project, and any possible implications such as short or long-term risks posed to your sources in this case. Vulnerable sources are entitled to clear explanations of the above, as early on as possible in the process. Researchers should keep in mind that their sources may not be familiar with their research or its possible implications. Researchers need to provide this context themselves and explain each aspect of a research project as needed. As early as possible in the process, researchers must seek agreement with sources on these major aspects: #### The nature and level of involvement: What is it exactly that you want to gain from your sources? Will this require a single interview or a series of meetings? Should they also expect phone calls from you? Do you expect them to be available for a set period of time, e.g., for the duration of a trial? Is your editor, a photographer, or a fact-checker going to call them later on? #### The duration and timing of the engagement: Very often, vulnerable sources contribute to time-sensitive research projects. For instance, when investigating how governments and organisations used funds provided by the EU for the accommodation of asylum seekers, I met a family who stayed in one of these housing units in Athens. The NGO responsible for them was receiving millions of euros from the EU, but photos the family captured showed that conditions inside their apartment were horrific. They wanted to expose the conditions, but it was best for them to do so after they had left the apartment, due to a fear of being identified. Similar conditions may occur when a company whistleblower wishes to denounce internal processes. It may be better to publish any material only *after vulnerable sources* have left these environments. Time of publication should be agreed upon early on to avoid frustration on both sides. No surprises! #### The format: When you interview a vulnerable source, will you quote them in full, or will you rephrase or paraphrase their words? Before recording a source’s voice or likeness through photos or videos, always obtain informed consent, taking care to explain how this material will be used. When recording their voice, for example, state from the outset whether it is strictly for your own note-taking or for inclusion in the publication of the story. #### Publication details: Where will the story be published, and in what context? In what languages and formats will the outcome be available? #### Fact-checking: Information provided by vulnerable sources, in fact by *any* sources, must be corroborated independently. This can be tiring or disappointing for people who share their trauma with you. Consider the [Pylos shipwreck](https://exposingtheinvisible.org/en/films/pylos-shipwreck/), the largest in the Mediterranean in recent years, which claimed the lives of over 600 men, women, and children. Some survivors said they had stayed in the water for an hour before a Greek Coast Guard vessel came to their rescue. Others said it was closer to half an hour. While probing such details can feel insensitive given the immense suffering, these discrepancies still require meticulous fact-checking. For sources unfamiliar with journalistic processes, it's crucial to explain that independent verification is paramount to upholding the integrity and validity of any investigation. This transparent communication also means being clear that some information, no matter how personally significant, may not be included in the final report if it cannot be independently verified. Clarifying these aspects in advance can help you manage expectations, maintain trust, and mitigate frustrations. ### 2. Let sources know they are also in control All human relations come with power dynamics. Usually, journalists have power over sources, as they ultimately decide what to include in a piece or not. But with sources, especially vulnerable sources, it’s important to know that they are also in control of their own story. One of the first steps I follow when working with vulnerable sources is to make clear to them that they have the option to decide if specific information shared with me is on the record, off the record, or on background. In other words, if they don’t want to answer a specific question, they don’t have to. They should also have the opportunity to pause an interview, if at any time they feel the need to do so. As researchers, we should show patience. Give them the time and space they need, and don’t pressure them. Resist the urge to fill pauses, or periods of silence. If a discussion touches on tragic events, don’t just keep pushing for answers. Take a break when the person seems distressed or simply wants to stop. ```cik-note ``` >**Some terms to explain to your sources:** > >* **On the record:** Unless explicitly agreed otherwise, any conversation with a journalist is considered ‘on the record’. This means everything a source says can be quoted directly, attributed by name, and published. It's the highest level of transparency and accountability, allowing readers to see precisely who said what. > >* **Off the record:** When something is ‘off the record,’ the journalist agrees that the information provided cannot be published or attributed to the source. Journalists and sources must mutually agree that a conversation is ‘off the record’ before any information is shared. > >* **On background:** When a conversation is ‘on background,’ journalists can publish and quote the information provided by a source, but cannot identify the source by name or otherwise attribute that information to the source using other identifiable information. Researchers can describe the source as accurately as possible, for example as ‘a source familiar with the incident’ or ‘an employee on the field’, without revealing the source’s identity. > ### 3. Give it time In 2019, during a months-long investigative project, I was tracking unaccompanied migrant children in Greece. These kids faced unimaginable challenges; some had experienced homelessness, others had been detained in police stations for months. As I worked to build trust and gather information, my editor offered a brilliant piece of advice: “just hang out with them.” No recording, no note-taking, just hanging out. Do what they do. Spend an afternoon with them, then another, without the pressure of collecting material for the written piece I had in mind. This approach was illuminating. It allowed me to visualise and truly understand the humanity of my potential sources first. It fostered a more natural, less rushed relationship, letting things unfold organically. Investigations, especially with vulnerable sources, demand time. That fellowship taught me a crucial lesson about building trust. When sources sense genuine care for their story, they often become your strongest advocates, introducing you to others and vouching for your credibility. It's a powerful ripple effect that extends far beyond any single interview. ### 4. Explain the possible implications Letting vulnerable sources know they’re also in control is related to your responsibility to inform them about the possible implications that may arise for them by speaking to you. For instance, a few years back I worked on an investigation exposing labor exploitation and racism in a call-center giant in Greece. One of the main sources, an ex-employee of the company, insisted on speaking on the record. Although it’s always better for the validity of an investigation that sources speak on the record, it was my duty back then to inform the person that, given the practices in that specific sector, speaking publicly against the industry’s major company could mean that employers in his sector would no longer want him at their company. The source took a few days to think on this. He decided to still do it, as he saw his professional future in an entirely unrelated field. It was his choice in the end, but it was my responsibility to provide context that he wasn’t necessarily aware of. In other cases, when a source shares potentially identifying or incriminating information, it is imperative that a researcher might ask: “Are you sure you want to share this?” - This reminds the source of the risks and empowers them to decide what to share and withhold. ### 5. Don’t create expectations “Why should I speak to you?” - This is a question I often get from residents of refugee camps across Greece. - “You do your story, what will I get?” Vulnerable sources might be in dire need of financial aid or of state documents that could make their life move forward, to name just two examples. You might be asking them to talk to you at a time when they feel their life is falling apart. Researchers should make sure not to exploit the hopes of vulnerable sources. Simply put, don’t offer things in return for information, and don’t promise what you cannot deliver. It is imperative to explain that the actual impact of your story, especially for the individual’s own story, might be limited. But you should also explain that your research will contribute to holding power to account, and may contribute to a wider impact beyond this one story. While you may not be in a position to help them get certain documents, you can help them in other ways, like to let them know which NGOs offer legal support or to tell them what you know about labor inspections. ### 6. Ensure physical safety when meeting in person Are sources talking to you at risk if someone sees them with a reporter or a researcher? Would source protection be compromised? Physical safety during meetings demands careful planning. Be sure to consider: * **Awareness of Surroundings:** Ensure you are not being followed. Have a good understanding of the area you are in. * **Discreet Meeting Locations:** Avoid places where you could be easily seen. If meeting at a café, choose a seat with a direct view of the entrance. Avoid outdoor seating where you might be visible to individuals you cannot see. * **Location Privacy:** If using taxis or ride-sharing apps to meet a vulnerable source, avoid mentioning or inputting the exact meeting address directly into the app. You can ride to a walking distance address, and get on foot from there. * **Mobile Device Protocols:** When meeting sensitive sources, consider not carrying your mobile phone or disabling GPS to protect their safety, even if you perceive no personal risk. * **Inform your source:** Share with your source the different steps you take to ensure their protection. This can help build trust among the two parties, but it can also lead to valuable feedback on the specifics. ### 7. Ensure digital safety In-person meetings are not the only form of communication in need of security. Securing digital communication is also paramount. This includes: * **Password Protection:** Keep your phone password-protected. * **Auto-Delete Messages:** Utilise features that automatically delete messages. Most applications (like Signal, WhatsApp) have this feature. * **Encrypted Communication:** Avoid standard phone lines and SMS. Prioritise encrypted messaging apps like Signal, which is favored by researchers and academics due to its strong security. * **End-to-End Encryption:** Opt for services designed with privacy, where end-to-end encryption is standard, ensuring only you and your source can read messages. ```cik-note ``` >**Digital Security Resources** > >For fundamentals on digital security, physical security and wellbeing as well as risk assessment, you can start with articles and guides such as: > >* Security in a Box guides and tutorials: [https://securityinabox.org/](https://securityinabox.org/) >* Safety First guide: [https://kit.exposingtheinvisible.org/en/safety.html](https://kit.exposingtheinvisible.org/en/safety.html) >* Risk Assessment guide: [https://exposingtheinvisible.org/en/articles/risk-assessment-mindset](https://exposingtheinvisible.org/en/articles/risk-assessment-mindset) >* Holistic Security manual: [https://holistic-security.tacticaltech.org/](https://holistic-security.tacticaltech.org/) > ### 8. Handle documents securely For sensitive investigations, physical transfer of documents can sometimes be the safest method, such as through mail or direct delivery. Physical transfer doesn’t leave metadata. Edits can’t be tracked. This was reportedly how Edward Snowden provided documents to reporters, according to ["Snowden's Box: Trust in the Age of Surveillance"](https://www.versobooks.com/products/852-snowden-s-box?srsltid=AfmBOop4rkUd1yEAj2yNWs2yL5PTqlrs3MMKtHk7bmUuARbO0H0-Ckhi) by Jessica Bruder and Dale Maharidge, two experienced journalists who worked behind the scenes of the Snowden story. As the summary of the book notes, "The biggest national security leak of the digital era was launched via a remarkably analog network, the US Postal Service." If a source provides you with a document, you have to make sure that you properly anonymise it in case you wish to publish it together with your investigation or share it with others outside your fully trusted team. Make sure you really anonymise it. Don’t just cover content with blacked out text boxes that can be removed from the file! Image edits with specific tools can be tracked, as can metadata of the files, so one needs to very carefully work on this to ensure full source protection. ```cik-note ``` >**Metadata** refers to information that describes the properties of a file, be it an image, a document, a sound recording, a map, etc. For example the contents of an image are the visible elements in it, while the date when the image was taken, the location and device information constitute its metadata. In communications – when sending a message or a file - metadata includes details about who sent it, when, where from, and to whom, among others. See more here: [https://ssd.eff.org/module/why-metadata-matters](https://ssd.eff.org/module/why-metadata-matters "https\://ssd.eff.org/module/why-metadata-matters"). > Of course, file and data anonymisation has its drawbacks. Often it means altering evidence, so you will need to make sure you safely preserve original, unaltered versions of files for any possible situations that might arise, including, for example, court procedures. In 2022, after months of investigation, Greece's National Transparency Authority (NTA) released a redacted report addressing allegations of illegal pushbacks of asylum seekers. The report's conclusion: no such pushbacks had occurred. However, researchers quickly discovered a critical flaw: basic digital design tools were all that was needed to remove the blacked-out sections intended to protect the anonymity of the report's sources, thereby compromising the identities of the individuals who had informed the research. ```cik-tip ``` > See more tips and examples about data anonymisation in these articles: > >* “How Data Journalists Can Use Anonymization to Protect Privacy”, GIJN: [https://gijn.org/stories/how-data-journalists-can-use-anonymization-to-protect-privacy/](https://gijn.org/stories/how-data-journalists-can-use-anonymization-to-protect-privacy/) > >* “Privacy and data leaks: How to decide what to report”: [https://datajournalism.com/read/longreads/privacy-and-data-leaks](https://datajournalism.com/read/longreads/privacy-and-data-leaks) > >* “Why does The New York Times use anonymous sources?”, The New York Times, [https://www.nytimes.com/article/why-new-york-times-anonymous-sources.html](https://www.nytimes.com/article/why-new-york-times-anonymous-sources.html) > >* “Here are 12 principles journalists should follow to make sure they’re protecting their sources”, NiemanLab: [https://www.niemanlab.org/2019/01/here-are-12-principles-journalists-should-follow-to-make-sure-theyre-protecting-their-sources/](https://www.niemanlab.org/2019/01/here-are-12-principles-journalists-should-follow-to-make-sure-theyre-protecting-their-sources/) > >* “Investigative Journalism: How to Develop and Manage your Sources”, Al Jazeera Journalism Review: [https://institute.aljazeera.net/en/ajr/article/investigative-journalism-how-develop-and-manage-your-sources](https://institute.aljazeera.net/en/ajr/article/investigative-journalism-how-develop-and-manage-your-sources) > >* “Media Defense Guide for Investigative Journalists”, Media Defence: [https://gijn.org/stories/legal-help-for-journalists/](https://gijn.org/stories/legal-help-for-journalists/) (see more resources from Media Defense here: [https://www.mediadefence.org/resource-hub/](https://www.mediadefence.org/resource-hub/) > >* “The Perugia Principles for Journalists Working with Whistleblowers in the Digital Age”: [https://whistleblowingnetwork.org/WIN/media/pdfs/Journalism-Sources-INT-Blueprint-2018-Perugia-Principles-for-Journalists.pdf](https://whistleblowingnetwork.org/WIN/media/pdfs/Journalism-Sources-INT-Blueprint-2018-Perugia-Principles-for-Journalists.pdf) > ### 9. Show empathy in interviews When interviewing vulnerable sources, traditional journalistic approaches — like quickly shifting between topics — can inadvertently inflict further harm. Instead, a more deliberate and structured interview approach is crucial. Before a sensitive interview, it's vital to carefully plan the progression of your questions. The goal is to guide the source through difficult topics efficiently, ensuring they only have to recount painful experiences a single time. For example, a parent discussing the loss of a child should ideally only need to recall the specifics of their child's death once during the conversation. ### 10. Keep showing up Vulnerable sources often go to extraordinary lengths to provide information for investigative projects. Yet, once the research phase concludes, contact frequently ceases. This abrupt disengagement can leave sources feeling exploited, deepening a sense that they were merely a means to an end for a story, especially when their personal realities remain unchanged. Imagine a journalist who maintains daily contact with a source for weeks, only to vanish once the story is filed. This sudden absence can breed resentment and a profound sense of betrayal. Consider the devastating [Mati fire](https://greekcitytimes.com/2025/03/13/mati-fire-guilty/), which claimed 104 lives in the outskirts of Athens in the summer of 2018. Both national and international media swarmed the area. Some reporters and crews stayed and covered the tragedy for up to three months. However, as new events unfolded in Greece and elsewhere, the journalistic gaze shifted, and many moved on, never to return. My former colleague, Tasos Telloglou of Inside Story, chose a different path. He continued to visit Mati long after the initial media frenzy subsided. A year later, when we began working together on follow-up stories about the fire's aftermath, I - despite being new to the community - was granted immediate access and full trust. They didn’t know me; this was solely because I was collaborating with him. Tasos had earned that trust by consistently making the drive to Mati when no one else did. His dedication was an illuminating example of how genuine care for a community's story and its ongoing realities, in stark contrast to extractive "parachute journalism," can build enduring trust and the whole different set of investigative possibilities this can bring. ## Post-publication phase Researchers and investigative reporters tend to jump from project to project. But for the vulnerable sources that contributed to an investigation, their everyday lives continue. Even reporters who carefully follow specific steps during the pre-reporting and the reporting phase of investigative projects may overlook the importance of the post-publication phase. It is also worth keeping in mind that while veteran researchers might not feel like the days following publication are special, sources who are unfamiliar with the investigative process may feel differently. They can feel pressure, might think they have been exposed, or even regret having talked to someone. After publishing a story, keep the following steps in mind. ### 1. Share the story Is the story finally out? Send it to your sources. Don’t let them find it via social media, or hear about it from other acquaintances. Even worse, don’t let them find out the story has been published only days later. Thank them once more for their contribution to the investigation and explain to them the story’s significance. ### 2. Ask for feedback Don’t just send your sources the story and leave it at that. Ask them what they think of it. While you may not be in a position to make major changes after publication, it’s important to learn how they viewed their contribution to the story. Are their concerns properly framed in it? Do they feel that your portrayal of their reality was just? ### 3. Remain accessible Journalists and researchers often disappear after publication, which can make vulnerable sources hesitant to trust not just them, but all journalists in the future. Be reachable to your sources, follow up on their wellbeing, and do not vanish. Working with people on the move in Greece for several years now, I often have sources of stories asking me why other journalists stopped replying to messages after they had opened their houses to them, had coffee with them, or shared childhood photos with them. Sources who have experienced these ‘parachute journalists’, will find it more difficult to trust researchers in the future and will feel that they were taken advantage of or used. <hr class="thick"> #### About the author **Stavros Malichudis** is a reporter and editor. He has worked for the Agence France-Presse and inside story, and has participated in cross-border investigations with Lighthouse Reports and Investigate Europe. He’s member of Reporters United. His reports have been published in European media. He was shortlisted for the European Press Prize ‘21 and won the IJ4EU Impact Award ‘22. In 2019 he was selected as a fellow for BIRN's Balkan Fellowship for Journalistic Excellence (BFJE). He has been trained in data journalism at Columbia University in New York on a fellowship. <hr class="thick"> *Published in June 2025* |
|
None
String added in the repository |
Navigating Sensitive Stories: A Guide to Ethical Engagement with …
Navigating Sensitive Stories: A Guide to Ethical Engagement with …
|
|
None
String updated in the repository |
# Supply Chain and Product Investigations
By Mario Rautner  ```cik-in-short ``` **In Short:** An introduction to supply chain investigations including an overview of the main tools, techniques, data resources and essential precautions to take. It focuses on the main actors, stages and processes of a supply chain and includes a hypothetical step-by-step investigation. --- A supply chain is a set of steps that commodities (goods or raw materials) undergo on their way to becoming products used by consumers or industry. Stages of a supply chain can include places where products are transformed – for instance, an electronics company where circuit boards are made from transistors and copper lines – and where products change hands through transport, such as the loading of circuit boards onto a truck and their shipment to a computer manufacturing company. Supply chain investigations focus on collecting evidence to link each stage of a product’s journey, from the origin of the commodity (for example, the _**[conflict metal Cassiterite](https://en.wikipedia.org/wiki/Conflict_resource)**_ used in the solder of laptops) to the final product (the laptop itself), which can contain hundreds of different components. Supply chain investigations can be particularly impactful because they have the potential to change the behaviours of entire industries when done effectively, with thorough and reliable evidence. A _**[recent supply chain investigation](https://www.cips.org/en/supply-management/news/2018/may/unilever-suspends-palm-oil-suppliers-over-deforestation-claims/)**_, for example, alleged that a large palm oil supplier to major food companies such as Unilever, Nestle and Mars, was conducting massive deforestation and not complying with sustainable palm oil policies. As a direct result, Unilever stalled collaboration with the supplier pending further verification. In another case, in 2018, the British supermarket chain Iceland (Iceland Foods Ltd) _**[committed to removing palm oil](https://www.theguardian.com/environment/2018/apr/10/iceland-to-be-first-uk-supermarket-to-cut-palm-oil-from-own-brand-products)**_ from the manufacturing process of its own-brand products by the end of 2018. Such investigations and findings can have an impact not only on the reputation but also the financial status of the companies involved, in particular when they reveal illegal or controversial activities. Remember, something doesn’t have to be illegal for it to be considered controversial. Even the use of rabbit fur, while legal, can be problematic for a fashion brand these days. In the most successful cases, this change impacts the supply chain all the way back to the origin of the product, where it may prevent social injustice, from poor working conditions and human rights abuses to land grabbing and environmental destruction. Most supply chain investigations are carried out by journalists or NGOs, but they can also be carried out by determined independent investigators or people and communities directly affected by an issue. ## Shining a light on where our products really come from Everything around us is part of a supply chain: from the food we eat to the clothes we wear and the devices we use. But the nature of supply chains means we rarely know how commodities are produced or even where they originate. We might not always consider where our items come from, who harvested their source materials and under what conditions, whether there was exploitation or abuse at any point in the production process, or if illicit acts were carried out along the way. However, these questions might be brought to light when public controversies emerge about certain products or brands. Perhaps you live in an area where important materials are extracted or produced. These materials may then enter the supply chain of a final product through countless contracts, companies and manufacturing processes, which are often opaque or not immediately visible or accessible. Or, you might live near a forest that is being cleared and want to know whether the companies carrying out the clearing are using legal and ethical practices. The impact of such findings can be crucial in saving the forest. Of course, it is not always an investigator living in that area who will expose the truth, but having someone familiar or close by to observe or launch the quest for more information is a good start. Often a few initial questions and some raw evidence can lead to wider investigations by another group or organsation. Equipped with awareness of the risks and safety considerations, and a good assessment of what is at stake, you can also start this way. Even if you are not physically present at a location where commodities are produced or extracted, many supply chain investigations can be carried out remotely. Indeed, in most cases, elements of a supply chain are located far from the source - for example, wood processing and furniture manufacturing can take place continents away from where the forest was originally cleared. ```cik-tip ``` >**Tip: Helpful tools anyone can use to track products** > >**Barcodes:** Barcodes on consumer products can be used to identify the manufacturer, as well as additional >information about them. Though barcodes are most familiar to us from a retail environment, they are also >used in many other areas of business. In retail, the most commonly used barcodes are the _**[European >Article Numbers / EAN](https://www.gs1.org/standards/barcodes/ean-upc)**_, which are usually 13 digits and >used both >within and outside of Europe; and the 12-digit Universal Product Code (UPC), which is very similar to the >EAN, but contains a zero at the beginning to signify registration in the USA or Canada. The first three >digits of an EAN indicates the name of the company that accepted the registration of the code (note that >the location of the company may be, but not necessarily, the same as the location the product was >manufactured). Companies that accept registration are members of the GS1 non-profit organisation, which >maintains the global standard of barcodes. The [GS1 website](https://www.gs1.org/) has a _**[searchable database](http://gepir.gs1.org/index.php/search-by-gtin)**_ of many (but not all) barcodes. In some cases >this is useful for obtaining information on who registered the product, and can also reveal details about >a company's ownership structure. > >**Food hygiene labels:** >EU rules regarding _**[food hygiene](https://ec.europa.eu/food/safety/biosafety/food_hygiene_en)**_ cover >all stages of the production, processing, >distribution and sale of animal products intended for human consumption. This includes, for example, fish, >meat and dairy products, but also non-food animal byproducts such as pet food. Every individual >establishment or outlet (rather than the parent company) that puts such products on the market has to have >a registration number that uniquely identifies it. The codes can be found on the product label and consist >of a combination of letters and numbers within an oval. Have a look at a product at home to see how easy it >is to spot the label. The EU has an online database of the codes, making it possible to identify the name >of the company that put the product on the market. Since the regulation covers imports and any processes >that take place in the EU, the database includes details of companies from over 80 countries. This comes in >very useful if you’re investigating the chain of custody of a food product manufactured by a large consumer >company on a contract, or a brand owned by a supermarket. In such cases, you can identify the actual >producer rather than the brand owner. ## What makes a supply chain investigation  As you prepare for your investigation it’s important to have an informed idea of what is possible to achieve, what is relevant and what may be out of reach. Supply chains are not always simple; they are often complicated networks or webs of actors, processes and movements. While it takes thousands of links in multiple supply chains to manufacture one laptop, investigations usually focus on very specific components of a supply chain, such as the source of one single element used in the manufacture of one part of the laptop, or a single factory where one stage of the production takes place. Maintaining such a focus makes the research more effective. Supply chain research can be instrumental in linking a company to ethical issues that taint their products, even though the physical components of the product itself may not be controversial – for example, the violation of indigenous land rights that often occurs when producing rubber for car tires. A key concept to remember about supply chains is **traceability:** the passing of information from one stage in the supply chain to the next. This can relate to the initial source of the ingredients or raw material of a product, the alterations it incurred along the way, or other relevant details that may help trace an entire history of transportation and production. Some companies have systems for this but many supply chains are so long and complicated that it can be difficult even for the companies directly involved to achieve a fully traceable supply chain. Sometimes, **certification schemes** are used to ensure that products are manufactured, produced and traded following specific standards. For instance, they can indicate that the human rights and traditional land rights of local and indigenous people are observed or that no tropical forests are cleared to make a specific product. Well known certification schemes include those for organic and fair trade produce or for sustainable timber, palm oil or coffee.  *Fairtrade certificate logos. Image from Pinterest: https://www.pinterest.com/pin/110619734567839836/* In order to carry out supply chain research, it is necessary to get a good understanding not only of the product you are trying to track and its production prices, but also about **company finances** and **trading** and **transport processes** in different countries. **Re-evaluate** your assumptions and **verify** your data at every step. As with any investigation, you should always question your initial assumptions and results in order to make sure that the information you collect is relevant to what you intended to prove. What at first appears likely may in fact have another explanation. For instance, you might have linked a company illegally clearing forests to a timber mill that sells to a furniture company, only to find out that the furniture company does not actually use the timber species harvested from the forest in question. You therefore cannot make a direct connection between the company and the illegal deforestation activity. The **flexible** and **unpredictable** nature of supply chain research also means that it is not always possible to anticipate how long your research will take. Sometimes linking two processes can take a day of online database work; sometimes it can take months and dozens of research strategies and tools, including online and field research. Due to the complexities of supply chain research and the legal, ethical and safety risks involved in accusing large corporations of wrongdoing, if you decide to conduct this kind of investigation, it’s best to take your time to slowly expand your skills and understanding of supply chains. Starting with basic searches and testing some of the tools and techniques we present here would be an easy way to introduce yourself to the field. How and where do I start? ------------------------- You can start your investigation at any point along the supply chain where you suspect unethical or illegal behaviour might be taking place. In most cases, investigators follow the trail downstream along the supply chain from this entry point. For instance, if an investigation starts at a mobile phone assembly plant suspected of using child labour, it will most likely continue downstream to the brand that sells the mobile phones, rather than upstream to the origin of the plastics and metals used in the phones’ manufacturing process. The focus here will be to expose illegal and unethical practices of exploiting children to produce that phone. In some cases, however, research can occur upstream. An example might be the case of an ill-famed company establishing a manufacturing plant in a town whose residents want to find out if the products and commodities entering the plant are of controversial origin. Once you identify an entry point, you can follow a sequential path along the supply chain to collect evidence connecting each step, be it along a production process or a product’s transport path. That said, your research will most likely be nonlinear, and you will have to use multiple tools. Depending on what you are investigating, your work will involve extensive online research, and sometimes field research. You might need to use anything from maps to identify the source locations of raw materials or interactive transport tracking services to follow commodity shipments, to online customs and product databases or corporate records and stock exchange websites to learn more about the products and companies you are focusing on. You might also sometimes undertake field research and collaborate with others to collect evidence and witness accounts about what happens on the ground. Remember that companies tend to tightly guard information about their suppliers and customers to avoid exposing their internal mechanisms and advantages to their competitors. Companies also sometimes hire investigators to research the supply chains of competing products and of other companies. Good supply chain investigators look for evidence that links the chain, and this often requires a lot of creativity since no two supply chains are the same and access to information can vary vastly depending on the data sources available or the geographies where research is being conducted. The ability to ‘look sideways’ by making previously unseen connections is a critical skill, as is being able to come up with innovative solutions to research problems. ## Mapping and profiling actors When thinking about supply chains, a useful first step is to understand the various actors – the participating companies or individuals – that operate along the way. While the actual chains are usually unique combinations, some of the actors can be categorised into groups, which are often connected by shipping or transport providers. Most often, along a supply chain you will encounter the following: ```cik-definition ``` > **Producer:** company, person or group of persons > that takes, grows, mines or otherwise produces the raw materials (such > as the owner of a timber plantation). > > **Initial processor:** company that carries out the first > transformation of the product (for instance, a timber mill turning a log > into planks). > > **Further processors:** companies that carry out additional > transformations of the product (wood can ultimately even turn into fiber > for textiles). In many supply chains there are multiple further > processors, while other supply chains may not have any. > > **Importers, exporters, distributors:** companies responsible > for getting the products into different countries, by operating or > hiring shipping or trucking services to carry the products (for > example, shipping the wood planks to a deposit in country X, from where > it can be sold to furniture manufacturers and others). > > **Manufacturers:** companies that carry out the last > transformation before the product is sold to consumers or industrial > users (such as the company making furniture or toothpicks). > > **Retailers:** companies and individuals responsible for > selling the products to consumers or industrial users (like a hardware > shop or furniture store).  *Simplified example of supply chain actors and their connections* Let’s assume you’ve chosen your entry point in the supply chain: for example, a local fabrics manufacturer where you suspect that workers are mistreated and underpaid, and you want to find out what fashion brands buy these fabrics and where they are sold. You’ve also mapped the actors you will focus on: the manufacturer and the retailer (plus, potentially, the final consumers), so you can raise public awareness about your investigation findings. Now you know what and whom you are investigating. The next step is to establish a list of companies and people of interest – your actual actors. Before you move any further, begin by researching their company ownerships, business models, networks of collaborators, places where they are registered and/or operate, products they manufacture or sell, and possible controversies surrounding them already. By doing this background research you create a profile of your main actors and their connections, build your foundation for further investigation, prepare yourself for interviews with relevant people (sources, witnesses, specialists etc.), establish connections and assess your potential risks. Internet research is a good place to start looking for basic information about businesses and people associated with the companies in the supply chain you're investigating. Try to find out as much as possible from companies’ websites and activity reports, supply chain due diligence reports (if available), stock exchanges (if the company is listed), media articles or social media profiles of the company, its board and staff. At the same time, look for official documents that include the companies’ registration data, shareholders, subsidiaries, members of the board, directors, annual financial reports and other relevant details. If you already know in which countries to look for official documents, you can start by checking available online corporate registries and other official databases, such as land records, court records, patent registries etc. There are a number of platforms to help you start this research. Some corporate registry websites may seem intimidating at first, but don’t be daunted: there are more user-friendly, convenient and extremely helpful ways to get introduced to the company research field. _**[The Investigative Dashboard](https://id.occrp.org/)**_ (ID) is a resource built by the [Organized Crime and Corruption Reporting Project](https://www.occrp.org/en), providing a global index of public registries for companies, land and courts. From this platform you can go to the country and type of records you want, but note that while some corporate records provide free access to basic or more advanced information, many will require registration and fee payments in order to provide detailed company records. ID often marks the resources as free or paid for, so you know what to expect. In addition, the platform allows investigators to freely search and use its _**[database](https://data.occrp.org/)**_ of millions of documents and datasets from public sources, leaks and investigations it has conducted in the past. Chances are that you might find good leads in there too.  *Investigative Dashboard database: https://id.occrp.org/. Screenshot by Tactical Tech.* There are other similar resources where you can continue and expand your research. _**[Open Corporates](https://opencorporates.com/)**_ is a platform that lets you search freely for company records and related data it collects from all over the world. [The International Consortium for Investigative Journalism (ICIJ)](https://www.icij.org/) runs the _**[Offshore Leaks Database](https://offshoreleaks.icij.org/)**_ a massive resource of company records and other useful documents that have been revealed by large leaks and projects the group has been working on, including the Paradise Papers, Panama Papers, Offshore Leaks and Bahamas Leaks. Do not forget about the power of social media, and especially professional networks such as LinkedIn, where you can search for people and companies as well as identify possible connections among your subjects of interest. Elsewhere in this kit, you can learn how to build a strong body of evidence, how to investigate companies and people, how to do online and offline research and how to use various tools and techniques such as maps, social media, interviews etc. You can combine resources and develop new skills with each issue you investigate, based on the demands of your particular context, while also making sure to remember the safety and ethical considerations that go hand in hand with your plans. ## Inspiration and guidance: a sample investigation This section uses a fictitious example to show you the kinds of tools and approaches you might use to carry out a supply chain investigation. While the example is relatively realistic, the [chain of custody](#term-chain-of-custody) has been simplified and shortened for the sake of illustration. Unlike the product itself, chain of custody research cannot follow a simple research recipe, so the goal of this example is to demonstrate the need to think creatively and adjust to each specific case. ```cik-note ``` >**Note:** all the names of the entities involved are fictional, and >any relation to real companies or products is coincidental. ### Sparked by a story of injustice You read a news report that details how villagers in a South American country have been rounded up at gunpoint and forced to work in a bauxite mine run by a company named Brown Gold. You want to find out if the minerals from the mine are being used in well known consumer products. You start your investigation by focusing on the supply chain at the point where the mineral leaves the mine or company warehouse. This is your entry point. Initially, you might begin with simple online research. Keep in mind that this case could be sensitive, so it’s useful to familiarise yourself with and start using some digital security tools, such as the [Tor Browser](https://www.torproject.org/projects/torbrowser.html.en), [DuckDuckGo](https://duckduckgo.com/) search engine or a [VPN](#term-vpn) (Virtual Private Network), to avoid being tracked. Your search does not yield much more than the name of the company itself, but in the process you gather additional information, such as the company’s registration number and the names of the founders and key people in the company. You also carry out online research and ask some specialists you trust on related subjects such as: - The uses of bauxite (as a main source of aluminium). - Detailed information about the aluminium production process and its impacts. - The largest companies involved in global bauxite and aluminium production. - The main uses for aluminium in industry. ```cik-example ``` >**Example of online research entry points:** > >In many instances, [Wikipedia](https://www.wikipedia.org/) provides useful entry-level information and links to external sources that you can check for more in-depth knowledge. Specialised websites that focus on the particular minerals are helpful for understanding the field and its main players. >In this case, for example, a DuckDuckGo (or a Firefox, Google etc.) search for “aluminium and bauxite” leads you to pages like that of [Geology.com](https://geology.com/minerals/bauxite.shtml), >[Aluminum.com](https://www.aluminum.org/industries/production/bauxite) (Aluminum Association in the US), [Australia’s Aluminum Council](https://aluminium.org.au/interactive-flowchart/bauxite-mining-chart/) and many others, where you read about the production process and uses of this resource. > >Searching for “bauxite producers world 2019,” for instance, takes you to information on the current _**[top bauxite producer countries and companies](http://www.thedailyrecords.com/2018-2019-2020-2021/world-famous-top-10-list/world/largest-bauxite-producing-countries-world/6882/)**_. > >Searching for “aluminium producers world 2019” leads you to other resources such as a [map and downloadable database](http://www.world-aluminium.org/statistics/) of top aluminium producers from international institute [World Aluminium](http://www.world-aluminium.org/) and additional data from statistics aggregator [Statista.](https://www.statista.com/statistics/280920/largest-aluminum-companies-worldwide/) ### Investigating the key players Your preliminary online research provided key information about the products and actors at the centre of your investigation; enough to build some background knowledge and basic details about your topic and main subjects (producers, buyers etc.). You then continue by mapping your actors and the connections between them – who is the producer, who delivers the commodity to whom, by what means etc. – and following the supply chain downstream from your entry point. #### The Producer: MINING COMPANY Since the company – Brown Gold – is mining the resource, it is considered the producer of the commodity. By using the resources mentioned above, you can gather various types of information to get a better understanding of who Brown Gold is: - Physical company operations and locations – source: maps and satellite images (potential transport routes). - Type of company (private, public and, if so, shareholders, etc.) – source: company registries, stock exchange, website information. - Financial information – source: company filings, annual reports on company’s website, stock exchange (if the company is listed on any stock exchange). - Registered office addresses (used for formal purposes, e.g. where mail can be received) and trading or operational addresses (where actual activities occur) – source: online and offline corporate records, company’s website. - Board members and directors (and any other businesses they might be involved in) – source: company registries, company website, company and people databases (see above for Investigative Dashboard and other resources) - Bank connections, loans, mortgages – source: corporate records, court records, annual financial reports. - Employee feedback, reviews and grievances on sites like [glassdoor.com](https://www.glassdoor.com/Reviews/index.htm), [comparably.com](https://www.comparably.com/), [careerbliss.com](https://www.careerbliss.com/), sometimes yield information about the company. - The current careers or job opportunity page and archives of job postings from _**[archive.org](https://archive.org/web/)**_ may yield information about projects, locations, and future plans of the company. ```cik-note ``` >**Note:** > >Carefully searching through company reports and other documents can lead to very useful findings. >Annual reports, corporate sustainability reports, company presentations and shareholder filings can all be useful not only for basic information on a company and its finances, but also, in some cases, about its customers (photos within these reports can be especially useful for this). > In this case, you find the location of the company on satellite images and identify its board members and directors, but cannot find any further information about them. Nor have you found detailed financial data or names of any customers of the company. Searching through company records and annual reports, you do, however, manage to find a subsidiary (a company owned and controlled by another company) of Brown Gold, by the name of BG Shipping. #### The Initial Processor: ALUMINIUM SMELTER From your initial online research you found out that bauxite is used to manufacture aluminium, which in turn is produced in large smelters (facilities where the metal is extracted by heating and melting the ore). You also discovered that smelters are massive investments – to the extent that there are fewer than 150 of them globally. This information will help you in your investigation, because it offers a relatively narrow field of inquiry. In order to find one of the smelters that is supplied by Brown Gold, you use a process of elimination to rule out certain routes, while looking for likely connections between the mining company and certain smelters. Examples of investigation activities to do this include: - Finding out if there are any smelters located in the country the mine operates in – there are none. - Identifying other bauxite mines in the country – you find three more. - Analysing whether this country exports any bauxite overseas using the UN Trade database [Comtrade](https://comtrade.un.org/) -- you find that six countries received bauxite. So far you know of four mines in total; this means you can’t yet determine which country is supplied by Brown Gold.  *Sample comtrade.org search. Screenshot by Tactical Tech.* In order to identify the smelter, you need to first determine the country that receives the bauxite. With a relatively simple online search (e.g. "how is bauxite transported") you learn that bauxite is most commonly transported by ships or trains. You carry out further online research, and check transport routes and map locations (using Google Maps or Google Earth) to map out deepwater seaports. Then, by looking at the train lines between the four mines and the location of the mines to each other, you conclude that Brown Gold is most likely exporting its bauxite using ships. Moreover, you previously identified the BG Shipping subsidiary and were able to locate its warehouses on the map at a port that is connected to Brown Gold via train track. You can therefore conclude with even higher certainty that the company exports its bauxite from this port. #### The Trader: SHIPPING COMPANY Today, the majority of commodity-carrying ships are fitted with live tracking devices, which employ what’s called an Automatic Identification System (AIS). All vessels over 300 tonnes (gross) are required by law to emit radio signals with their speed, direction and exact location. This data is also captured by receivers on other vessels and by stationary receivers in ports and on coastlines. Services such as _**[Marine Traffic](https://www.marinetraffic.com/)**_, _**[Vesseltracker](https://www.vesseltracker.com/)**_, and others use AIS information to produce live maps showing the movement of ships across the globe. To search the map, you need to know the name or registration number of the vessel you are tracking. For additional resources and depending on your needs, other tracking services may be of further help, so reviews like this from [Marine Insight](https://www.marineinsight.com/) - _**["Top Eight Websites to Track Your Ship"](https://www.marineinsight.com/know-more/top-8-websites-to-track-your-ship/)**_ - may help when looking for alternatives. Most of the ship tracking services either have a free version or offer a free trial, which can be used to track these vessels. They also tend to employ their own network of receiver stations and have maps on their websites that indicate the network coverage. MarineTraffic, for instance, provides near real-time information about vessels’ positions and voyage-related information. In addition to the AIS receivers, which are ground-based, Marine Traffic (and other such platforms) also collects vessel data from satellite receivers. Obtaining this data often requires paying a one-off fee or signing up for yearly subscription, which is usually cheaper if you plan to use it longer.  *Sample MarineTraffic.org search. Screenshot by Tactical Tech.* ```cik-Tip ``` >**Tip:** > >Always search more than one vessel-tracking website to trace the same ship, as some platforms may provide >you with more free information than others, and some may have the data you need while others may not. If there is an area of particular importance to your investigation that is not covered by these services, it’s even possible to buy your own receiver online, from places like Amazon or specialised websites such as _**[Milltech Marine](https://www.milltechmarine.com/)**_. If this is the case, do some serious research first: check what range you need to cover, what other users recommend, verify equipment reviews, consult forums like _**[ShipSpotting](http://www.shipspotting.com/)**_ and maybe even check with some vessel tracking enthusiasts for advice and recommendations. ```cik-note ``` >**Note:** > >Taking this route requires some financial cost, a stable Internet connection and power supply in the place where you are installing the receiver, knowledge of how to use the equipment and some extra effort to maintain it. However, if it helps to advance your investigation and you have no other options, it’s well worth the relatively short time it takes to learn and experiment. Marine Traffic provides useful advice for such situations in its [FAQ section](https://help.marinetraffic.com/hc/en-us/articles/205326387-What-is-the-cost-of-an-AIS-Receiving-Station-). >The website also offers free receivers to those who want to contribute to its [live map](https://www.marinetraffic.com/en/p/expand-coverage), but that requires you to share your receiver’s data >with the platform, something you might or might not want to do, depending on how sensitive your >investigation is. If you decide to go ahead and operate a receiver, you also get a free subscription plan, or an upgrade to your existing account. After deciding which tracking site(s) to use, you spend several weeks mapping out the journeys of vessels that appear to have been loading materials at BG Shipping. Ships that spend several hours berthed right next to the BG Shipping facility in the port are more than likely loading or unloading. You can track journeys of vessels by saving their movements on some of the above mentioned online platforms. In Marine Traffic, for example, you can create your own "fleet" of vessels to monitor and get updates on, or set up notifications when a vessel reaches a port. You can also access past travel information in the database - a service that is sometimes free, other times for a fee. Eventually, by tracking these ships, you identify three ports in three different countries across Asia that received exports from BG Shipping. This means that, at the very least, three separate smelters bought bauxite from the Brown Gold mine. Using Marine Traffic you manage to show that a vessel that left BG Shipping went directly to the port facility of a company called Caliper Alumina. You conclude that Caliper Alumina is one of the smelters receiving bauxite from Brown Gold mine. Think back to your earlier research, when you identified and mapped smelters, their transport routes, subsidiaries, key personnel, addresses and finances. You can now apply the same techniques and resources to find out more about Caliper Alumina and its activities. ```cik-tip ``` >**Tip:** > >Vessels sometimes make multiple stopovers. It’s also not always clear from sites like Marine >Traffic whether a vessel is loading or unloading. Look for additional information about the vessel’s draught (the depth its hull is submerged into the water), which provides another indicator of whether it has been loading or unloading (the higher the draught, the heavier the ship). > #### The Further Processor: CAN PRODUCER In this example, you can use another resource, namely *customs data*, to try to identify one of the customers of Caliper Alumina. Customs data consists of shipping manifests, which every shipment vessel is legally required to carry while it travels around the world. Governments collect this data in order to produce their trade data statistics. It’s possible to purchase this data, but doing so is expensive and only available for a limited number of countries. Generally it provides the name of the importer and exporter – so if one country in the trade does not make customs data available for purchase, you may be able to get the same information from the other country involved in the transaction. The cost of purchasing customs data has decreased lately, in line with a growing number of companies offering it for sale. Examples of such services include [Trade Atlas](https://www.tradeatlas.com/en), [Export Genius](https://www.exportgenius.in/export-import-trade-data/), [Xportmine](https://www.xportmine.com/), [ListThe](https://www.listthe.com/), [Panjiva](https://panjiva.com/) etc. The first step in this process is knowing which data to buy, and to do that, you need some context and background information. Products are usually organised by HS (harmonised system) codes, a nomenclature developed and maintained by the _**[World Customs Organization](http://www.wcoomd.org/)**_(WCO). The codes are organised into 21 sections, which are subdivided into 97 chapters, in turn further subdivided into several thousand headings and subheadings. The most recent HS codes edition, which we use in this section as of 2019, is [the 2017 HS Nomenclature](http://www.wcoomd.org/en/topics/nomenclature/instrument-and-tools/hs-nomenclature-2017-edition/hs-nomenclature-2017-edition.aspx) (note that a new one will come into place in [2022](http://www.wcoomd.org/en/topics/nomenclature/instrument-and-tools/hs-nomenclature-2022-edition.aspx)). **Here is how it works:** Usually each subdivision comes in a two-digit increment. The case in our example: Section XV "Base Metals and Articles of Base Metal" lists a series of chapters numbered from 72 to 83. - Chapter 76 is Aliminium and articles thereof. This is followed by 16 headings, for example: - Heading 7606: Aluminium plates, sheets and strips... - Heading 7612: Aluminium casks, drums, cans, boxes and similar containers (including rigid or collapsible tubular containers), for any material (other than compressed or liquefied gas), of a capacity not exceeding 300l, whether or not lined or heat-insulated, but not fitted with mechanical or thermal equipment. - Subheading 7612.10: Collapsible Tubular containers.  *Base metals codes from World Customs Organization wcoomd.org. Screenshot by Tactical Tech.* When buying customs data, most of the time it is necessary to know the HS codes of the products in question. Often HS codes are defined down to eight or 10 digits, but at these levels the classification is usually set by the countries rather than the WCO. In addition, codes are reviewed and sometimes adjusted every few years, including those of the World Customs Organization above, so make sure you stay updated. ```cik-note ``` >**Note:** > >Each product and shipment that enters a country has an HS code attached >to it, so both domestic and international trade statistics are usually >based on those HS codes. For instance, the United Nations maintains the >_**[Comtrade database](https://comtrade.un.org/)**_, which contains such statistics - >[comtrade.un.org/data](https://comtrade.un.org/data) - that can also >prove useful for chain of custody investigations. In many cases, customs data comes in spreadsheets and often contains false positives or multiple spellings for the same company name. Depending on the purpose of your investigation, you might need to clean the data, especially for statistical analysis. The details of the data vary depending on the country that supplies it, but it usually includes: - the names of the exporter and importer, - the date of the shipment (at least the month), - the weight of the shipment, - the HS codes, - the port of loading, - the port of discharge. Additional data fields can include addresses, phone numbers, detailed product descriptions, the value of the shipments, the name of the vessel and shipping company, as well as the names and addresses of notifying parties (the buyer or importer who receives notice of the shipment’s arrival). In this case, the data shows that a US company by the name of Beverages Inc. purchased aluminium sheets from Caliper Alumina. You also find a number of older shipments made by that company, which you are able to verify with historical Marine Traffic data, by tracking the vessel’s past movement as detailed above. You also notice that the Beverages Inc. website features photos that show cans of soft drinks produced by prominent brands. #### The Manufacturer: SOFT DRINK COMPANY One of the companies whose cans are displayed on the website of Beverages Inc. is called Spark Drinks. While you might be tempted to link this well known brand to the controversial bauxite producer, more research is necessary to ensure that this supply chain is indeed linked. There are various reasons why they may not be connected. There could, for instance, be different grades (strength, concentration) of aluminium in different lines of cans used for different brands. So, there is a chance that Sparks Drinks may not be using the aluminium made from the bauxite sourced from Brown Gold’s mine. This is where the problem of *traceability* emerges. ```cik-note ``` >**Note:** > >Supply chain *traceability* is a process and technique that ensures the >integrity and transparency of a company’s supply chain, by tracking >products from origin to consumer. This reduces the risk of mislabelling >commodities that have been mixed from different suppliers and regions, >allows products to be tracked back to their source and helps identify >particular batches of the product if something goes wrong and products >need to be tested or recalled, etc. Traceability is a key challenge in >supply chains. > >In certain cases, however, commodities and products are *segregated* >(they have their *identity preserved*), which means that companies can >trace the product and its components all the way back to the origin and >the producer. This segregation is often required for certified products >(e.g. those who must meet strict quality standards) and animal products >(for instance, to identify farms in case food safety issues arise). > >In most instances, though, raw materials or components from different >sources (e.g. from different mines and suppliers) are mixed together at >various points in the supply chain. The result is that, due to a lack of >segregation and traceability, most companies do not know for sure >whether they are connected to actors involved in illegal, unethical or >controversial activity. From an investigator’s point of view, this means >that all actors downstream of where the mixing occurs are part of the >supply chain in question. > >For companies, traceability is important for a number of reasons. These >include compliance with policies of sustainable sourcing (for example, >making sure they do not contribute to famine, drought or environmental >damage) or with conflict minerals regulations implemented by various countries, which impose strict rules >on the purchase and use of commodities obtained from areas controlled by >armed groups or where abuses take place. Such aspects are important for a company’s accountability >to regulatory authorities and to their consumers. There are a number of research techniques you can use to find out if mixing occurs along the supply chain. Sometimes the *company information* (websites, public filings) includes details about where the company purchases commodities from and if it keeps a detailed log of its own supply chain from the sourcing of materials until the point these reach its ‘hands’. In other situations, searching online for *scientific articles* or other investigations on your topic or commodity of interest, as well as more general magazines such as _**[Supply Chain Quarterly](https://www.supplychainquarterly.com/)**_, and available *industry* *research* might shed light on how the sector operates in general. In addition, NGOs like [Greenpeace](https://www.greenpeace.org/international/), [Rainforest Action Network](https://www.ran.org/), [Rainforest Alliance](https://www.rainforest-alliance.org/), [Mining Watch](https://www.miningwatch.ca/) and many others already conduct research on various supply chains and their impact on human rights, the environment, and the wellbeing of communities living and working along the chain. Contact them for advice if you find their knowledge useful. ```cik-tip ``` >**Tip: using NGO reports** > >The techniques and resources outlined here are often used by >NGOs working to pressure companies into changing their ways of >production or cleaning up their supply chains. NGO reports can offer >insightful and inspiring examples of how such tools have been used in >conjunction with other supply chain research techniques. When reading reports, keep in mind the >mission of the NGO. Some of the language >may not be appropriate for investigators or journalists and some texts are >designed for advocacy and policy purposes. However, it’s an interesting >exercise to try and figure out how the research was carried out and what >investigation tools and strategies were applied. Also look out for *trade and industry magazines* and websites that profile the way your company of interest operates (see [supplychain247.com](https://www.supplychain247.com/), for example). However, keep an eye out for articles and research that appear one-sided or solely focused on positive angles – this is a strong clue that they may be sponsored by the very companies you are investigating, and thus not always reliable or complete. Search multiple sources and read as much as possible to create a complete picture. You might also consider *talking to industry experts* or even to the company itself. ```cik-warning ``` >**Warning:** > >Before talking directly to the company of interest, you should conduct >thorough background research on their operations, and be prepared to ask >highly relevant questions rather than details you might already find >elsewhere. If you plan on accusing the company of something, it makes >sense to avoid telling them so, or even closely hinting at that, before >you’ve gathered solid proof. Powerful companies will go a long way to >make you stop your research, try to discredit you or sometimes even >threaten you with legal action or other types of >pressure. Now, back to our scenario. You find out from industry magazines that mixing occurs at the aluminium smelter and that all products leaving the smelter contain bauxite from various sources. Since this question is crucial, you also confirm it by carrying out interviews with industry experts and eventually also with representatives of the smelter itself. This means that any product containing aluminium that originates from the smelter, including the products of Spark Drinks, is potentially contaminated with bauxite of controversial origin. #### The Retailer: SUPERMARKET In this supply chain, the last step of identifying retailers that sell the products to consumers is relatively simple, since most retailers in your country have online stores and stocked products can be found via the retailer websites – as well, of course, as in person by visiting the stores. This allows you to complete the supply chain and link the bauxite to a number of major supermarket chains. ### Loose ends and further questions Even though the case study ends here, there are many potential questions still to be answered to increase your certainty in the information. These include: - Are the customers listed by Beverages Inc. still up to date? - Is it possible that the bauxite and aluminium are of grades that rule out the smelter you identified, and thus render the particular chain impossible (and therefore another smelter would have to be identified)? - Does Beverages Inc. export to other countries? - Are there ways to confirm the supply chain links using other research techniques or further verify the results you obtained so far? You can employ numerous different research techniques to identify customers of the aluminium smelter and, depending on how many more transformations take place, there may still be more chains to uncover before you reach the final consumer product. For electronic products, it is likely that the aluminium will change hands many times before it can be found (unrecognisable in shape and form) in a store. In some cases, due to the nature and complexity of supply chains, you’ll have to invent or significantly adapt the research techniques as you go. ## Risks and considerations Supply chain research often comes with significant risks, many of which depend on both the research strategies employed and the geographies in which they take place. Because this type of research is often connected to and can effect the profits, loss, and earnings of companies it makes sense to have an understanding of how to secure your research and work pseudonymously. If you decide go public with your findings timed release of key research (instead of releasing everything at one) is a method of dealing with crisis management and public relations teams. Developing a methodical and meta data preserving way to catalog data, notes, research and files is key. Developing a workflow that involves having an encrypted backup is recommended. If you are technical, look for ways to split decryption keys into _**[(Shamir Secret) Shards](https://en.wikipedia.org/wiki/Shamir%27s_Secret_Sharing)**_. This allows you to split the key into several pieces, and only if a certain number of pieces come together does the key reform. Look at projects like [ssss](http://point-at-infinity.org/ssss/) and [storaj](https://storj.io) for examples. Hashes of these shards can be used to create so-called "insurance files" which may be of use to the investigator, however is out of scope of this section. ### Safety concerns Successful supply chain research often requires field research in addition to online data collection and shipments tracking. Field research can be high-risk, especially in some areas where corruption is rife, where investigators and journalists have little protection, or where violence occurs frequently. This kind of work is usually carried out in teams and under detailed plans for physical and digital security. High-risk work should only be undertaken by experienced investigators. ### Data access Companies usually have multiple customers and suppliers, often hundreds. There are, therefore, many forks in the road when it comes to following the trail of the products you are investigating. Make such choices carefully and base them on knowledge, facts and information rather than on gut feelings. An assessment of the difficulty of research and operating in specific jurisdictions for instance is a rational way of making such decisions. It is important to give special consideration to supply chain bottlenecks (for example, a blockage in the chain caused by shortage of a commodity or increased demand), where only a small number of companies may have overwhelming market shares. For instance, a handful of traders now dominates the global trade of agricultural commodities. In most cases it is not possible to prove the supply chain for a specific product all the way to the end product. This diluting of the chain is common and is not necessarily a concern. For example, it would be very difficult to show that one harvest of coffee beans from a smallholder (a non-industrial farmer harvesting a small area) who has illegally cleared forested land, ends up in instant coffee products of a large consumer brand company. However, it might be possible to prove that the smallholder sells to a specific agent, who sells the coffee to a company where the beans are collected and washed. This company in turn sells to a processing and packaging company that ships the coffee overseas to an importer, who passes it on to the coffee roasting and grinding company, which in turn sells it to a large consumer brand that packages the instant coffee products, which are sold in supermarkets around the world. In this scenario, the coffee coming from the illegal plants can never be traced directly and has been mixed with coffee from many other plantations, yet the consumer brand is still exposed to the problem because it is part of the supply chain. ### Legal issues Supply chain research requires a very high standard of proof. If the evidence at any point of the chain is not strong enough, additional investigations are required to corroborate your theory or conclusion. This type of investigation often ultimately exposes very large corporations for whom reputation is as valuable as the products themselves, and they often respond strongly when their reputation is under threat. Be aware that it is not uncommon for lawsuits to ensue from investigations on supply chains, and there are significant risks for investigators even when the information they provide is accurate. Supply chain investigators frequently obtain legal advice before publishing the results of their work, which often exposes powerful companies and shady practices. Investigators may find themselves accused of libel or slander; countries have different laws and potentials for legal issues so asking for legal opinions is useful even if only as a precaution. Organisations and individuals engaged in this research can also face SLAPP (Strategic Lawsuit Against Public Participation) suits. These are designed to use the companies’ large funds to engage investigators in lawsuits in order to intimidate them and delay or stop them from carrying out their work, even if the lawsuit has very little or no legal merit and chance of success. ### Collaboration In order to avoid or mitigate many of the risks above, investigators often collaborate with others doing similar work, including asking for advice and best practices on physical and digital security. Collaboration is particularly advisable when investigating certain topics for the first time, and additional knowledge and expertise is required. Don’t shy away from asking questions to trustworthy organisations and people, and avoid the temptation of trying to take down a company with your very first investigation. Collaboration also help share the workload, so you don’t have to follow the entire supply chain by yourself, but instead work with local communities, experts, NGOs and others to collect stronger evidence and reach a more impactful result. Some countries may have freedom of information laws which allow citizens of that country access to information not available to the investigator. It may make sense to collaborate with a citizen who can access other information and at lower risks. When collaborating it makes sense to take extra care to encrypt files/folders and get hashes (fingerprints) of files to look for changes. The kind of "source control" software a developer uses ([github](https://github.com/), [bitbucket](https://bitbucket.org/product), etc.) may aid you in doing this and keeping track of files, edits, and ownership. When working in a team be sure to have a secure and encrypted way to communicate. Messenger apps like [Signal](https://signal.org/en/) or [Wire](https://wire.com/) are recommended. <hr class="thick"> _Published April 2019_ ## Resources ### Articles and Guides - *[Supply Studies Research Guide: A Research Guide for Investigations in the Critical Study of Logistics](https://supplystudies.com/research-guide/)*, by Matthew Hockenberry, Ingrid Burrington, Karina Garcia, and Colette Perold, 2025. - *[Destroying elephant habitat while breaching the Indonesian Government moratorium on forest clearance for palm oil](https://www.ran.org/wp-content/uploads/2018/06/RAN_Leuser_Watch_PT_Agra_Bumi_Niaga.pdf)*, from Rainforest Action Network. A supply chain investigation report. - *[Eating up the Amazon](https://www.greenpeace.org/usa/wp-content/uploads/legacy/Global/usa/report/2010/2/eating-up-the-amazon.pdf)*, from Greenpeace. An investigation report on tracking soy from the Amazon. - *[Investigating supply chains](https://gijn.org/investigating-supply-chains/)*, from the Global Investigative Journalist Network (GIJN). A short guide including tips, resources and techniques. - *[Investigating illegal timber](https://www.earthsight.org.uk/tic/guidebook)*, from the Timber Investigations Center. A guidebook to researching timber supply chains. - *[Top eight websites to track your ship accurately](https://www.marineinsight.com/know-more/top-8-websites-to-track-your-ship/)*, from Marine Insight. A brief review including tools comparisons and tips. - *[Who's Got the Power: Tackling imbalances in agricultural supply chain](https://web.archive.org/web/20151117052506/http://www.fairtrade-advocacy.org/power/183-projects/psc-main-page/870-the-report-on-imbalances-of-power-in-agricultural-supply-chains)*, from the Fair Trade Advocacy Office (FTAO). A study about power concentration and unfair trading practices in agricultural supply chains. ### Tools and Databases - *[EAN/UPC barcodes overview](https://www.gs1.org/standards/barcodes/ean-upc)*, from GS1 non-profit organisation. A list of global standards of barcodes. - *[EU rules regarding food hygiene](https://ec.europa.eu/food/safety/biosafety/food_hygiene_en)*, from the European Commission. - *[GS1 searchable database of barcodes](http://gepir.gs1.org/index.php/search-by-gtin)*, from GS1 non-profit organisation. ## Glossary ### term-certified-product **Certified product** – a product that receives approvals and certifications confirming it abides by a set of quality and performance standards and regulations. For instance, certified organic foods are expected to be produced without the use of chemicals but also abide by specific conditions in terms of storage, packaging and transport, among others. ### term-chain-of-custody **Chain of custody** – a process that seeks to demonstrate that physical and other sorts of evidence is protected from tampering during the course of an investigation, from the point of collection to the point of publication or use in other circumstances, such as in court. ### term-commodities **Commodities** – traded goods or raw material. ### term-distributor **Distributor** – company responsible for getting the products into different countries, by operating or hiring shipping or trucking companies to carry the products. ### term-exporter **Exporter** – an actor (company, organisation, person) sending goods and materials across the border to a foreign country for trade purposes. ### term-further-processor **Further processor** – company that carries out additional transformations of a product (such as wood turning into fiber for textiles). In many supply chains there are multiple further processors, while other supply chains may not have any. ### term-hs-codes **HS (harmonised system) codes** – a nomenclature established to organise and list products for easier classification and labelling across borders. It is developed and maintained by the World Customs Organisation (WCO). ### term-importer **Importer** – an actor (company, organisation, person) bringing in goods and materials across the border from a foreign country for trade purposes. ### term-initial-processor **Initial processor** – company that carries out the first transformation of the product (for instance, a timber mill turning a log into planks). ### term-manufacturer **Manufacturer** – company that carries out the last transformation before the product is sold to consumers or industrial users (such as the company making furniture or toothpicks). ### term-producer **Producer** – the actor, be it a company or person/group of persons, that takes, grows, mines or otherwise produces the raw materials (such as the owner of a timber plantation). ### term-product-certification **Product certification** – a process of verification and approvals confirming that certain standards and regulations are met by products and services. ### term-retailer **Retailer** – company or individual responsible for selling the products to consumers or industrial users (like a hardware shop or furniture store). ### term-supply-chain **Supply chain** – a set of steps that commodities undergo on their way to becoming products used by consumers or industry. ### term-supply-chain-bottleneck **Supply chain bottleneck** – an impediment or blockage along the supply chain caused by shortage of a commodity or increased demand, and where only a small number of companies may have overwhelming market shares. ### term-vpn **Virtual Private Network (VPN)** - software that creates an encrypted "tunnel" from your device to a server run by your VPN service provider. Websites and other online services will receive your requests from - and return their responses to - the IP address of that server rather than your actual IP address.
# Supply Chain and Product Investigations
By Mario Rautner  ```cik-in-short ``` **In Short:** An introduction to supply chain investigations including an overview of the main tools, techniques, data resources and essential precautions to take. It focuses on the main actors, stages and processes of a supply chain and includes a hypothetical step-by-step investigation. --- A supply chain is a set of steps that commodities (goods or raw materials) undergo on their way to becoming products used by consumers or industry. Stages of a supply chain can include places where products are transformed – for instance, an electronics company where circuit boards are made from transistors and copper lines – and where products change hands through transport, such as the loading of circuit boards onto a truck and their shipment to a computer manufacturing company. Supply chain investigations focus on collecting evidence to link each stage of a product’s journey, from the origin of the commodity (for example, the _**[conflict metal Cassiterite](https://en.wikipedia.org/wiki/Conflict_resource)**_ used in the solder of laptops) to the final product (the laptop itself), which can contain hundreds of different components. Supply chain investigations can be particularly impactful because they have the potential to change the behaviours of entire industries when done effectively, with thorough and reliable evidence. A _**[recent supply chain investigation](https://www.cips.org/en/supply-management/news/2018/may/unilever-suspends-palm-oil-suppliers-over-deforestation-claims/)**_, for example, alleged that a large palm oil supplier to major food companies such as Unilever, Nestle and Mars, was conducting massive deforestation and not complying with sustainable palm oil policies. As a direct result, Unilever stalled collaboration with the supplier pending further verification. In another case, in 2018, the British supermarket chain Iceland (Iceland Foods Ltd) _**[committed to removing palm oil](https://www.theguardian.com/environment/2018/apr/10/iceland-to-be-first-uk-supermarket-to-cut-palm-oil-from-own-brand-products)**_ from the manufacturing process of its own-brand products by the end of 2018. Such investigations and findings can have an impact not only on the reputation but also the financial status of the companies involved, in particular when they reveal illegal or controversial activities. Remember, something doesn’t have to be illegal for it to be considered controversial. Even the use of rabbit fur, while legal, can be problematic for a fashion brand these days. In the most successful cases, this change impacts the supply chain all the way back to the origin of the product, where it may prevent social injustice, from poor working conditions and human rights abuses to land grabbing and environmental destruction. Most supply chain investigations are carried out by journalists or NGOs, but they can also be carried out by determined independent investigators or people and communities directly affected by an issue. ## Shining a light on where our products really come from Everything around us is part of a supply chain: from the food we eat to the clothes we wear and the devices we use. But the nature of supply chains means we rarely know how commodities are produced or even where they originate. We might not always consider where our items come from, who harvested their source materials and under what conditions, whether there was exploitation or abuse at any point in the production process, or if illicit acts were carried out along the way. However, these questions might be brought to light when public controversies emerge about certain products or brands. Perhaps you live in an area where important materials are extracted or produced. These materials may then enter the supply chain of a final product through countless contracts, companies and manufacturing processes, which are often opaque or not immediately visible or accessible. Or, you might live near a forest that is being cleared and want to know whether the companies carrying out the clearing are using legal and ethical practices. The impact of such findings can be crucial in saving the forest. Of course, it is not always an investigator living in that area who will expose the truth, but having someone familiar or close by to observe or launch the quest for more information is a good start. Often a few initial questions and some raw evidence can lead to wider investigations by another group or organsation. Equipped with awareness of the risks and safety considerations, and a good assessment of what is at stake, you can also start this way. Even if you are not physically present at a location where commodities are produced or extracted, many supply chain investigations can be carried out remotely. Indeed, in most cases, elements of a supply chain are located far from the source - for example, wood processing and furniture manufacturing can take place continents away from where the forest was originally cleared. ```cik-tip ``` >**Tip: Helpful tools anyone can use to track products** > >**Barcodes:** Barcodes on consumer products can be used to identify the manufacturer, as well as additional >information about them. Though barcodes are most familiar to us from a retail environment, they are also >used in many other areas of business. In retail, the most commonly used barcodes are the _**[European >Article Numbers / EAN](https://www.gs1.org/standards/barcodes/ean-upc)**_, which are usually 13 digits and >used both >within and outside of Europe; and the 12-digit Universal Product Code (UPC), which is very similar to the >EAN, but contains a zero at the beginning to signify registration in the USA or Canada. The first three >digits of an EAN indicates the name of the company that accepted the registration of the code (note that >the location of the company may be, but not necessarily, the same as the location the product was >manufactured). Companies that accept registration are members of the GS1 non-profit organisation, which >maintains the global standard of barcodes. The [GS1 website](https://www.gs1.org/) has a _**[searchable database](http://gepir.gs1.org/index.php/search-by-gtin)**_ of many (but not all) barcodes. In some cases >this is useful for obtaining information on who registered the product, and can also reveal details about >a company's ownership structure. > >**Food hygiene labels:** >EU rules regarding _**[food hygiene](https://ec.europa.eu/food/safety/biosafety/food_hygiene_en)**_ cover >all stages of the production, processing, >distribution and sale of animal products intended for human consumption. This includes, for example, fish, >meat and dairy products, but also non-food animal byproducts such as pet food. Every individual >establishment or outlet (rather than the parent company) that puts such products on the market has to have >a registration number that uniquely identifies it. The codes can be found on the product label and consist >of a combination of letters and numbers within an oval. Have a look at a product at home to see how easy it >is to spot the label. The EU has an online database of the codes, making it possible to identify the name >of the company that put the product on the market. Since the regulation covers imports and any processes >that take place in the EU, the database includes details of companies from over 80 countries. This comes in >very useful if you’re investigating the chain of custody of a food product manufactured by a large consumer >company on a contract, or a brand owned by a supermarket. In such cases, you can identify the actual >producer rather than the brand owner. ## What makes a supply chain investigation  As you prepare for your investigation it’s important to have an informed idea of what is possible to achieve, what is relevant and what may be out of reach. Supply chains are not always simple; they are often complicated networks or webs of actors, processes and movements. While it takes thousands of links in multiple supply chains to manufacture one laptop, investigations usually focus on very specific components of a supply chain, such as the source of one single element used in the manufacture of one part of the laptop, or a single factory where one stage of the production takes place. Maintaining such a focus makes the research more effective. Supply chain research can be instrumental in linking a company to ethical issues that taint their products, even though the physical components of the product itself may not be controversial – for example, the violation of indigenous land rights that often occurs when producing rubber for car tires. A key concept to remember about supply chains is **traceability:** the passing of information from one stage in the supply chain to the next. This can relate to the initial source of the ingredients or raw material of a product, the alterations it incurred along the way, or other relevant details that may help trace an entire history of transportation and production. Some companies have systems for this but many supply chains are so long and complicated that it can be difficult even for the companies directly involved to achieve a fully traceable supply chain. Sometimes, **certification schemes** are used to ensure that products are manufactured, produced and traded following specific standards. For instance, they can indicate that the human rights and traditional land rights of local and indigenous people are observed or that no tropical forests are cleared to make a specific product. Well known certification schemes include those for organic and fair trade produce or for sustainable timber, palm oil or coffee.  *Fairtrade certificate logos. Image from Pinterest: https://www.pinterest.com/pin/110619734567839836/* In order to carry out supply chain research, it is necessary to get a good understanding not only of the product you are trying to track and its production prices, but also about **company finances** and **trading** and **transport processes** in different countries. **Re-evaluate** your assumptions and **verify** your data at every step. As with any investigation, you should always question your initial assumptions and results in order to make sure that the information you collect is relevant to what you intended to prove. What at first appears likely may in fact have another explanation. For instance, you might have linked a company illegally clearing forests to a timber mill that sells to a furniture company, only to find out that the furniture company does not actually use the timber species harvested from the forest in question. You therefore cannot make a direct connection between the company and the illegal deforestation activity. The **flexible** and **unpredictable** nature of supply chain research also means that it is not always possible to anticipate how long your research will take. Sometimes linking two processes can take a day of online database work; sometimes it can take months and dozens of research strategies and tools, including online and field research. Due to the complexities of supply chain research and the legal, ethical and safety risks involved in accusing large corporations of wrongdoing, if you decide to conduct this kind of investigation, it’s best to take your time to slowly expand your skills and understanding of supply chains. Starting with basic searches and testing some of the tools and techniques we present here would be an easy way to introduce yourself to the field. How and where do I start? ------------------------- You can start your investigation at any point along the supply chain where you suspect unethical or illegal behaviour might be taking place. In most cases, investigators follow the trail downstream along the supply chain from this entry point. For instance, if an investigation starts at a mobile phone assembly plant suspected of using child labour, it will most likely continue downstream to the brand that sells the mobile phones, rather than upstream to the origin of the plastics and metals used in the phones’ manufacturing process. The focus here will be to expose illegal and unethical practices of exploiting children to produce that phone. In some cases, however, research can occur upstream. An example might be the case of an ill-famed company establishing a manufacturing plant in a town whose residents want to find out if the products and commodities entering the plant are of controversial origin. Once you identify an entry point, you can follow a sequential path along the supply chain to collect evidence connecting each step, be it along a production process or a product’s transport path. That said, your research will most likely be nonlinear, and you will have to use multiple tools. Depending on what you are investigating, your work will involve extensive online research, and sometimes field research. You might need to use anything from maps to identify the source locations of raw materials or interactive transport tracking services to follow commodity shipments, to online customs and product databases or corporate records and stock exchange websites to learn more about the products and companies you are focusing on. You might also sometimes undertake field research and collaborate with others to collect evidence and witness accounts about what happens on the ground. Remember that companies tend to tightly guard information about their suppliers and customers to avoid exposing their internal mechanisms and advantages to their competitors. Companies also sometimes hire investigators to research the supply chains of competing products and of other companies. Good supply chain investigators look for evidence that links the chain, and this often requires a lot of creativity since no two supply chains are the same and access to information can vary vastly depending on the data sources available or the geographies where research is being conducted. The ability to ‘look sideways’ by making previously unseen connections is a critical skill, as is being able to come up with innovative solutions to research problems. ## Mapping and profiling actors When thinking about supply chains, a useful first step is to understand the various actors – the participating companies or individuals – that operate along the way. While the actual chains are usually unique combinations, some of the actors can be categorised into groups, which are often connected by shipping or transport providers. Most often, along a supply chain you will encounter the following: ```cik-definition ``` > **Producer:** company, person or group of persons > that takes, grows, mines or otherwise produces the raw materials (such > as the owner of a timber plantation). > > **Initial processor:** company that carries out the first > transformation of the product (for instance, a timber mill turning a log > into planks). > > **Further processors:** companies that carry out additional > transformations of the product (wood can ultimately even turn into fiber > for textiles). In many supply chains there are multiple further > processors, while other supply chains may not have any. > > **Importers, exporters, distributors:** companies responsible > for getting the products into different countries, by operating or > hiring shipping or trucking services to carry the products (for > example, shipping the wood planks to a deposit in country X, from where > it can be sold to furniture manufacturers and others). > > **Manufacturers:** companies that carry out the last > transformation before the product is sold to consumers or industrial > users (such as the company making furniture or toothpicks). > > **Retailers:** companies and individuals responsible for > selling the products to consumers or industrial users (like a hardware > shop or furniture store).  *Simplified example of supply chain actors and their connections* Let’s assume you’ve chosen your entry point in the supply chain: for example, a local fabrics manufacturer where you suspect that workers are mistreated and underpaid, and you want to find out what fashion brands buy these fabrics and where they are sold. You’ve also mapped the actors you will focus on: the manufacturer and the retailer (plus, potentially, the final consumers), so you can raise public awareness about your investigation findings. Now you know what and whom you are investigating. The next step is to establish a list of companies and people of interest – your actual actors. Before you move any further, begin by researching their company ownerships, business models, networks of collaborators, places where they are registered and/or operate, products they manufacture or sell, and possible controversies surrounding them already. By doing this background research you create a profile of your main actors and their connections, build your foundation for further investigation, prepare yourself for interviews with relevant people (sources, witnesses, specialists etc.), establish connections and assess your potential risks. Internet research is a good place to start looking for basic information about businesses and people associated with the companies in the supply chain you're investigating. Try to find out as much as possible from companies’ websites and activity reports, supply chain due diligence reports (if available), stock exchanges (if the company is listed), media articles or social media profiles of the company, its board and staff. At the same time, look for official documents that include the companies’ registration data, shareholders, subsidiaries, members of the board, directors, annual financial reports and other relevant details. If you already know in which countries to look for official documents, you can start by checking available online corporate registries and other official databases, such as land records, court records, patent registries etc. There are a number of platforms to help you start this research. Some corporate registry websites may seem intimidating at first, but don’t be daunted: there are more user-friendly, convenient and extremely helpful ways to get introduced to the company research field. _**[The Investigative Dashboard](https://id.occrp.org/)**_ (ID) is a resource built by the [Organized Crime and Corruption Reporting Project](https://www.occrp.org/en), providing a global index of public registries for companies, land and courts. From this platform you can go to the country and type of records you want, but note that while some corporate records provide free access to basic or more advanced information, many will require registration and fee payments in order to provide detailed company records. ID often marks the resources as free or paid for, so you know what to expect. In addition, the platform allows investigators to freely search and use its _**[database](https://data.occrp.org/)**_ of millions of documents and datasets from public sources, leaks and investigations it has conducted in the past. Chances are that you might find good leads in there too.  *Investigative Dashboard database: https://id.occrp.org/. Screenshot by Tactical Tech.* There are other similar resources where you can continue and expand your research. _**[Open Corporates](https://opencorporates.com/)**_ is a platform that lets you search freely for company records and related data it collects from all over the world. [The International Consortium for Investigative Journalism (ICIJ)](https://www.icij.org/) runs the _**[Offshore Leaks Database](https://offshoreleaks.icij.org/)**_ a massive resource of company records and other useful documents that have been revealed by large leaks and projects the group has been working on, including the Paradise Papers, Panama Papers, Offshore Leaks and Bahamas Leaks. Do not forget about the power of social media, and especially professional networks such as LinkedIn, where you can search for people and companies as well as identify possible connections among your subjects of interest. Elsewhere in this kit, you can learn how to build a strong body of evidence, how to investigate companies and people, how to do online and offline research and how to use various tools and techniques such as maps, social media, interviews etc. You can combine resources and develop new skills with each issue you investigate, based on the demands of your particular context, while also making sure to remember the safety and ethical considerations that go hand in hand with your plans. ## Inspiration and guidance: a sample investigation This section uses a fictitious example to show you the kinds of tools and approaches you might use to carry out a supply chain investigation. While the example is relatively realistic, the [chain of custody](#term-chain-of-custody) has been simplified and shortened for the sake of illustration. Unlike the product itself, chain of custody research cannot follow a simple research recipe, so the goal of this example is to demonstrate the need to think creatively and adjust to each specific case. ```cik-note ``` >**Note:** all the names of the entities involved are fictional, and >any relation to real companies or products is coincidental. ### Sparked by a story of injustice You read a news report that details how villagers in a South American country have been rounded up at gunpoint and forced to work in a bauxite mine run by a company named Brown Gold. You want to find out if the minerals from the mine are being used in well known consumer products. You start your investigation by focusing on the supply chain at the point where the mineral leaves the mine or company warehouse. This is your entry point. Initially, you might begin with simple online research. Keep in mind that this case could be sensitive, so it’s useful to familiarise yourself with and start using some digital security tools, such as the [Tor Browser](https://www.torproject.org/projects/torbrowser.html.en), [DuckDuckGo](https://duckduckgo.com/) search engine or a [VPN](#term-vpn) (Virtual Private Network), to avoid being tracked. Your search does not yield much more than the name of the company itself, but in the process you gather additional information, such as the company’s registration number and the names of the founders and key people in the company. You also carry out online research and ask some specialists you trust on related subjects such as: - The uses of bauxite (as a main source of aluminium). - Detailed information about the aluminium production process and its impacts. - The largest companies involved in global bauxite and aluminium production. - The main uses for aluminium in industry. ```cik-example ``` >**Example of online research entry points:** > >In many instances, [Wikipedia](https://www.wikipedia.org/) provides useful entry-level information and links to external sources that you can check for more in-depth knowledge. Specialised websites that focus on the particular minerals are helpful for understanding the field and its main players. >In this case, for example, a DuckDuckGo (or a Firefox, Google etc.) search for “aluminium and bauxite” leads you to pages like that of [Geology.com](https://geology.com/minerals/bauxite.shtml), >[Aluminum.com](https://www.aluminum.org/industries/production/bauxite) (Aluminum Association in the US), [Australia’s Aluminum Council](https://aluminium.org.au/interactive-flowchart/bauxite-mining-chart/) and many others, where you read about the production process and uses of this resource. > >Searching for “bauxite producers world 2019,” for instance, takes you to information on the current _**[top bauxite producer countries and companies](http://www.thedailyrecords.com/2018-2019-2020-2021/world-famous-top-10-list/world/largest-bauxite-producing-countries-world/6882/)**_. > >Searching for “aluminium producers world 2019” leads you to other resources such as a [map and downloadable database](http://www.world-aluminium.org/statistics/) of top aluminium producers from international institute [World Aluminium](http://www.world-aluminium.org/) and additional data from statistics aggregator [Statista.](https://www.statista.com/statistics/280920/largest-aluminum-companies-worldwide/) ### Investigating the key players Your preliminary online research provided key information about the products and actors at the centre of your investigation; enough to build some background knowledge and basic details about your topic and main subjects (producers, buyers etc.). You then continue by mapping your actors and the connections between them – who is the producer, who delivers the commodity to whom, by what means etc. – and following the supply chain downstream from your entry point. #### The Producer: MINING COMPANY Since the company – Brown Gold – is mining the resource, it is considered the producer of the commodity. By using the resources mentioned above, you can gather various types of information to get a better understanding of who Brown Gold is: - Physical company operations and locations – source: maps and satellite images (potential transport routes). - Type of company (private, public and, if so, shareholders, etc.) – source: company registries, stock exchange, website information. - Financial information – source: company filings, annual reports on company’s website, stock exchange (if the company is listed on any stock exchange). - Registered office addresses (used for formal purposes, e.g. where mail can be received) and trading or operational addresses (where actual activities occur) – source: online and offline corporate records, company’s website. - Board members and directors (and any other businesses they might be involved in) – source: company registries, company website, company and people databases (see above for Investigative Dashboard and other resources) - Bank connections, loans, mortgages – source: corporate records, court records, annual financial reports. - Employee feedback, reviews and grievances on sites like [glassdoor.com](https://www.glassdoor.com/Reviews/index.htm), [comparably.com](https://www.comparably.com/), [careerbliss.com](https://www.careerbliss.com/), sometimes yield information about the company. - The current careers or job opportunity page and archives of job postings from _**[archive.org](https://archive.org/web/)**_ may yield information about projects, locations, and future plans of the company. ```cik-note ``` >**Note:** > >Carefully searching through company reports and other documents can lead to very useful findings. >Annual reports, corporate sustainability reports, company presentations and shareholder filings can all be useful not only for basic information on a company and its finances, but also, in some cases, about its customers (photos within these reports can be especially useful for this). > In this case, you find the location of the company on satellite images and identify its board members and directors, but cannot find any further information about them. Nor have you found detailed financial data or names of any customers of the company. Searching through company records and annual reports, you do, however, manage to find a subsidiary (a company owned and controlled by another company) of Brown Gold, by the name of BG Shipping. #### The Initial Processor: ALUMINIUM SMELTER From your initial online research you found out that bauxite is used to manufacture aluminium, which in turn is produced in large smelters (facilities where the metal is extracted by heating and melting the ore). You also discovered that smelters are massive investments – to the extent that there are fewer than 150 of them globally. This information will help you in your investigation, because it offers a relatively narrow field of inquiry. In order to find one of the smelters that is supplied by Brown Gold, you use a process of elimination to rule out certain routes, while looking for likely connections between the mining company and certain smelters. Examples of investigation activities to do this include: - Finding out if there are any smelters located in the country the mine operates in – there are none. - Identifying other bauxite mines in the country – you find three more. - Analysing whether this country exports any bauxite overseas using the UN Trade database [Comtrade](https://comtrade.un.org/) -- you find that six countries received bauxite. So far you know of four mines in total; this means you can’t yet determine which country is supplied by Brown Gold.  *Sample comtrade.org search. Screenshot by Tactical Tech.* In order to identify the smelter, you need to first determine the country that receives the bauxite. With a relatively simple online search (e.g. "how is bauxite transported") you learn that bauxite is most commonly transported by ships or trains. You carry out further online research, and check transport routes and map locations (using Google Maps or Google Earth) to map out deepwater seaports. Then, by looking at the train lines between the four mines and the location of the mines to each other, you conclude that Brown Gold is most likely exporting its bauxite using ships. Moreover, you previously identified the BG Shipping subsidiary and were able to locate its warehouses on the map at a port that is connected to Brown Gold via train track. You can therefore conclude with even higher certainty that the company exports its bauxite from this port. #### The Trader: SHIPPING COMPANY Today, the majority of commodity-carrying ships are fitted with live tracking devices, which employ what’s called an Automatic Identification System (AIS). All vessels over 300 tonnes (gross) are required by law to emit radio signals with their speed, direction and exact location. This data is also captured by receivers on other vessels and by stationary receivers in ports and on coastlines. Services such as _**[Marine Traffic](https://www.marinetraffic.com/)**_, _**[Vesseltracker](https://www.vesseltracker.com/)**_, and others use AIS information to produce live maps showing the movement of ships across the globe. To search the map, you need to know the name or registration number of the vessel you are tracking. For additional resources and depending on your needs, other tracking services may be of further help, so reviews like this from [Marine Insight](https://www.marineinsight.com/) - _**["Top Eight Websites to Track Your Ship"](https://www.marineinsight.com/know-more/top-8-websites-to-track-your-ship/)**_ - may help when looking for alternatives. Most of the ship tracking services either have a free version or offer a free trial, which can be used to track these vessels. They also tend to employ their own network of receiver stations and have maps on their websites that indicate the network coverage. MarineTraffic, for instance, provides near real-time information about vessels’ positions and voyage-related information. In addition to the AIS receivers, which are ground-based, Marine Traffic (and other such platforms) also collects vessel data from satellite receivers. Obtaining this data often requires paying a one-off fee or signing up for yearly subscription, which is usually cheaper if you plan to use it longer.  *Sample MarineTraffic.org search. Screenshot by Tactical Tech.* ```cik-Tip ``` >**Tip:** > >Always search more than one vessel-tracking website to trace the same ship, as some platforms may provide >you with more free information than others, and some may have the data you need while others may not. If there is an area of particular importance to your investigation that is not covered by these services, it’s even possible to buy your own receiver online, from places like Amazon or specialised websites such as _**[Milltech Marine](https://www.milltechmarine.com/)**_. If this is the case, do some serious research first: check what range you need to cover, what other users recommend, verify equipment reviews, consult forums like _**[ShipSpotting](http://www.shipspotting.com/)**_ and maybe even check with some vessel tracking enthusiasts for advice and recommendations. ```cik-note ``` >**Note:** > >Taking this route requires some financial cost, a stable Internet connection and power supply in the place where you are installing the receiver, knowledge of how to use the equipment and some extra effort to maintain it. However, if it helps to advance your investigation and you have no other options, it’s well worth the relatively short time it takes to learn and experiment. Marine Traffic provides useful advice for such situations in its [FAQ section](https://help.marinetraffic.com/hc/en-us/articles/205326387-What-is-the-cost-of-an-AIS-Receiving-Station-). >The website also offers free receivers to those who want to contribute to its [live map](https://www.marinetraffic.com/en/p/expand-coverage), but that requires you to share your receiver’s data >with the platform, something you might or might not want to do, depending on how sensitive your >investigation is. If you decide to go ahead and operate a receiver, you also get a free subscription plan, or an upgrade to your existing account. After deciding which tracking site(s) to use, you spend several weeks mapping out the journeys of vessels that appear to have been loading materials at BG Shipping. Ships that spend several hours berthed right next to the BG Shipping facility in the port are more than likely loading or unloading. You can track journeys of vessels by saving their movements on some of the above mentioned online platforms. In Marine Traffic, for example, you can create your own "fleet" of vessels to monitor and get updates on, or set up notifications when a vessel reaches a port. You can also access past travel information in the database - a service that is sometimes free, other times for a fee. Eventually, by tracking these ships, you identify three ports in three different countries across Asia that received exports from BG Shipping. This means that, at the very least, three separate smelters bought bauxite from the Brown Gold mine. Using Marine Traffic you manage to show that a vessel that left BG Shipping went directly to the port facility of a company called Caliper Alumina. You conclude that Caliper Alumina is one of the smelters receiving bauxite from Brown Gold mine. Think back to your earlier research, when you identified and mapped smelters, their transport routes, subsidiaries, key personnel, addresses and finances. You can now apply the same techniques and resources to find out more about Caliper Alumina and its activities. ```cik-tip ``` >**Tip:** > >Vessels sometimes make multiple stopovers. It’s also not always clear from sites like Marine >Traffic whether a vessel is loading or unloading. Look for additional information about the vessel’s draught (the depth its hull is submerged into the water), which provides another indicator of whether it has been loading or unloading (the higher the draught, the heavier the ship). > #### The Further Processor: CAN PRODUCER In this example, you can use another resource, namely *customs data*, to try to identify one of the customers of Caliper Alumina. Customs data consists of shipping manifests, which every shipment vessel is legally required to carry while it travels around the world. Governments collect this data in order to produce their trade data statistics. It’s possible to purchase this data, but doing so is expensive and only available for a limited number of countries. Generally it provides the name of the importer and exporter – so if one country in the trade does not make customs data available for purchase, you may be able to get the same information from the other country involved in the transaction. The cost of purchasing customs data has decreased lately, in line with a growing number of companies offering it for sale. Examples of such services include [Trade Atlas](https://www.tradeatlas.com/en), [Export Genius](https://www.exportgenius.in/export-import-trade-data/), [Xportmine](https://www.xportmine.com/), [ListThe](https://www.listthe.com/), [Panjiva](https://panjiva.com/) etc. The first step in this process is knowing which data to buy, and to do that, you need some context and background information. Products are usually organised by HS (harmonised system) codes, a nomenclature developed and maintained by the _**[World Customs Organization](http://www.wcoomd.org/)**_(WCO). The codes are organised into 21 sections, which are subdivided into 97 chapters, in turn further subdivided into several thousand headings and subheadings. The most recent HS codes edition, which we use in this section as of 2019, is [the 2017 HS Nomenclature](http://www.wcoomd.org/en/topics/nomenclature/instrument-and-tools/hs-nomenclature-2017-edition/hs-nomenclature-2017-edition.aspx) (note that a new one will come into place in [2022](http://www.wcoomd.org/en/topics/nomenclature/instrument-and-tools/hs-nomenclature-2022-edition.aspx)). **Here is how it works:** Usually each subdivision comes in a two-digit increment. The case in our example: Section XV "Base Metals and Articles of Base Metal" lists a series of chapters numbered from 72 to 83. - Chapter 76 is Aliminium and articles thereof. This is followed by 16 headings, for example: - Heading 7606: Aluminium plates, sheets and strips... - Heading 7612: Aluminium casks, drums, cans, boxes and similar containers (including rigid or collapsible tubular containers), for any material (other than compressed or liquefied gas), of a capacity not exceeding 300l, whether or not lined or heat-insulated, but not fitted with mechanical or thermal equipment. - Subheading 7612.10: Collapsible Tubular containers.  *Base metals codes from World Customs Organization wcoomd.org. Screenshot by Tactical Tech.* When buying customs data, most of the time it is necessary to know the HS codes of the products in question. Often HS codes are defined down to eight or 10 digits, but at these levels the classification is usually set by the countries rather than the WCO. In addition, codes are reviewed and sometimes adjusted every few years, including those of the World Customs Organization above, so make sure you stay updated. ```cik-note ``` >**Note:** > >Each product and shipment that enters a country has an HS code attached >to it, so both domestic and international trade statistics are usually >based on those HS codes. For instance, the United Nations maintains the >_**[Comtrade database](https://comtrade.un.org/)**_, which contains such statistics - >[comtrade.un.org/data](https://comtrade.un.org/data) - that can also >prove useful for chain of custody investigations. In many cases, customs data comes in spreadsheets and often contains false positives or multiple spellings for the same company name. Depending on the purpose of your investigation, you might need to clean the data, especially for statistical analysis. The details of the data vary depending on the country that supplies it, but it usually includes: - the names of the exporter and importer, - the date of the shipment (at least the month), - the weight of the shipment, - the HS codes, - the port of loading, - the port of discharge. Additional data fields can include addresses, phone numbers, detailed product descriptions, the value of the shipments, the name of the vessel and shipping company, as well as the names and addresses of notifying parties (the buyer or importer who receives notice of the shipment’s arrival). In this case, the data shows that a US company by the name of Beverages Inc. purchased aluminium sheets from Caliper Alumina. You also find a number of older shipments made by that company, which you are able to verify with historical Marine Traffic data, by tracking the vessel’s past movement as detailed above. You also notice that the Beverages Inc. website features photos that show cans of soft drinks produced by prominent brands. #### The Manufacturer: SOFT DRINK COMPANY One of the companies whose cans are displayed on the website of Beverages Inc. is called Spark Drinks. While you might be tempted to link this well known brand to the controversial bauxite producer, more research is necessary to ensure that this supply chain is indeed linked. There are various reasons why they may not be connected. There could, for instance, be different grades (strength, concentration) of aluminium in different lines of cans used for different brands. So, there is a chance that Sparks Drinks may not be using the aluminium made from the bauxite sourced from Brown Gold’s mine. This is where the problem of *traceability* emerges. ```cik-note ``` >**Note:** > >Supply chain *traceability* is a process and technique that ensures the >integrity and transparency of a company’s supply chain, by tracking >products from origin to consumer. This reduces the risk of mislabelling >commodities that have been mixed from different suppliers and regions, >allows products to be tracked back to their source and helps identify >particular batches of the product if something goes wrong and products >need to be tested or recalled, etc. Traceability is a key challenge in >supply chains. > >In certain cases, however, commodities and products are *segregated* >(they have their *identity preserved*), which means that companies can >trace the product and its components all the way back to the origin and >the producer. This segregation is often required for certified products >(e.g. those who must meet strict quality standards) and animal products >(for instance, to identify farms in case food safety issues arise). > >In most instances, though, raw materials or components from different >sources (e.g. from different mines and suppliers) are mixed together at >various points in the supply chain. The result is that, due to a lack of >segregation and traceability, most companies do not know for sure >whether they are connected to actors involved in illegal, unethical or >controversial activity. From an investigator’s point of view, this means >that all actors downstream of where the mixing occurs are part of the >supply chain in question. > >For companies, traceability is important for a number of reasons. These >include compliance with policies of sustainable sourcing (for example, >making sure they do not contribute to famine, drought or environmental >damage) or with conflict minerals regulations implemented by various countries, which impose strict rules >on the purchase and use of commodities obtained from areas controlled by >armed groups or where abuses take place. Such aspects are important for a company’s accountability >to regulatory authorities and to their consumers. There are a number of research techniques you can use to find out if mixing occurs along the supply chain. Sometimes the *company information* (websites, public filings) includes details about where the company purchases commodities from and if it keeps a detailed log of its own supply chain from the sourcing of materials until the point these reach its ‘hands’. In other situations, searching online for *scientific articles* or other investigations on your topic or commodity of interest, as well as more general magazines such as _**[Supply Chain Quarterly](https://www.supplychainquarterly.com/)**_, and available *industry* *research* might shed light on how the sector operates in general. In addition, NGOs like [Greenpeace](https://www.greenpeace.org/international/), [Rainforest Action Network](https://www.ran.org/), [Rainforest Alliance](https://www.rainforest-alliance.org/), [Mining Watch](https://www.miningwatch.ca/) and many others already conduct research on various supply chains and their impact on human rights, the environment, and the wellbeing of communities living and working along the chain. Contact them for advice if you find their knowledge useful. ```cik-tip ``` >**Tip: using NGO reports** > >The techniques and resources outlined here are often used by >NGOs working to pressure companies into changing their ways of >production or cleaning up their supply chains. NGO reports can offer >insightful and inspiring examples of how such tools have been used in >conjunction with other supply chain research techniques. When reading reports, keep in mind the >mission of the NGO. Some of the language >may not be appropriate for investigators or journalists and some texts are >designed for advocacy and policy purposes. However, it’s an interesting >exercise to try and figure out how the research was carried out and what >investigation tools and strategies were applied. Also look out for *trade and industry magazines* and websites that profile the way your company of interest operates (see [supplychain247.com](https://www.supplychain247.com/), for example). However, keep an eye out for articles and research that appear one-sided or solely focused on positive angles – this is a strong clue that they may be sponsored by the very companies you are investigating, and thus not always reliable or complete. Search multiple sources and read as much as possible to create a complete picture. You might also consider *talking to industry experts* or even to the company itself. ```cik-warning ``` >**Warning:** > >Before talking directly to the company of interest, you should conduct >thorough background research on their operations, and be prepared to ask >highly relevant questions rather than details you might already find >elsewhere. If you plan on accusing the company of something, it makes >sense to avoid telling them so, or even closely hinting at that, before >you’ve gathered solid proof. Powerful companies will go a long way to >make you stop your research, try to discredit you or sometimes even >threaten you with legal action or other types of >pressure. Now, back to our scenario. You find out from industry magazines that mixing occurs at the aluminium smelter and that all products leaving the smelter contain bauxite from various sources. Since this question is crucial, you also confirm it by carrying out interviews with industry experts and eventually also with representatives of the smelter itself. This means that any product containing aluminium that originates from the smelter, including the products of Spark Drinks, is potentially contaminated with bauxite of controversial origin. #### The Retailer: SUPERMARKET In this supply chain, the last step of identifying retailers that sell the products to consumers is relatively simple, since most retailers in your country have online stores and stocked products can be found via the retailer websites – as well, of course, as in person by visiting the stores. This allows you to complete the supply chain and link the bauxite to a number of major supermarket chains. ### Loose ends and further questions Even though the case study ends here, there are many potential questions still to be answered to increase your certainty in the information. These include: - Are the customers listed by Beverages Inc. still up to date? - Is it possible that the bauxite and aluminium are of grades that rule out the smelter you identified, and thus render the particular chain impossible (and therefore another smelter would have to be identified)? - Does Beverages Inc. export to other countries? - Are there ways to confirm the supply chain links using other research techniques or further verify the results you obtained so far? You can employ numerous different research techniques to identify customers of the aluminium smelter and, depending on how many more transformations take place, there may still be more chains to uncover before you reach the final consumer product. For electronic products, it is likely that the aluminium will change hands many times before it can be found (unrecognisable in shape and form) in a store. In some cases, due to the nature and complexity of supply chains, you’ll have to invent or significantly adapt the research techniques as you go. ## Risks and considerations Supply chain research often comes with significant risks, many of which depend on both the research strategies employed and the geographies in which they take place. Because this type of research is often connected to and can effect the profits, loss, and earnings of companies it makes sense to have an understanding of how to secure your research and work pseudonymously. If you decide go public with your findings timed release of key research (instead of releasing everything at one) is a method of dealing with crisis management and public relations teams. Developing a methodical and meta data preserving way to catalog data, notes, research and files is key. Developing a workflow that involves having an encrypted backup is recommended. If you are technical, look for ways to split decryption keys into _**[(Shamir Secret) Shards](https://en.wikipedia.org/wiki/Shamir%27s_Secret_Sharing)**_. This allows you to split the key into several pieces, and only if a certain number of pieces come together does the key reform. Look at projects like [ssss](http://point-at-infinity.org/ssss/) and [storaj](https://storj.io) for examples. Hashes of these shards can be used to create so-called "insurance files" which may be of use to the investigator, however is out of scope of this section. ### Safety concerns Successful supply chain research often requires field research in addition to online data collection and shipments tracking. Field research can be high-risk, especially in some areas where corruption is rife, where investigators and journalists have little protection, or where violence occurs frequently. This kind of work is usually carried out in teams and under detailed plans for physical and digital security. High-risk work should only be undertaken by experienced investigators. ### Data access Companies usually have multiple customers and suppliers, often hundreds. There are, therefore, many forks in the road when it comes to following the trail of the products you are investigating. Make such choices carefully and base them on knowledge, facts and information rather than on gut feelings. An assessment of the difficulty of research and operating in specific jurisdictions for instance is a rational way of making such decisions. It is important to give special consideration to supply chain bottlenecks (for example, a blockage in the chain caused by shortage of a commodity or increased demand), where only a small number of companies may have overwhelming market shares. For instance, a handful of traders now dominates the global trade of agricultural commodities. In most cases it is not possible to prove the supply chain for a specific product all the way to the end product. This diluting of the chain is common and is not necessarily a concern. For example, it would be very difficult to show that one harvest of coffee beans from a smallholder (a non-industrial farmer harvesting a small area) who has illegally cleared forested land, ends up in instant coffee products of a large consumer brand company. However, it might be possible to prove that the smallholder sells to a specific agent, who sells the coffee to a company where the beans are collected and washed. This company in turn sells to a processing and packaging company that ships the coffee overseas to an importer, who passes it on to the coffee roasting and grinding company, which in turn sells it to a large consumer brand that packages the instant coffee products, which are sold in supermarkets around the world. In this scenario, the coffee coming from the illegal plants can never be traced directly and has been mixed with coffee from many other plantations, yet the consumer brand is still exposed to the problem because it is part of the supply chain. ### Legal issues Supply chain research requires a very high standard of proof. If the evidence at any point of the chain is not strong enough, additional investigations are required to corroborate your theory or conclusion. This type of investigation often ultimately exposes very large corporations for whom reputation is as valuable as the products themselves, and they often respond strongly when their reputation is under threat. Be aware that it is not uncommon for lawsuits to ensue from investigations on supply chains, and there are significant risks for investigators even when the information they provide is accurate. Supply chain investigators frequently obtain legal advice before publishing the results of their work, which often exposes powerful companies and shady practices. Investigators may find themselves accused of libel or slander; countries have different laws and potentials for legal issues so asking for legal opinions is useful even if only as a precaution. Organisations and individuals engaged in this research can also face SLAPP (Strategic Lawsuit Against Public Participation) suits. These are designed to use the companies’ large funds to engage investigators in lawsuits in order to intimidate them and delay or stop them from carrying out their work, even if the lawsuit has very little or no legal merit and chance of success. ### Collaboration In order to avoid or mitigate many of the risks above, investigators often collaborate with others doing similar work, including asking for advice and best practices on physical and digital security. Collaboration is particularly advisable when investigating certain topics for the first time, and additional knowledge and expertise is required. Don’t shy away from asking questions to trustworthy organisations and people, and avoid the temptation of trying to take down a company with your very first investigation. Collaboration also help share the workload, so you don’t have to follow the entire supply chain by yourself, but instead work with local communities, experts, NGOs and others to collect stronger evidence and reach a more impactful result. Some countries may have freedom of information laws which allow citizens of that country access to information not available to the investigator. It may make sense to collaborate with a citizen who can access other information and at lower risks. When collaborating it makes sense to take extra care to encrypt files/folders and get hashes (fingerprints) of files to look for changes. The kind of "source control" software a developer uses ([github](https://github.com/), [bitbucket](https://bitbucket.org/product), etc.) may aid you in doing this and keeping track of files, edits, and ownership. When working in a team be sure to have a secure and encrypted way to communicate. Messenger apps like [Signal](https://signal.org/en/) or [Wire](https://wire.com/) are recommended. <hr class="thick"> _Published April 2019_ ## Resources ### Articles and Guides - *[Supply Studies Research Guide: A Research Guide for Investigations in the Critical Study of Logistics](https://supplystudies.com/research-guide/)*, by Matthew Hockenberry, Ingrid Burrington, Karina Garcia, and Colette Perold, 2025. - *[Destroying elephant habitat while breaching the Indonesian Government moratorium on forest clearance for palm oil](https://www.ran.org/wp-content/uploads/2018/06/RAN_Leuser_Watch_PT_Agra_Bumi_Niaga.pdf)*, from Rainforest Action Network. A supply chain investigation report. - *[Eating up the Amazon](https://www.greenpeace.org/usa/wp-content/uploads/legacy/Global/usa/report/2010/2/eating-up-the-amazon.pdf)*, from Greenpeace. An investigation report on tracking soy from the Amazon. - *[Investigating supply chains](https://gijn.org/investigating-supply-chains/)*, from the Global Investigative Journalist Network (GIJN). A short guide including tips, resources and techniques. - *[Investigating illegal timber](https://www.earthsight.org.uk/tic/guidebook)*, from the Timber Investigations Center. A guidebook to researching timber supply chains. - *[Top eight websites to track your ship accurately](https://www.marineinsight.com/know-more/top-8-websites-to-track-your-ship/)*, from Marine Insight. A brief review including tools comparisons and tips. - *[Who's Got the Power: Tackling imbalances in agricultural supply chain](https://web.archive.org/web/20151117052506/http://www.fairtrade-advocacy.org/power/183-projects/psc-main-page/870-the-report-on-imbalances-of-power-in-agricultural-supply-chains)*, from the Fair Trade Advocacy Office (FTAO). A study about power concentration and unfair ### Tools and Databases - *[EAN/UPC barcodes overview](https://www.gs1.org/standards/barcodes/ean-upc)*, from GS1 non-profit organisation. A list of global standards of barcodes. - *[EU rules regarding food hygiene](https://ec.europa.eu/food/safety/biosafety/food_hygiene_en)*, from the European Commission. - *[GS1 searchable database of barcodes](http://gepir.gs1.org/index.php/search-by-gtin)*, from GS1 non-profit organisation. ## Glossary ### term-certified-product **Certified product** – a product that receives approvals and certifications confirming it abides by a set of quality and performance standards and regulations. For instance, certified organic foods are expected to be produced without the use of chemicals but also abide by specific conditions in terms of storage, packaging and transport, among others. ### term-chain-of-custody **Chain of custody** – a process that seeks to demonstrate that physical and other sorts of evidence is protected from tampering during the course of an investigation, from the point of collection to the point of publication or use in other circumstances, such as in court. ### term-commodities **Commodities** – traded goods or raw material. ### term-distributor **Distributor** – company responsible for getting the products into different countries, by operating or hiring shipping or trucking companies to carry the products. ### term-exporter **Exporter** – an actor (company, organisation, person) sending goods and materials across the border to a foreign country for trade purposes. ### term-further-processor **Further processor** – company that carries out additional transformations of a product (such as wood turning into fiber for textiles). In many supply chains there are multiple further processors, while other supply chains may not have any. ### term-hs-codes **HS (harmonised system) codes** – a nomenclature established to organise and list products for easier classification and labelling across borders. It is developed and maintained by the World Customs Organisation (WCO). ### term-importer **Importer** – an actor (company, organisation, person) bringing in goods and materials across the border from a foreign country for trade purposes. ### term-initial-processor **Initial processor** – company that carries out the first transformation of the product (for instance, a timber mill turning a log into planks). ### term-manufacturer **Manufacturer** – company that carries out the last transformation before the product is sold to consumers or industrial users (such as the company making furniture or toothpicks). ### term-producer **Producer** – the actor, be it a company or person/group of persons, that takes, grows, mines or otherwise produces the raw materials (such as the owner of a timber plantation). ### term-product-certification **Product certification** – a process of verification and approvals confirming that certain standards and regulations are met by products and services. ### term-retailer **Retailer** – company or individual responsible for selling the products to consumers or industrial users (like a hardware shop or furniture store). ### term-supply-chain **Supply chain** – a set of steps that commodities undergo on their way to becoming products used by consumers or industry. ### term-supply-chain-bottleneck **Supply chain bottleneck** – an impediment or blockage along the supply chain caused by shortage of a commodity or increased demand, and where only a small number of companies may have overwhelming market shares. ### term-vpn **Virtual Private Network (VPN)** - software that creates an encrypted "tunnel" from your device to a server run by your VPN service provider. Websites and other online services will receive your requests from - and return their responses to - the IP address of that server rather than your actual IP address. |
|
None
Resource updated |
The “
source/en/the-kit-en.json” file was changed.
4 months ago
|
|
None
String updated in the repository |
Ocean Datasets for Investigations
=============================== By Mae Lubetkin and Kevin Rosa  ```cik-in-short ``` **In short:** Learn how to identify and use Ocean datasets as a tool for revealing the unseen or underreported dynamics of the world's most significant bodies of water. --- ## 1. Introduction: Ocean Science, Data, Storytelling Given our current planetary condition, many of the world's most pressing stories are linked to the ocean. Covering 71% of Earth's surface and interacting constantly with the atmosphere, the ocean is our greatest carbon sink and an essential indicator of climate change. Despite its critical role in maintaining a habitable planet and supporting coastal livelihoods, the ocean is often invisible to the lived experience of most individuals, particularly those far from its shores. There are so many ways to tell stories about the ocean, and countless diverse perspectives from which to understand it. Now, more than ever, we need to incorporate cross-cultural and trans-disciplinary strategies for investigative projects, particularly those concerning the ocean. This guide presents ocean datasets as a tool for revealing the unseen or underreported dynamics of the world's most significant bodies of water. For informed investigations and impactful storytelling, oceanographic datasets can be an essential resource for journalists, activists, and anyone interested in data-driven methods to communicate the climate crisis, environmental change, natural disasters, extractivism, and associated ocean justice issues. From bathymetric maps, subsea imagery, and 3-D habitat models, to satellite-derived and *in situ* real-time monitoring data – a vast amount of oceanographic media and data is publicly available. In this Exposing the Invisible Guide, we begin with an introduction on the broader scientific history and context within which ocean data is collected, stored, and made accessible. Section two outlines the diversity of datasets, including some history, trade-offs, use cases, data collection methods, data sources, and resources to learn more about each data type that we present. Section three offers a specific application of ocean data in a case study, including: steps explaining why the data is useful for supporting this particular story; how to source and present the data; and, finally, how to bring it into a meaningful investigative report, journalistic piece, or other storytelling format. The guide concludes with a summarized approach for using ocean datasets in investigations and outlines strategies for identifying the right ocean scientists who could support you and your investigation.  *Boulder resting on the seafloor offshore the Revillagigedo Archipelago in the Pacific Ocean. Credit: Ocean Exploration Trust.* ### 1.1 Ocean Science: History and Context Ocean science generally refers to the observation and investigation of biological, geological, physical, and chemical processes that shape and constitute global marine environments. This broad disciplinary field includes numerous sub-disciplines that focus on intricate, detailed studies of specific scientific questions concerning the ocean. Ocean scientists monitor habitat change, measure biodiversity, investigate geological phenomena, and study human impacts on ocean systems (i.e., global warming, pollution, overfishing, and extractive projects). Interactions between marine ecosystems and human activities, as well as atmospheric and coastal processes, are all carefully investigated by ocean scientists today. Despite niche specializations, there are increasingly multidisciplinary projects that involve diverse experts in order to more comprehensively understand the interconnections between oceanic processes and phenomena. Collectively, this research improves our baseline knowledge of the ocean, which can then support preservation strategies while maintaining sustainable and regenerative relationships with diverse marine ecosystems. Although ‘contemporary’ ocean science has deep roots in European colonialism and imperial exploration, ocean knowledge systems long predate Western scientific inquiry. Indigenous and coastal communities across Oceania, as well as the Atlantic and Indian Oceans, carefully studied and navigated the seas for thousands of years. These forms of ocean science are less known or dominant on a global scale, but they nevertheless involve highly sophisticated techniques for understanding the stars, ocean swells, winds, and currents. Navigators across Oceania used interconnected and embodied forms of ocean science, on their own terms, to journey vast distances across the seas with tremendous precision. While other coastal Indigenous peoples around the world developed specific place-based systems of knowledge, including both land and marine management practices that viewed these spaces as highly linked. Most of these communities understood the ocean not as a monstrous or alien-filled void (as European explorers often depicted it), but as a vast world interconnected with our own.  *Rebbilib (navigational chart) by a Marshall Islands artist in the 19th to early 20th century. These stick charts were used by Marshall Islander navigators during long ocean voyages. Credit: Gift of the Estate of Kay Sage Tanguy, 1963. Source: <https://www.metmuseum.org/art/collection/search/311297>* When European seafarers began worldwide exploration voyages in the 15th-16th centuries, they dismissed or actively erased these Indigenous ocean knowledge systems. European or 'Western' scientific models were linked to a colonial mindset that often viewed the natural world as a space to examine in order to master and own its elements, rendering them as 'resources'. Notions of relationality were strongly opposed to the point that scientists considered their surroundings as objects to be studied rather than elements to relate to or work with. At the core, these opposing ocean scientific methods or knowledge systems reflected the specific values and worldviews of each culture, respectively. The Challenger Expedition (1872–1876) was the first European-led systematic scientific exploration of the global oceans. In some ways, it was groundbreaking. However, it also played a key role in the broader colonial project, which aimed to map, control, and extract 'resources' from around the world. Today, Western or 'contemporary' ocean science continues to use investigatory methods that stem from European knowledge systems. Oceanographic research often occurs in the context of economic or territorial expansion and military-supported science projects. Nevertheless, these methods are beginning to open up to other forms of knowledge creation that move beyond the geopolitical interests of wealthy nations.  *Map of ocean currents created by John Nelson using the "WGS 1984 Spilhaus Ocean Map in Square" projected coordinate system in ArcGIS. The Spilhaus Projection (developed by oceanographer Athelstan Spilhaus in 1942) reflects aspects of decolonial cartography by shifting focus from land-centered perspectives to an oceanic worldview. Source <https://storymaps.arcgis.com/stories/756bcae18d304a1eac140f19f4d5cb3d>* Thanks to the enduring activism and decolonizing work led by many Indigenous and coastal communities, there is increasing recognition of the need to reclaim ocean science by amplifying the knowledge and perspectives of those who have long understood the seas. Ocean science is just beginning this deep process of decolonization, which first seeks to acknowledge and reckon with the violent and ongoing impacts of colonization. Decolonizing ocean science requires a fundamental shift in who holds agency and sovereignty over their own ocean waters. This also relates to international waters and who is included, excluded, or undervalued throughout negotiations concerning the legal and regulatory structures governing global oceans. Today, many ocean scientific projects are co-designed and co-led by ocean knowledge holders from diverse backgrounds. Collecting ocean datasets requires a team of experts who follow cultural protocols, ensure environmental safety, and apply diverse scientific methods, all while striving for more relational practices. ### 1.2 Ocean Data Collection Today Today, ocean datasets are collected by ocean scientists in collaboration with ocean engineers. These datasets are gathered from several sources to understand the global ocean and its role in maintaining Earth's habitability and critical planetary cycles. Ocean engineers develop the tools, platforms, and instruments that are required for data collection, such as underwater vehicles, satellite-mounted sensors, and buoys. By designing technologies that can operate in diverse and sometimes extreme conditions, these engineers support and expand ocean scientific capabilities. Together, ocean scientists and engineers advance our understanding of the planet for both research and conservation. There is a considerable variety of ocean data types, tools for data collection, and associated databases to store these recorded entities. This diversity of datasets is outlined in section 2. Like most scientific fields, funding can be secured from public governmental bodies or private sources. The ocean datasets we focus on here are publicly accessible and typically funded by governments via taxpayer contributions. This means that ocean datasets are for the people and should be accessible. Unfortunately, many public ocean datasets are stocked in complex databases and require specialized software, programming experience, or extensive knowledge to access. That said, there are plenty of datasets that can be retrieved and visualized more easily, with little to no background knowledge. There are also some ocean datasets that can be accessed with helpful tips and instructions, which is what we will focus on here. The Exposing the Invisible Toolkit is designed as a self-learning resource, we hope that this guide will support future investigations and make ocean datasets more accessible to communities and investigators around the world. ### 1.3 Data Gaps, Capacity Initiatives, Ocean Defenders Conducting ocean science can be a costly endeavor. Depending on the environment, scientific goals, and technical requirements, some ocean scientific work can only be conducted by wealthy nations or private organizations. Typically, this kind of science takes place at sea or uses remote sensing techniques. For example, deep ocean exploration and research requires a ship, vehicles, or platforms to deploy down to the deep ocean, technical and computing systems aboard to process the data, and a diverse team of experts to manage these operations. In contrast, satellite remote sensing used for ocean research typically covers the entire Earth surface. Publicly funded satellite-derived ocean datasets can be useful across territorial waters, throughout international seas, and are accessible regardless of one's nationality. Near-shore ocean science and *in situ* coastal monitoring efforts are more financially affordable, especially as diverse knowledge systems merge with lower-cost technologies and capacity initiatives. In this context, capacity refers to the skills, resources, and knowledge needed to effectively conduct ocean science. As ocean science undergoes a process of decolonization, this emphasis on capacity building, development, and sharing is also strengthening. Additionally, several international ocean law and policy frameworks specifically aim to increase ocean science capacity. Ocean defenders are also central to these efforts. As groups, individuals, or organizations dedicated to protecting marine environments, defenders play a key role in advocating for capacity building within and outside scientific structures. Many defenders are fisherpeople, coastal communities, or groups directly affected by changes in climate and ocean health. Beyond advocating for sustainable and generative oceanic futures, they also fight to overcome political resistance and funding barriers. Ocean defenders, like land defenders, face challenging or dangerous obstacles while pushing for local and global ocean preservation. Ocean science and policy clearly needs collaborative approaches that bring multiple knowledge systems forward while prioritizing those most impacted by climate change, pollution, and other threats to both marine habitats and coastal livelihoods.  *Ocean-defending artisanal fishers and their supporters in South Africa celebrate upon receiving the news that Shell’s permit to conduct a seismic survey on the Wild Coast had been set aside by the Makhanda High Court, in September 2022. Photo credit: Taryn Pereira. Source: <https://oceandefendersproject.org/case-study/no-to-seismic-surveys/>* **More on capacity initiatives, knowledge gaps, and ocean defenders:** - Guilhon, M., M. Vierros, H. Harden-Davies, D. Amon, S. Cambronero-Solano, C. Gaebel, K. Hassanali, V. Lopes, A. McCarthy, A. Polejack, G. Sant, J.S. Veiga, A. Sekinairai, and S. Talma. (2025). [Measuring the success of ocean capacity initiatives.](https://doi.org/10.5670/oceanog.2025.122) *Oceanography* 38(1). - Saba, A.O., I.O. Elegbede, J.K. Ansong, V.O. Eyo, P.E. Akpan, T.O. Sogbanmu, M.F. Akinwunmi, N. Merolyne, A.H. Mohamed, O.A. Nubi, and A.O. Lawal-Are. 2025. [Building ocean science capacity in Africa: Impacts and challenges.](https://doi.org/10.5670/oceanog.2025.133) *Oceanography* 38(1). - Behl, M., Cooper, S., Garza, C., Kolesar, S. E., Legg, S., Lewis, J. C., White, L., & Jones, B. (2021). [Changing the culture of coastal, ocean, and marine sciences: strategies for individual and collective actions.](https://www.jstor.org/stable/27051390) *Oceanography*, *34*(3), 53–60. - The Ocean Defenders Project (2025). [Ocean Defenders: Protectors of our ocean environment and human rights.](https://oceandefendersproject.org/project-publications/ocean-defenders-protectors-of-our-ocean-environment-and-human-rights/) The Peopled Seas Initiative, Vancouver, Canada. - Belhabib, D. (2021) [Ocean science and advocacy work better when decolonized.](https://doi.org/10.1038/s41559-021-01477-1) *Nat Ecol Evol* 5, 709–710. - Kennedy, R. and Rotjan, R. (2023). [Mind the gap: comparing exploration effort with global biodiversity patterns and climate projects to determine ocean areas with greatest exploration needs.](https://doi.org/10.3389/fmars.2023.1219799) *Front. Mar. Sci.* (10). - Bell, K.L.C, Johannes, K.N., Kennedy, B.R.C., & Poulton, S.E. (2025) [How little we’ve seen: A visual coverage estimate of the deep seafloor.](https://www.science.org/doi/10.1126/sciadv.adp8602) *Science Advances*, Vol 11, Issue 19. ### 1.4 Ocean Datasets for Investigations and Storytelling Ocean datasets play a crucial role in both scientific investigations and storytelling by providing evidence-based insights into the health and dynamics of our global ocean. These datasets help scientists better understand the ocean, but beyond research, they serve an important role in ocean-related investigations. Ocean datasets can support communities, journalists, and activists in raising data-backed awareness about critical marine and environmental justice issues. By sharing this data in accessible ways, oceanographic narratives can amplify the voices of coastal communities, engage the public, and inspire action in support of more regenerative ocean futures. Numerous well-resourced journalistic and forensic organizations use ocean data to support their stories or reporting, such as Forensic Architecture, LIMINAL, Forensis, Earshot, Border Forensics, and others. In this guide, we will demonstrate how you can access datasets and conduct your own oceanic investigations. By the end, you will be able to illustrate a well-defined ocean or climate question using publicly available oceanographic datasets and media collections, which will enhance your evidence-based and visually engaging story. ## 2. Diversity of Data Types The following sub-sections serve as a collection of key ocean data types, organized by how they are collected and what they reveal. Each broad data source type (i.e., the overarching technique used to gather certain kinds of ocean data) begins with a bit of history and trade-offs, and then is further broken down into specific data products. For each data product, we share use cases, data collection methods, key considerations, and some databases (often from U.S. and European agencies), followed by resources to learn more. This structure is designed to help you understand how each dataset is produced, grasp the significance of the data, and know where to go for deeper investigations or analyses. There are many data types presented below which are organized in these four broad categories: *in situ* sensors; deep ocean observation, exploration, and research; mapping; and, satellite remote sensing. See Section 3 to review an example case study which demonstrates how some of these datasets may be used to support an investigative ocean and climate story.  *Illustration of a range of ocean sensor platforms—ships, profiling drifters, gliders, moored buoys, and satellites. Source: <https://www.ecmwf.int/en/about/media-centre/news/2021/world-meteorological-day-focuses-role-ocean-weather-and-climate>* ### 2.1 *In Situ* Sensors *In situ* sensing refers to the direct measurement of ocean properties using instruments that are physically located within the water. Sensors for measuring temperature, salinity, pressure, currents, and of biochemical concentrations are deployed on a range of platforms with various advantages and drawbacks. While satellites provide broad spatial coverage of the ocean’s surface, *in situ* platforms are essential for monitoring the ocean’s interior, tracking coastal change, and measuring water properties that cannot be detected remotely. - **History:** - Sailors have used thermometers to measure ocean temperature since at least as early as Captain James Cook’s [1772 voyage](https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/rog.20022) to the Antarctic Circle (another example of colonial science forming the foreground to Western ocean science). - The Nansen bottle (1896) and later the Niskin bottle (1966) enabled the capture of water samples at specific depths, which could then be pulled up and tested for temperature and salinity on the ship. - The first bathythermograph was developed in 1938 and featured a temperature sensor on a wire which recorded a temperature profile as it was lowered into the ocean. This was used by the US Navy in WWII to improve sonar accuracy, since temperature layers alter acoustic propagation. - Today, there are thousands of advanced sensors across the world’s oceans which transmit readings in real time via satellite. - **Trade-offs:** - *In situ* sensors can only measure the ocean properties at their exact location so great consideration is taken in their placement and timing. - Powering the instruments is a constant challenge and factors into decisions about sampling frequency. - The extreme pressure in the deep ocean limits the operating depth of some sensors. - Harsh operating conditions limit the lifespan of sensors and necessitates regular maintenance/replacement, often in remote locations at high costs. This leads to less accessible areas being undersampled. #### 2.1.1 Moorings and Fixed Platforms  *Example of a coastal data mooring. Source: Baily et al., 2019, <https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2019.00180/full>* - **About:** - A range of platforms anchored in place, collecting time-series data at a fixed location. - Used in the deep ocean (e.g., the TAO array across the Pacific Ocean), the continental shelf (e.g., NOAA’s National Data Buoy Network), and at the coastline (e.g., tide gauges). - **Use cases:** - Long-term climate monitoring of ocean heat content and circulation. - Data inputs for forecast models, improves accuracy of ocean and weather predictions. - Early warning systems for tsunamis and hurricane storm surge. Tracking tidal heights and local sea level rise. - Water quality monitoring for pollutants, algal blooms, and hypoxia. - **Data collection:** - Sensor packages measure temperature, salinity, pressure, biochemistry, and more. - Some moorings have Acoustic Doppler Current Profilers (ADCPs) to measure water current velocities throughout the water column. - **Key considerations:** - Data today is mostly broadcasted in near-real time, but there are some platforms that require physical retrieval before the data is downloaded. - Spatial coverage is extremely limited and focused around a small set of nations. - The diversity of databases and data types can pose a challenge for accessing and working with the data. - **Data sources:** - [US NOAA National Data Buoy Center (NDBC)](https://www.ndbc.noaa.gov/) - [EU Copernicus Marine Service In Situ dashboard](https://marineinsitu.eu/dashboard/) - [Global Tropical Moored Buoy Array](https://www.pmel.noaa.gov/gtmba/) - **Resources to learn more:** - [WHOI - Moorings & Buoys](https://www.whoi.edu/what-we-do/explore/instruments/instruments-moorings-buoys/) - [Ocean Observatories Initiative](https://oceanobservatories.org/) #### 2.1.2 Drifters and Floats  *Map of Argo float locations. Source: <https://argo.ucsd.edu/about/status/>* - **About:** - Unanchored and unpropelled instruments that drift with the currents and take ocean measurements. - Drifters stay at the surface and provide information about surface conditions, and surface currents are calculated from their GPS trajectory. - Floats profile the water column by adjusting their buoyancy to move up and down. The Argo program is the largest and most significant, with over 4,000 Argo floats profiling the world’s oceans. - Capable of providing global coverage at lower cost than moored sensors, especially in remote open-ocean regions. - **Use cases:** - Drifters: Mapping near-surface currents and tracking pollutants and marine debris transport. - Floats: Measuring subsurface temperature and salinity for climate studies. Some Argo floats also have biochemical sensors. - **Data collection:** - Drifters: GPS-tracked, measure SST, pressure, sometimes salinity, sometimes waves. - Argo floats: Profile down to 2,000 m every 10 days, transmitting data via satellite. - **Key considerations:** - Drifters and floats are always moving, so you can’t get a clean timeseries for a single location like you can with moorings. Additionally, Argo floats only take one profile every 10 days in order to preserve battery life. - Argo floats don’t generally operate near the coast on the continental shelf. - Some drifters/floats lack real-time telemetry (data transmission). - **Data sources:** - [Global Drifter Program](https://www.aoml.noaa.gov/phod/gdp/) - [Argo Program](https://argo.ucsd.edu/) - [Copernicus Marine Service drifters](https://data.marine.copernicus.eu/products?facets=featureTypes%7ETrajectory) - [SOCCOM Biogeochemical Floats](https://soccom.org/) - **Resources to learn more:** - [About Argo](https://argo.ucsd.edu/about/) #### 2.1.4 Autonomous Vehicles - ASVs and Gliders  *Illustration of a glider’s sawtooth propulsion pattern. Source: <https://earthzine.org/going-deep-to-go-far-how-dive-depth-impacts-seaglider-range/>* - **About:** - Autonomous Surface Vehicles (ASVs) and gliders are robotic platforms that enable long-duration, energy-efficient monitoring over vast areas. - Bridging the gap between targeted, expensive, ship-based measurements and low-cost, but uncontrolled, drifters/floats. - **Use cases:** - Significant overlap with the use cases for moored sensors and drifters/floats. - Targeted measurements in dangerous conditions like hurricanes. - Mapping surveys, autonomous of ship or in tandem with other vehicles. - **Data collection:** - Autonomous Surface Vehicles (ASVs) use solar panels, wind, or waves as a power source to supplement and recharge their batteries. - Gliders are underwater vehicles that create propulsion by adjusting their buoyancy and gliding horizontally while sinking/rising (similar to an airplane). This enables longer battery range than propellers or thrusters. - **Key considerations:** - Gliders and ASVs are often used in targeted studies rather than continuous global monitoring and thus have lower data availability. - Shorter mission durations than moorings or drifters/floats. - **Data sources:** - [NOAA Glider Data Assembly Center](https://gliders.ioos.us/) - [OceanGliders](https://www.oceangliders.org/) - **Resources to learn more:** - [National Oceanography Centre UK - Gliders](https://noc.ac.uk/facilities/marine-autonomous-robotic-systems/gliders) ### 2.2 Deep Ocean Observation, Exploration, and Research Deep ocean science is typically conducted to observe long-term changes at specific seafloor sites, to explore marine habitats that are unknown to science, and to conduct applied or experimental research on focused environmental questions. A range of deep ocean data collection tools include platforms or landers, cabled observatories, and deep submergence systems—such as human-occupied vehicles (HOVs), remotely-occupied vehicles (ROVs), and autonomous underwater vehicles (AUVs).  *Human-occupied vehicle (HOV) 'Alvin' being recovered in 2024 during an expedition to the East Pacific Rise hydrothermal vent fields. Photo credit: Mae Lubetkin* - **History** - 1872–1876: The *HMS Challenger* expedition, a milestone for deep ocean science but deeply tied to imperialism - Mid-20th century: Cold War military priorities further developed submersible vehicle capabilities, leading to *Trieste*’s 1960 Mariana Trench dive - 1964 onward: HOVs expanded access to the deep ocean, e.g., *Alvin* (US), *Nautile* (France), *Shinkai* (Japan), and *Mir* (Russia) - 1980s–2000s: ROVs and AUVs developed by industry (oil, mining, and defense) and scientific institutions, in parallel - 2000s–present: Cabled observatories (e.g., Ocean Networks Canada, DONET in Japan), public research campaigns (e.g., NOAA, IFREMER), and oceanographic instruments expanded reach and scope. - Today: Many regions still face barriers to participation and funding in deep ocean science (as outlined in the introduction). Meanwhile, deep submergence science in wealthy nations increasingly utilizes AI, autonomous systems, 4K and 3-D imaging techniques. - **Trade-offs:** - Provides direct access to deep ocean environments which are inaccessible by surface vessels or remote sensing - High spatial and contextual resolution: can capture detailed imagery, samples, and detailed *in situ* measurements - Resource-intensive: operations usually require ships, launch/recovery teams, and specialized personnel - Limited temporal and spatial coverage: data collection is episodic, site-specific, and dependent on expedition funding, schedules, and weather conditions at-sea - High costs and technical barriers mean deep ocean science is dominated by a few well-funded institutions or nations, with limited global access - Colonial legacies persist in relation to who sets research agendas, who makes funding decisions, and who benefits from collected data #### 2.2.1 Deep Submergence Systems (HOVs, ROVs, AUVs) - **About:** - Vehicles that operate in the deep ocean water column or along the seafloor, including: - Human-occupied vehicles (HOVs): carry scientists directly, typically 1-3 observers and a pilot - Remotely operated vehicles (ROVs): tethered and piloted from a surface vessel like a ship - Autonomous underwater vehicles (AUVs): untethered and pre-programmed - These systems can operate from hours to days and are depth-rated around 4000-6000 m, but some may reach full ocean depths (11 km) and others may work well in shallower waters. - **Use cases:** - High-resolution visual surveys - Precise targeted sampling with environmental context and imagery at diverse environments including hydrothermal vents, methane seeps, cold-water coral habitats, and others - Biogeographic habitat mapping - Wreck exploration and infrastructure inspection - Imagery of deep ocean environments can support visual storytelling, public engagement, and education - **Data collection:** - Data is streamed directly to the support vessel for tethered operations - While for untethered submersibles (HOVs and AUVs) most data is retrieved when the vehicle is recovered - All physical samples are retrieved and processed upon vehicle recovery - **Key considerations:** - Requires experienced pilots and operational support (expensive) - AUVs need detailed mission planning (mission failure could lead to vehicle loss) - Navigation and environmental risks must be managed carefully - **Data sources:** - SeaDataNet - [EU research vessel data](https://csr.seadatanet.org/), including cruise summary reports and more - EuroFleets - European initiative to compile [EU research cruise data](https://www.eurofleets.eu/data/) - Rolling Deck to Repository - [US research vessel data](https://www.rvdata.us/data), including: expedition summary, shiptrack navigation, scientific sampling event log, post-processed data - JAMSTEC Databases - [Japan research vessel data](https://www.jamstec.go.jp/e/database/), including: HOV *Shinkai 6500* and ROV *Kaiko* mission data, cruise reports, and dive logs - **Resources to learn more:** - [Woods Hole Oceanographic Institution - National Deep Submergence Facility](https://ndsf.whoi.edu/) - [Ocean Exploration Trust - Science and Technology](https://nautiluslive.org/science-tech)  *An imaging elevator equipped with two camera systems, lights, battery packs, and crates to store additional sampling tools to be used by an HOV during a dive in the same region. Source: Woods Hole Oceanographic Institution* #### 2.2.2 Landers and Elevators - **About:** - Landers are relatively simple systems that descend to the seafloor and remain stationary for the duration of their deployment. - They are sometimes referred to as 'elevators' since they descend to the seafloor then ascend back to the surface - There are no people on landers, but they typically carry sensors, cameras, samplers, and other instruments - Depending on power supply and scientific goals, they can spend hours to sometimes months on the seafloor - **Use cases:** - Collecting environmental data (e.g., conductivity, temperature, pH, oxygen) - Capturing imagery of habitats or operations - Deploying baited cameras or traps to study biodiversity - Using the platform to carry additional gear or instruments to the seafloor that a deep submergence system could not transport on its own due to space limitations - **Data collection:** - Typically data is retrieved when the lander is recovered back on deck - Some landers will transmit data acoustically or remotely from the seafloor - The frequency that imagery or other data are collected is pre-programmed before deployment - **Key considerations:** - Requires careful site selection, recovery planning, and often ship time - Currents can impact the intended landing location on the seafloor, sometimes drifting the platform or lander far off-site - Limited in capabilities, not as advanced as deep submergence vehicles, but also much cheaper and easier to custom build - **Data sources:** - Deep ocean lander data (e.g., imagery and environmental sensor data) would be found in the same databases and repositories listed in section 2.2.1 Deep Submergence Systems (HOVs, ROVs, AUVs). - **Resources to learn more:** - [Schmidt Ocean Institute - Elevators and Landers](https://schmidtocean.org/technology/elevators-landers/) - [The Deep Autonomous Profiler (DAP), a Platform for Hadal Profiling and Water Sample Collection (Muir et al., 2021)](https://journals.ametsoc.org/view/journals/atot/38/10/JTECH-D-20-0139.1.xml) - [Lander Lab: Technologies, Strategies and Use of Ocean Landers (Hardy, 2022)](https://magazines.marinelink.com/Magazines/MarineTechnology/202201/content/technologies-strategies-landers-594271)  *Map of Ocean Networks Canada NEPTUNE and VENUS Observatories near Vancouver Island, Canada. Each orange square represents a node or station along the cabled observatory where instruments or sensors are mounted. Source: <https://www.oceannetworks.ca/>* #### 2.2.3 Cabled Observatories - **About:** - Mostly permanent, wired infrastructure on the seafloor that transmit real-time power and data via fiber optic cables connected to shore stations - Similar in some ways to the data collection tools described in section 2.1 *In Situ* Sensors, but these networks are fixed in location and networked - People do not visit these observatories, instead they support a wide range of sensors (e.g., temperature, pressure, seismometers, hydrophones, cameras, samplers) - Can integrate with ROVs or AUV docking stations, and are also typically maintained and serviced by ROVs - Designed for continuous, high-frequency monitoring of deep ocean processes across years or decades - They connect highly diverse environments from hydrothermal vent regions to abyssal plains and continental shelves - **Use cases:** - Long-term and consistent monitoring of geophysical activity (e.g., earthquakes, hydrothermal vents) - Real-time data for early warning systems (e.g., tsunamis, gas releases) - To study oceanographic processes (e.g., currents, biogeochemical fluxes, and ecosystem change) - Supports public engagement and education through livestreams - **Data collection:** - Real-time data is livestreamed to shore stations and then available via online portals - Most are operated by national or international research infrastructures - **Key considerations:** - Extremely costly, high maintenance needs (ROVs are often used for annual servicing) - Site selection is key since they are fixed installations - **Data sources:** - Ocean Networks Canada - [Oceans 3.0 Data Portal](https://data.oceannetworks.ca/), including all datasets, dashboards, and visualizers (more info on [ONC data](https://www.oceannetworks.ca/data/)) - US Ocean Observatories Initiative - [OOI Data Portal](https://oceanobservatories.org/data-portal/), includes cable-linked arrays on East and West Coasts and deep Pacific - EU [EMSO ERIC Data Portal](https://data.emso.eu/home), real-time and archived data, tools and research environment to investigate seafloor observatories across European margins - **Resources to learn more:** - [Ocean Observatories Initiative](https://oceanobservatories.org/observatories/) - arrays, infrastructure, instruments - [Ocean Networks Canada](https://www.oceannetworks.ca/observatories/) - observatories - [Interactive map of ONC locations](https://www.arcgis.com/home/webmap/viewer.html?webmap=fcea4e5f087f41c58bcc5e51b13fffa1&extent=-158.3094,39.6681,-29.8133,75.4182) - [Regional Cabled Observatories](https://www.whoi.edu/what-we-do/explore/ocean-observatories/about-ocean-observatories/types-of-observatories/regional-cabled-observatories/) - summary by Woods Hole Oceanographic Institution ### 2.3 Mapping Marine hydrography or ocean scientific mapping involves the creation of high-resolution representations of the seafloor, water column, and other associated features or phenomena (e.g., fish migrations, vents or seeps bubbling up) using vessel-based sonar, autonomous vehicles, acoustic or optical tools. It is a type of remote sensing since the mapping instrument is not on the seafloor. Unlike satellite remote sensing, which observes only the ocean surface from space, hydrographic mapping is conducted from platforms within or on the ocean surface. These mapping systems can resolve fine-scale topography (seafloor bathymetry), subsurface geologic layers, water column imaging, and habitat mapping that integrates both physical and biological data. Laser and 3-D reconstruction are other forms of high-resolution mapping. - **History:** - 1870s–1900s: Early bathymetric charts created using lead lines, linked to colonial navigation and maritime claims as well as scientific knowledge creation - 1920s–1940s: Echo sounding developed for military and commercial navigation, later repurposed for seafloor mapping - 1950s–1970s: Multibeam sonar developed, enabling wider swath coverage (i.e. can map wider seafloor area, not just single points) and broader seafloor topography or bathymetry mapping - 2010s–present: Autonomous vehicles, numerous specialized sonar systems, and 3-D photogrammetry advance deep mapping capabilities - Today: Mapping remains uneven globally—nations with limited funding or access to ships and processing capacity are underrepresented and do not have detailed seafloor maps of the their waters - **Trade-offs:** - High-resolution, fine-scale maps of seafloor and water column features - Enables geologic, biologic, and habitat-based spatial analysis - Requires significant ship time, technical expertise, and post-processing - Data coverage is patchy, most of the seafloor remains unmapped - High cost and national interests impact where mapping occurs and who benefits from the data  *Bathymetric mapping using a hull-mounted multibeam sonar system. Black lines indicate the ship’s track, while the coloration represents depth differences (red is shallow, purple is deep) used for visualizing the bathymetric or topographic features of the seafloor. Source: https://www.worldofitech.com/mapping-the-ocean-floor-water-bathymetry-data/>* #### 2.3.1 Bathymetric Mapping - **About:** - Measurement and charting of the depth and shape of the seafloor - Typically uses sonar-based systems (e.g., single-beam or multibeam echosounders), mounted on ships, AUVs, or towed platforms - Short-range systems (e.g., ROV-mounted sonar) provide highly detailed data over small areas (centimeters in resolution), while medium-range systems (e.g., hull-mounted multibeam on ships or AUVs) cover much larger swaths with lower resolution - **Use cases:** - Mapping underwater topography and geological features - Planning submersible dives and identifying hazards - Supporting infrastructure projects like cables or offshore wind farms - Creating base maps for habitat mapping or biogeographic studies (i.e. understanding what marine life lives where and how their habitats are linked to geologic features as well as currents and physical oceanographic phenomena) - **Data collection:** - By research vessels or autonomous vehicles using sonar systems - Key manufacturers include Kongsberg, Teledyne, R2Sonic, and Edgetech - Data is processed using specialized hydrographic software (e.g., QPS Qimera, CARIS, MB-System) - **Key considerations:** - Requires calibration (e.g., sound speed profiles) and correction for vessel motion - Deep ocean mapping can be slow and resource-intensive - Interpretation of raw bathymetry data requires trained analysts and geospatial tools, it is not yet fully automated - **Data sources:** - European Marine Observation and Data Network - [EMODnet Bathymetry](https://emodnet.ec.europa.eu/en/bathymetry), multibeam datasets and other maps with a built in visualizer - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access), archive of US and global bathymetric surveys with visualizer - General Bathymetric Chart of the Oceans - [GEBCO Gridded Bathymetry Data](https://www.gebco.net/data-products/gridded-bathymetry-data), global map interface of compiled bathymetry - Global Multi-Resolution Topography - [GMRT data](https://www.gmrt.org/), global compilation of multibeam data (includes graphical map tool) - **Resources to learn more:** - [NOAA - What is bathymetry?](https://oceanservice.noaa.gov/facts/bathymetry.html) - [Seabed 2030](https://seabed2030.org/) – global initiative to map the entire seafloor by 2030 - [Using open-access mapping interfaces to advance deep ocean understanding](https://link.springer.com/article/10.1007/s40012-025-00410-2) (Johannes, 2025) - [GeoMapApp](https://www.geomapapp.org/) - free map-based application for browsing, visualizing and analyzing a diverse suite of curated global and regional geoscience data sets  *Plumes of bubbles emanating from the seafloor, indicating that there were methane gas seep ecosystems in this region. Sound waves reflect strongly off the gas bubbles and are visible in the water column data. Source: <https://nautiluslive.org/blog/2018/08/08/more-just-bathymetry-seafloor-mapping-tool-exploration>* #### 2.3.2 Water Column Mapping - **About:** - Acoustic systems—usually multibeam echosounders or special water column sonars—to detect and visualize features suspended in the ocean between the surface and the seafloor - Key for detecting gas plumes, biological layers (schools of fish or migrations in the twilight zone), suspended sediments, etc. - **Use cases:** - Observing midwater scattering layers (biological migrations) - Detecting hydrothermal plumes amd tracking gas plumes from methane seeps - **Data collection:** - Most multibeam systems include water column data modes, so it is often collected in tandem with bathymetry - From ships, AUVs, and ROVs - Data must be interpreted alongside oceanographic profiles (e.g., CTD casts) and often requires manual cleaning to reduce noise - **Key considerations:** - Processing and interpreting water column data is a bit tedious not yet standardized - Detection is sensitive to sonar frequency, range, and sea conditions - Validation with ground-truth sampling (e.g., bottle casts, nets, sensors) is helpful - **Data sources:** - European Marine Observation and Data Network - [EMODnet Physics](https://emodnet.ec.europa.eu/en/physics), includes some water column data layers - NOAA National Centers for Environmental Information - [NCEI water column sonar data](https://www.ncei.noaa.gov/maps/water-column-sonar/) - **Resources to learn more:** - [OET - More than just Bathymetry](https://nautiluslive.org/blog/2018/08/08/more-just-bathymetry-seafloor-mapping-tool-exploration) - Seafloor Mapping as a Tool for Exploration - [Migration in the Ocean Twilight Zone](https://twilightzone.whoi.edu/explore-the-otz/migration/) - often monitored with water column data #### 2.3.3 Seafloor Backscatter - **About:** - Analyzing the intensity of sound that is reflected or ‘scattered back’ from the seafloor when using sonar systems - Provides information about seafloor texture, hardness, and composition (e.g., sand, rock, mud) - Often conducted simultaneously with bathymetric mapping during ship-based or AUV surveys - **Use cases:** - Seafloor habitat classification or substrate mapping - Detecting anthropogenic objects or features (e.g., cables, wrecks) - Complements bathymetry for geologic or habitat models - **Data collection:** - Similar to bathymetry and water column mapping, backscatter data is collected using the same sonar systems and processed using similar software - **Key considerations:** - Requires calibration and post-processing to produce usable mosaics - Interpretation of sediment type from backscatter typically should be verified by ground-truth sampling (e.g., grabs, cores with ROVs or HOVs) - **Data sources:** - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access) - **Resources to learn more:** - NOAA - [How does backscatter help us understand the sea floor?](https://oceanservice.noaa.gov/facts/backscatter.html)  *Sediment layers seen in the sub-bottom profiler data collected in 2021 at the New England and Corner Rise Seamounts expedition on the NOAA Ship 'Okeanos Explorer'. Source: <https://oceanexplorer.noaa.gov/technology/sub-bottom-profiler/sub-bottom-profiler.html>* #### 2.3.4 Sub-bottom Profiling - **About:** - Uses low-frequency acoustic pulses to penetrate below the seabed and image sediment layers or other buried geologic features - Reveals vertical structures beneath the seafloor - Typically deployed from research vessels or towed systems - **Use cases:** - Studying sedimentation and geological processes - Locating subseafloor gas pockets or archaeological sites - For infrastructure planning or hazard assessment (e.g., submarine landslides) - **Data collection:** - Chirp profilers (high resolution, shallow penetration) and boomer/sparker systems (deeper penetration) are used - Operated from vessels with sonar equipment often collected simultaneously while collecting bathymetry and other mapping data, even if the sonar systems are different - **Key considerations:** - Resolution and penetration are inversely related (deeper = less detail) - Can be noisy and hard to interpret without ground-truthing (e.g., sediment cores) - **Data sources:** - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access) - European Marine Observation and Data Network - [EMODnet Geology](https://emodnet.ec.europa.eu/en/geology), includes sub-bottom and other forms of seafloor geological data - **Resources to learn more:** - NOAA - [Sub-Bottom Profiler](https://oceanexplorer.noaa.gov/technology/sub-bottom-profiler/sub-bottom-profiler.html)  *3-D reconstructed seafloor lava flows and hydrothermal vent field from the East Pacific Rise. This 3-D model was produced using downward facing video imagery and photogrammetry techniques. Credit: Mae Lubetkin* #### 2.3.5 Photogrammetry and 3-D Reconstruction - **About:** - Stitching together overlapping images or video frames from subsea camera systems often mounted on ROVs or AUVs - To create detailed mosaics or 3-D models of seafloor features - Uses optical data, offering true-color, high-resolution imagery (unlike acoustic mapping techniques described above) - **Use cases:** - Mapping hydrothermal vent fields, coral reefs, archaeological sites, etc. - Change detection in dynamic environments (e.g., volcanic or vent habitats, biological growth or loss) - Public engagement and educational tools - **Data collection:** - Collected by vehicle-mounted cameras with precise navigation and positioning - Software like Agisoft Metashape or custom photogrammetry pipelines are used for processing (which is easier now than ever before, becoming much more common in ocean sciences) - **Key considerations:** - Requires good lighting and water clarity - Processing is computationally intensive, and vehicle navigation data helps with plotting 3-D reconstructions onto broader bathymetric maps - Can be limited to small survey areas due to time constraints and battery limitations - **Data sources:** - Monterey Bay Aquarium Research Institute - [MBARI Sketchfab](https://sketchfab.com/mbari) - 3-D models of seafloor sites can be found in academic papers or at individual institutions or government agencies data repositories - **Resources to learn more:** - [Seafloor Futures](https://garden.ocean-archive.org/seafloor-futures/) (Lubetkin, 2024) - [Realtime Underwater Modeling and Immersion](https://nautiluslive.org/tech/realtime-underwater-modeling-and-immersion) - Ocean Exploration Trust - [Underwater 3-D Reconstruction from Video or Still Imagery: Matisse and 3-D Metrics Processing and Exploitation Software](https://www.mdpi.com/2077-1312/11/5/985) (Arnaubec et al., 2023) - [Seeing the Sea in 3-D](https://schmidtocean.org/cruise-log-post/seeing-the-sea-in-3d/) - Schmidt Ocean Institute ### 2.4 Satellite Remote Sensing Satellite data provides the most familiar, spatially complete picture of the ocean. This bird’s-eye perspective is invaluable for understanding large-scale phenomena like currents, sea surface temperature patterns, and phytoplankton blooms, providing visual evidence that can enhance storytelling. However, there are unique considerations since, unlike the use of satellite imagery on land, most of our understanding of the ocean does not come from the visual spectrum. In this section, we’ll introduce three of the most important types of satellite ocean data and discuss the use cases for each. - **History:** - 1978: NASA launched Seasat, the first satellite designed for ocean research. - Significant expansion in the 1990s with missions including TOPEX/Poseidon (ocean altimetry), AVHRR (high-resolution sea surface temperature), and SeaWiFS (ocean biology). - Modern constellations are operated by NASA, NOAA, ESA, EUMETSAT, CNES, ISRO, and others. - **Trade-offs:** - Excellent spatial coverage that’s impossible to achieve with ships or buoys. - Very high costs, these platforms are operated by government agencies. - Only seeing the very surface of the ocean, no subsurface data. - Limited horizontal resolution (spatial detail) and temporal resolution (orbital repeat time).  *Gulf of Mexico SST on a cloud-free data. Source: <https://marine.rutgers.edu/cool/data/satellites/imagery/>* #### 2.4.1 Radiometry - Sea Surface Temperature (SST) - **About:** - Sea surface temperature (SST) is the oldest and most extensive application of satellite oceanography. - **Use cases:** - Tracking climate change, El Niño, and marine heat waves. - Key input for weather models (e.g., very important for hurricane forecasting). - Mapping ocean eddies, currents, and upwelling, which are critical to fisheries. - **Data collection:** - Two separate types of sensors measure SST: Infrared and microwave. - IR sensors have higher spatial resolution ([1-4 km](https://coastwatch.noaa.gov/cwn/product-families/sea-surface-temperature.html)) and finer temporal coverage but cannot “see” through clouds, which block over 70% of the ocean at any given time. - Microwave sensors can see through most non-precipitating clouds but have a lower spatial resolution (about 25 km) and don’t work near the coastline. - Measures temperature of the top ~1 mm of the ocean - Blended products: Combine multiple sensors for better coverage (e.g., GHRSST L4)  *Example SST data at different processing levels. (Merchant et al. 2019): <https://www.nature.com/articles/s41597-019-0236-x>* - **Key considerations:** - Make yourself aware of the different processing levels when accessing data. Level 4 (L4) will be the easiest to work with but may not be fully accurate. - L2: Data along the original orbital track. - L3: Gridded data, sometimes averaged over time. - L4: Cloud-free, gaps are filled by various methods depending on the source. - Temporal resolution depends on the satellite orbit. There are good options that blend multiple satellites. - **Data sources:** - [EU Copernicus Marine Service](https://marine.copernicus.eu/) - [US NASA Physical Oceanography DAAC](https://podaac.jpl.nasa.gov/) - [NOAA CoastWatch](https://coastwatch.noaa.gov/cw_html/cwViewer.html) - Graphical Interface - **Resources to learn more:** - [Group for High Resolution Sea Surface Temperature (GHRSST)](https://www.ghrsst.org/ghrsst-data-services/for-sst-data-users/) #### 2.4.2 Radar Altimetry - Sea Surface Height (SSH) - **About:** - Measures ocean surface height by sending radio pulses and measuring return time. - SSH can tell us the strength of large scale currents like the Gulf Stream, as the slope of the sea surface is used to calculate the “geostrophic current”. - **Use cases:** - Key to understanding ocean circulation and long-term sea level rise. - **Data collection:** - Radar altimeters on satellites measure SSH directly (e.g., Jason-3, Sentinel-6) and then the geostrophic currents are calculated in a post-processing step. - Spatial resolution is significantly worse than SST (25+ km) - The recent SWOT satellite is a new type of altimeter with much higher resolution but has very limited coverage since there is only one currently in orbit. - **Key considerations:** - SSH is useful for large-scale ocean currents but not coastal tidal currents. - Similar to SST, be careful about processing level and look for re-gridded datasets. - Can generally see through clouds, so gaps are not a significant issue. - **Data sources:** - [EU Copernicus Marine Service](https://data.marine.copernicus.eu/products?facets=mainVariables%7ESea+surface+height--sources%7ESatellite+observations) and [Aviso](https://www.aviso.altimetry.fr/en/data/products/sea-surface-height-products/global.html) - [US NASA PODAAC](https://podaac.jpl.nasa.gov/NASA-SSH) - [Copernicus MyOcean Pro](https://data.marine.copernicus.eu/viewer/expert) and [Aviso](https://seewater.aviso.altimetry.fr/) - Graphical Interfaces - **Resources to learn more:** - [NASA JPL - What is Ocean Surface Topography?](https://podaac.jpl.nasa.gov/OceanSurfaceTopography)  *Global map of marine Chlorophyll concentration. Source: <https://sos.noaa.gov/catalog/datasets/biosphere-marine-chlorophyll-concentration/>* #### 2.4.3 Optical - “Ocean Color” - **About:** - Ocean color sensors measure the reflectance of sunlight from the ocean surface to infer biological and chemical properties, such as algal concentration, suspended sediments, and water clarity. - **Use cases:** - Tracking phytoplankton blooms and changes in marine ecosystems. - Useful for monitoring water quality, including coastal sediment and oil spills. - **Data collection:** - Sensors measure light reflected from the ocean at different wavelengths (e.g., MODIS, VIIRS, Sentinel-3 OLCI) and then apply algorithms in order to calculate variables such as Chlorophyll-a concentration. - **Key considerations:** - Ocean color data is significantly affected by cloud cover, aerosols, and atmospheric correction errors. - **Data sources:** - [EU Copernicus Marine Service](https://data.marine.copernicus.eu/products?facets=mainVariables%7EPlankton--sources%7ESatellite+observations) - [US NASA Ocean Color Web](https://oceancolor.gsfc.nasa.gov/data/find-data/) - [NASA Worldview](https://worldview.earthdata.nasa.gov/) - Graphical Interface - **Resources to learn more:** - [IOCCG (International Ocean-Color Coordinating Group)](https://ioccg.org/) ### 2.5 Additional databases and scientific support The four sub-sections above (*In Situ* Sensors; Deep Ocean Observation, Exploration, and Research Systems; Mapping; Satellite Remote Sensing) cover the main areas of ocean scientific data types and collection methods. There are some datasets that are not discussed in this guide since they are likely less useful for investigative storytelling or require technical skills to access and interpret the data. Below are some additional databases and information on contacting scientists to support your investigation. While in section 3, we outline a case study using real data to tell an ocean story. #### 2.5.1 Additional databases, collections, and visualizers Sites that either did not fit into one of the sub-sections above, or that contain information which is generated after scientific studies occur: - [PANGAEA](https://pangaea.de/) - data publisher for earth and environmental science (across disciplines) - [International Seabed Authority DeepData](https://www.isa.org.jm/deepdata-database/) - database hosting all data related to international deep-seabed activities, particularly those collected by contractors (i.e. nations or entities) during their exploration activities and other relevant environmental and resources-related data. Includes a dashboard and map to search for basic stats and information about what contractors have done during deep seabed mining exploration cruises. - [Marine Geoscience Data System](https://www.notion.so/Ocean-Datasets-for-Investigations-1caf92221af780c68873c2aecf9b3479?pvs=21) - geology and geophysical research data across collections - [USGS Earthquake Hazards Program](https://www.usgs.gov/programs/earthquake-hazards/earthquakes) - interactive map with magnitudes and additional information (earthquakes can occur on land and in the ocean) - [WoRMS – World Register of Marine Species](https://www.marinespecies.org/) - comprehensive taxonomic list of marine organism names - [OBIS – Ocean Biodiversity Information System](https://obis.org/) - global open-access data and information on marine biodiversity - [Windy](https://www.windy.com/) - animated weather maps, radar, waves and spot forecasts #### 2.5.2 Scientific support All of the datasets and databases we outlined above are free and open to the public. We hope that we outlined enough context and links to user-friendly platforms to access the data so that you feel empowered to conduct your own investigations with ocean datasets. That said, some data might be more challenging to work with depending on prior experience and computing skills, among other factors. When in doubt, you can always contact an ocean scientist to ask questions or seek support. Depending on your investigation or story, you will need to contact a specific type of ocean scientist since each has their own specialty. You can start by searching for and contacting scientists at nearby universities or research institutes. **Ocean scientists and their specializations:** - **Physical Oceanographers -** Study ocean currents, tides, waves, and ocean-atmosphere interactions. They can help explain phenomena like sea level rise or how ocean circulation affects weather and climate. - **Chemical Oceanographers -** Focus on the chemical composition of seawater and how it changes over time. Useful for stories involving ocean acidification, pollution, nutrient cycling, or chemical runoff impacts. - **Biological Oceanographers or Marine Biologists -** Study marine organisms and their interactions with the ocean environment. They are ideal sources for stories on biodiversity, fisheries, invasive species, and ecosystem health. - **Geological Oceanographers or Marine Geologists -** Study the structure and composition of the ocean floor. They can provide insights into underwater earthquakes, tsunamis, deep-sea mining, or the formation of underwater features. - **Climate Scientists with Ocean Expertise -** Examine how oceans influence and respond to climate change. They are helpful for broader climate stories that involve ocean heat content, carbon storage, or long-term trends in ocean conditions. - **Marine Ecologists -** Study relationships among marine organisms and their environment. They can clarify ecosystem-level impacts, like those from overfishing, coral bleaching, or marine protected areas. - **Fisheries Scientists -** Specialize in fish populations, fishing practices, and resource management. Helpful for reporting on commercial fishing, stock assessments, or policy/regulation issues. - **Ocean Data Scientists -** Work with large marine datasets and modeling, can assist with interpreting satellite data, ocean models, or big datasets. - **Marine Policy Experts and Ocean Economists -** Focus on the intersection of ocean science, law, and economics. Helpful for coverage of marine regulations, governance issues, or the ‘blue economy.’ - **Marine Technologists or Ocean Engineers -** Design and use tools like underwater drones, sensors, and buoys. They can help explain how ocean data is collected and what the limitations of certain technologies might be. ## 3. Case Study: Gulf of Maine Ocean Warming  As ocean scientists from the Northeastern United States, we have each witnessed how rapid ocean changes are affecting the ecosystems and communities around us. For this example case study, we focus on the Gulf of Maine—a region close to home. When telling ocean stories, it is helpful to have either first-hand or personal connections to the coastal or oceanic region you are investigating. ### 3.1 Motivation The Gulf of Maine is warming [faster than 99% of the global ocean](https://eos.org/features/why-is-the-gulf-of-maine-warming-faster-than-99-of-the-ocean), making it a key site to investigate local impacts of climate change on marine environments and coastal livelihoods. Stories of changing fish stocks and stressed fisheries are already discussed in communities within the Northeastern region. Before getting into the data, we will think through the historical and ecological context of the Gulf of Maine and its fisheries. For centuries, the regional identity has been deeply linked to the ocean. Indigenous [Wabanaki peoples](https://www.wabanakialliance.com/wabanaki-history/)—including the Abenaki, Mi'kmaq, Maliseet, Passamaquoddy, and Penobscot nations—relied on these coastal waters for food as well as cultural practices and trade. They managed their coastal and marine environments with ocean knowledge developed across generations. When European colonization began, [intensive cod fishing fueled transatlantic trade and early settlements](https://www.markkurlansky.com/books/cod-a-biography-of-the-fish-that-changed-the-world/). Europeans considered the cod fisheries to be so abundant that they were endless. The overfishing by settlers caused a massive collapse in cod stocks by the 1950s. Now, other important local fisheries like the American lobster are being impacted by the combination of ocean warming and historic overfishing. Harmful algal blooms have also increased in frequency which indicate that the broader Gulf ecosystems are under stress. In the following sections, we guide you through using publicly available ocean datasets to investigate the scientific questions behind the Gulf of Maine and its warming waters. By accessing current and archival datasets you will be able to visually show the seawater temperatures going up and connect that to other environmental stories or investigations about the Gulf. ### 3.2 Data acquisition In order to investigate warming in the Gulf of Maine, we will analyze surface temperatures from two different datasets: a local *in situ* temperature sensor and the global-average SST. With the global SST as our baseline, we’ll be able to determine how much faster the Gulf of Maine is warming compared to the rest of the world. This analysis involves database downloads, data post-processing/analysis, and data visualization. If you don’t have experience with coding and want to get started with the Python programming language, see the appendix for tips on getting setup. Otherwise, you can always consider contacting a scientist to support you with your investigation (see section 2.5.2 Scientific support). #### 3.2.1 Gulf of Maine buoy temperature dataset First, we’ll go to the [National Data Buoy Center (NDBC)](https://www.ndbc.noaa.gov/) website and look for a buoy in the Gulf of Maine with a long historical record of temperature measurements. Clicking on the “Historical Data & Climatic Summaries” link at the bottom of [Station 44007’s page](https://www.ndbc.noaa.gov/station_page.php?station=44007) reveals annual text files going back to 1982.  *Screenshot from the NDBC website showing potential buoys to use for Gulf of Maine case study.* The task now is to process all of this data into a more usable format. We’ll do this with a python script using the [pandas](https://pandas.pydata.org/) data analysis library. 1. Loop through the years 1982-2024 and create the dataset url for each year, using the NDBC website to deduce the url structure. 2. Load the text data directly from each url via `pandas.read_csv()` 3. Convert the year, month, day, hour columns into a single pandas datetime column. 4. Combine all of the data into a single dataframe. 5. Save our data to a new CSV file.  *An example of what the available buoy data looks like for the year 1985. The highlighted sections show the parts of the dataset that we’re interested in: the date/time and the water temperature.* #### 3.2.2 Global mean SST dataset Next, we want a corresponding dataset for the globally-averaged SST, in order to determine whether the Gulf of Maine is warming faster or slower than the average. The [Climate Reanalyzer](https://climatereanalyzer.org/clim/sst_daily/?dm_id=world2) displays globally-averaged SST from the [NOAA 1/4° Daily Optimum Interpolation Sea Surface Temperature (OISST)](https://www.ncei.noaa.gov/products/optimum-interpolation-sst), a long term Climate Data Record that incorporates observations from different platforms (satellites, ships, buoys and Argo floats) into a regular global grid. 1/4° refers to the grid resolution—about 25 km. There is an option to download the underlying data from the Climate Reanalyzer website, which will save us a lot of time vs. trying to access decades of data and doing the global averaging ourselves. The data is available as a JSON file, which is a different text file format that will require a more custom approach for converting into a pandas dataframe.  *Screenshot of the [Climate Reanalyzer](https://climatereanalyzer.org/clim/sst_daily/?dm_id=world2) website. In the dropdown menu, we want to download the JSON data.* One key concept to note is how this data handles dates. Each year includes a list of 366 temperatures, without any explicit list of the corresponding dates. This is using the format of “day of year” and we see that the last temperature is “null” for non-leap years. When processing this data, we need to take this into account and ignore the null final value. Similar to the buoy data, we’ll re-format this data in a pandas dataframe and save to a new CSV file.  *A look inside the globally-averaged SST JSON file. The data is arranged as a list of years where each year has a list of 366 temperatures.* ### 3.3 Climatological data analysis A standard method for analyzing climate change anomalies is to first remove the climatological “seasonal” signal from the data. This will allow us to show, for each data point, how much warmer or colder it was than the average temperature for that day of the year. The first step is choosing which time period we’ll use for our climatology “baseline”. Here we’ve chosen 1991 to 2020 since it is fully covered by our data and matches the climatology period used by the Climate Reanalyzer website. Next, we’ll use some built-in pandas methods to get the climatological average temperature for each day and then map that to each datapoint in our timeseries. The following code snippet shows the steps used for both the buoy data and the global SST: ```python # Select just the data in the range of the climatology period df_clim = df[(df.index.year >= 1991) & (df.index.year <= 2020)].copy() # Assign the day of year (1-366) to each data point in the timeseris df_clim["day_of_year"] = df_clim.index.dayofyear # Take the mean for each day_of_year df_clim = df_clim.groupby("day_of_year")["temp"].mean() # New variable in df: the climatological temperature for that day df["day_of_year"] = df.index.dayofyear df["climatology_value"] = df["day_of_year"].map(df_clim) # Temperature anomaly is observed temperature minus climatological temperature df["anomaly"] = df["temp"] - df["climatology_value"] ```  *Our resulting dataframe includes new columns for climatology and temperature anomaly.* ### 3.4 Analyzing and Visualizing the results First, we’ll simply plot the full temperature timeseries and see what we find.  The warming signal is instantly apparent in the global SST data because the seasonal signal is so small. The Gulf of Maine, however, varies by more than 15° C throughout the year so any long term changes are difficult to see in this format. Plotting the climatology signal illustrates this point (pay attention to the y-axis).  Next we’ll view our temperature anomaly data (observed temperature minus climatology). As expected, there is more noise in the buoy data since it’s taken from a single point and any given day can vary by as much as 4 °C from climatology. The globally-averaged temperature has much less variance.  For the final version of our plot, we’re incorporate 3 changes: 1. Fit a simple linear regression using [numpy’s polyfit](https://numpy.org/doc/stable/reference/generated/numpy.polyfit.html) in order to quantify the average rate of warming for the two datasets. 2. Plot the monthly averages instead of the daily values in order to simplify the visual clutter. 3. Use the same y-axis range for the two plots for direct visual comparison.  Comparing our warming rate calculations against the published literature finds good agreement: - Gulf of Maine SST: our rate of 0.496°C/decade is within 5% of the 0.47°C/decade reported by the [Gulf of Maine Research Institute](https://gmri.org/stories/2024-gulf-of-maine-warming-update/). This is likely due to differences in methods—we used a single buoy and they used the OISST data averaged across the entire Gulf. - For global SST, our rate of 0.188 °C/decade is within 5% of the 0.18 °C/decade (over the past 50 years) published by [Samset et al. (2023)](https://www.nature.com/articles/s43247-023-01061-4). These final plots provide simple visual evidence of the Gulf of Maine’s rapid warming over the past 40 years. We showed the data transform from text files, to noisy timeseries, and finally to expert-validated trend lines. By removing the strong seasonal signal and focusing on the anomalies, we can clearly see the long-term warming trend in both the Gulf of Maine buoy data and the global mean SST. Finally, note that the linear regression is useful for quantifying the recent warming in an easily understandable number but is not necessarily a predictor of future warming. The [Maine Climate Science Dashboard](https://climatecouncil.maine.gov/future/climate/science-dashboard) shows the potential for human emissions to either accelerate or slow down this rapid warming.  *The [Maine Climate Science Dashboard](https://climatecouncil.maine.gov/future/climate/science-dashboard) combines historical water temperature measurements with different climate scenario forecasts.* ## 4. Conclusion Our investigation into Gulf of Maine temperatures, using readily available public datasets, highlights one local manifestation of global climate change. This rapid warming isn't merely an abstract data point, it continues to have profound implications for the region’s biodiversity and the human communities who rely on the ocean. Marine species are highly sensitive to temperature changes, and the Gulf of Maine has been experiencing a noteworthy decline in native species and [increase in warmer-water species](https://online.ucpress.edu/elementa/article/9/1/00076/118284/Climate-impacts-on-the-Gulf-of-Maine-ecosystemA). The next steps in this story might look to other data sources to explore: Why is the Gulf of Maine warming so quickly? and What will the region look like in the future? or How exactly are local fisheries affected by warming waters? This case study is one example of how to find the connections between global environmental change, local ocean data, and tangible human impacts. This process offers a template for investigating similar stories in your own regions: 1. **Start with a local observation or community concern:** What are people witnessing or experiencing in your local environment? 2. **Explore the scientific context:** Consult with scientists, read relevant research, and understand the underlying environmental drivers. 3. **Seek out publicly available data:** As shown in section 2, there is a large assortment of high-quality public ocean datasets that can be used to investigate countless questions. 4. **Connect the data back to human issues:** How do the environmental changes revealed by the data affect local cultures, livelihoods, health, and economies? The key thing to remember is that there are multiple angles to uncover and expose often-invisible impacts to the ocean. Datasets provide one lens to report on climate and environmental changes, but these stories impact communities and are thus both political and social. Just as ocean science has changed and begun to decolonize, it's crucial to investigate and tell stories that reflect diverse experiences. Ocean data can help highlight intersecting issues—such as deep seabed mining, marine health, and colonial continuums—with evidence-based information and compelling visualizations. We hope this guide offers a practical starting point for navigating ocean science, accessing and interpreting data, and connecting your investigation to real-world consequences that are planetary in scale yet intimately local. ```cik-note ``` >**APPENDIX: Getting started with python** > >If you have not done any coding before, the initial task of setting up your coding environment can be a challenging hurdle. There are multiple options for code editors/IDEs (integrating development environment), ways of handling dependencies (the external packages you install to give you advanced functionality), and other decisions that are outside the scope of this article. Luckily, once you’ve chosen your tools, there are good resources online so here are a few recommendations and then you can seek out more detailed tutorials: > >1. Use Visual Studio Code as your code editor (the application where you will write and run code). This is the most popular option and there is an extensive ecosystem of 3rd party plug-ins and help resources. <https://code.visualstudio.com/> > >2. Use [conda](https://docs.conda.io/projects/conda/en/latest/user-guide/getting-started.html) for package management. Your computer’s operating system may come with a version of python pre-installed but it's not a good idea to install packages onto this global location. Instead, create separate conda "environments" for different projects. This will allow you to experiment in a safe and organized way. Here is a helpful article on the VS Code website: <https://code.visualstudio.com/docs/python/environments>. For example, to create a new environment that we'll name "ocean-study" and install the "matplotlib" plotting package would look like this: > > ```bash > conda create -n ocean-study > conda activate ocean-study > conda install matplotlib > ``` > Now, in VS Code, just make sure your Python Interpreter is using this environment (it will look something like `~/miniconda3/envs/ocean-study/bin/python` and you will be able to use the matplotlib package in your code. > >3. Finally, consider using Jupyter Notebooks for exploratory coding where you’re loading datasets and making plots. Notebooks have the file extension `.ipynb` and allow you to run chunks of code independently in code "cells" and view the output right below. You can also use Markdown text cells to write notes and explanations for yourself and collaborators. Instructions on using VS Code: <https://code.visualstudio.com/docs/datascience/jupyter-notebooks> > <hr class="thick"> #### About the authors **Mae Lubetkin** is an ocean scientist, transmedia artist, and writer based in Paris and at sea. Their practice-led research remaps our relations to bodies of water and digital worlds by means of investigation, counter-narrative, and memory. With a background in marine geology and subsea imaging, their artistic practice is in dialogue with Science while situated in queer, intersectional, anti-extractivist, and decolonial frameworks. Guided by wet-techno-critical studies and thinking with other-than-human worlds, they compose environmental traces in installations and digital outputs. Their core practice is in solidarity with submerged, ancient, ephemeral and imaginary environments. **Dr. Kevin Rosa** is an oceanographer and the founder of Current Lab, a startup specializing in computational ocean forecasting. He holds a B.A. in Physics and a PhD in Physical Oceanography from the University of Rhode Island, with a focus on ocean physics and hydrodynamic modeling. <hr class="thick"> *Published in June 2025*
Ocean Datasets for Investigations
=============================== By Mae Lubetkin and Kevin Rosa  ```cik-in-short ``` **In short:** Learn how to identify and use Ocean datasets as a tool for revealing the unseen or underreported dynamics of the world's most significant bodies of water. --- ## 1. Introduction: Ocean Science, Data, Storytelling Given our current planetary condition, many of the world's most pressing stories are linked to the ocean. Covering 71% of Earth's surface and interacting constantly with the atmosphere, the ocean is our greatest carbon sink and an essential indicator of climate change. Despite its critical role in maintaining a habitable planet and supporting coastal livelihoods, the ocean is often invisible to the lived experience of most individuals, particularly those far from its shores. There are so many ways to tell stories about the ocean, and countless diverse perspectives from which to understand it. Now, more than ever, we need to incorporate cross-cultural and trans-disciplinary strategies for investigative projects, particularly those concerning the ocean. This guide presents ocean datasets as a tool for revealing the unseen or underreported dynamics of the world's most significant bodies of water. For informed investigations and impactful storytelling, oceanographic datasets can be an essential resource for journalists, activists, and anyone interested in data-driven methods to communicate the climate crisis, environmental change, natural disasters, extractivism, and associated ocean justice issues. From bathymetric maps, subsea imagery, and 3-D habitat models, to satellite-derived and *in situ* real-time monitoring data – a vast amount of oceanographic media and data is publicly available. In this Exposing the Invisible Guide, we begin with an introduction on the broader scientific history and context within which ocean data is collected, stored, and made accessible. Section two outlines the diversity of datasets, including some history, trade-offs, use cases, data collection methods, data sources, and resources to learn more about each data type that we present. Section three offers a specific application of ocean data in a case study, including: steps explaining why the data is useful for supporting this particular story; how to source and present the data; and, finally, how to bring it into a meaningful investigative report, journalistic piece, or other storytelling format. The guide concludes with a summarized approach for using ocean datasets in investigations and outlines strategies for identifying the right ocean scientists who could support you and your investigation.  *Boulder resting on the seafloor offshore the Revillagigedo Archipelago in the Pacific Ocean. Credit: Ocean Exploration Trust.* ### 1.1 Ocean Science: History and Context Ocean science generally refers to the observation and investigation of biological, geological, physical, and chemical processes that shape and constitute global marine environments. This broad disciplinary field includes numerous sub-disciplines that focus on intricate, detailed studies of specific scientific questions concerning the ocean. Ocean scientists monitor habitat change, measure biodiversity, investigate geological phenomena, and study human impacts on ocean systems (i.e., global warming, pollution, overfishing, and extractive projects). Interactions between marine ecosystems and human activities, as well as atmospheric and coastal processes, are all carefully investigated by ocean scientists today. Despite niche specializations, there are increasingly multidisciplinary projects that involve diverse experts in order to more comprehensively understand the interconnections between oceanic processes and phenomena. Collectively, this research improves our baseline knowledge of the ocean, which can then support preservation strategies while maintaining sustainable and regenerative relationships with diverse marine ecosystems. Although ‘contemporary’ ocean science has deep roots in European colonialism and imperial exploration, ocean knowledge systems long predate Western scientific inquiry. Indigenous and coastal communities across Oceania, as well as the Atlantic and Indian Oceans, carefully studied and navigated the seas for thousands of years. These forms of ocean science are less known or dominant on a global scale, but they nevertheless involve highly sophisticated techniques for understanding the stars, ocean swells, winds, and currents. Navigators across Oceania used interconnected and embodied forms of ocean science, on their own terms, to journey vast distances across the seas with tremendous precision. While other coastal Indigenous peoples around the world developed specific place-based systems of knowledge, including both land and marine management practices that viewed these spaces as highly linked. Most of these communities understood the ocean not as a monstrous or alien-filled void (as European explorers often depicted it), but as a vast world interconnected with our own.  *Rebbilib (navigational chart) by a Marshall Islands artist in the 19th to early 20th century. These stick charts were used by Marshall Islander navigators during long ocean voyages. Credit: Gift of the Estate of Kay Sage Tanguy, 1963. Source: <https://www.metmuseum.org/art/collection/search/311297>* When European seafarers began worldwide exploration voyages in the 15th-16th centuries, they dismissed or actively erased these Indigenous ocean knowledge systems. European or 'Western' scientific models were linked to a colonial mindset that often viewed the natural world as a space to examine in order to master and own its elements, rendering them as 'resources'. Notions of relationality were strongly opposed to the point that scientists considered their surroundings as objects to be studied rather than elements to relate to or work with. At the core, these opposing ocean scientific methods or knowledge systems reflected the specific values and worldviews of each culture, respectively. The Challenger Expedition (1872–1876) was the first European-led systematic scientific exploration of the global oceans. In some ways, it was groundbreaking. However, it also played a key role in the broader colonial project, which aimed to map, control, and extract 'resources' from around the world. Today, Western or 'contemporary' ocean science continues to use investigatory methods that stem from European knowledge systems. Oceanographic research often occurs in the context of economic or territorial expansion and military-supported science projects. Nevertheless, these methods are beginning to open up to other forms of knowledge creation that move beyond the geopolitical interests of wealthy nations.  *Map of ocean currents created by John Nelson using the "WGS 1984 Spilhaus Ocean Map in Square" projected coordinate system in ArcGIS. The Spilhaus Projection (developed by oceanographer Athelstan Spilhaus in 1942) reflects aspects of decolonial cartography by shifting focus from land-centered perspectives to an oceanic worldview. Source <https://storymaps.arcgis.com/stories/756bcae18d304a1eac140f19f4d5cb3d>* Thanks to the enduring activism and decolonizing work led by many Indigenous and coastal communities, there is increasing recognition of the need to reclaim ocean science by amplifying the knowledge and perspectives of those who have long understood the seas. Ocean science is just beginning this deep process of decolonization, which first seeks to acknowledge and reckon with the violent and ongoing impacts of colonization. Decolonizing ocean science requires a fundamental shift in who holds agency and sovereignty over their own ocean waters. This also relates to international waters and who is included, excluded, or undervalued throughout negotiations concerning the legal and regulatory structures governing global oceans. Today, many ocean scientific projects are co-designed and co-led by ocean knowledge holders from diverse backgrounds. Collecting ocean datasets requires a team of experts who follow cultural protocols, ensure environmental safety, and apply diverse scientific methods, all while striving for more relational practices. ### 1.2 Ocean Data Collection Today Today, ocean datasets are collected by ocean scientists in collaboration with ocean engineers. These datasets are gathered from several sources to understand the global ocean and its role in maintaining Earth's habitability and critical planetary cycles. Ocean engineers develop the tools, platforms, and instruments that are required for data collection, such as underwater vehicles, satellite-mounted sensors, and buoys. By designing technologies that can operate in diverse and sometimes extreme conditions, these engineers support and expand ocean scientific capabilities. Together, ocean scientists and engineers advance our understanding of the planet for both research and conservation. There is a considerable variety of ocean data types, tools for data collection, and associated databases to store these recorded entities. This diversity of datasets is outlined in section 2. Like most scientific fields, funding can be secured from public governmental bodies or private sources. The ocean datasets we focus on here are publicly accessible and typically funded by governments via taxpayer contributions. This means that ocean datasets are for the people and should be accessible. Unfortunately, many public ocean datasets are stocked in complex databases and require specialized software, programming experience, or extensive knowledge to access. That said, there are plenty of datasets that can be retrieved and visualized more easily, with little to no background knowledge. There are also some ocean datasets that can be accessed with helpful tips and instructions, which is what we will focus on here. The Exposing the Invisible Toolkit is designed as a self-learning resource, we hope that this guide will support future investigations and make ocean datasets more accessible to communities and investigators around the world. ### 1.3 Data Gaps, Capacity Initiatives, Ocean Defenders Conducting ocean science can be a costly endeavor. Depending on the environment, scientific goals, and technical requirements, some ocean scientific work can only be conducted by wealthy nations or private organizations. Typically, this kind of science takes place at sea or uses remote sensing techniques. For example, deep ocean exploration and research requires a ship, vehicles, or platforms to deploy down to the deep ocean, technical and computing systems aboard to process the data, and a diverse team of experts to manage these operations. In contrast, satellite remote sensing used for ocean research typically covers the entire Earth surface. Publicly funded satellite-derived ocean datasets can be useful across territorial waters, throughout international seas, and are accessible regardless of one's nationality. Near-shore ocean science and *in situ* coastal monitoring efforts are more financially affordable, especially as diverse knowledge systems merge with lower-cost technologies and capacity initiatives. In this context, capacity refers to the skills, resources, and knowledge needed to effectively conduct ocean science. As ocean science undergoes a process of decolonization, this emphasis on capacity building, development, and sharing is also strengthening. Additionally, several international ocean law and policy frameworks specifically aim to increase ocean science capacity. Ocean defenders are also central to these efforts. As groups, individuals, or organizations dedicated to protecting marine environments, defenders play a key role in advocating for capacity building within and outside scientific structures. Many defenders are fisherpeople, coastal communities, or groups directly affected by changes in climate and ocean health. Beyond advocating for sustainable and generative oceanic futures, they also fight to overcome political resistance and funding barriers. Ocean defenders, like land defenders, face challenging or dangerous obstacles while pushing for local and global ocean preservation. Ocean science and policy clearly needs collaborative approaches that bring multiple knowledge systems forward while prioritizing those most impacted by climate change, pollution, and other threats to both marine habitats and coastal livelihoods.  *Ocean-defending artisanal fishers and their supporters in South Africa celebrate upon receiving the news that Shell’s permit to conduct a seismic survey on the Wild Coast had been set aside by the Makhanda High Court, in September 2022. Photo credit: Taryn Pereira. Source: <https://oceandefendersproject.org/case-study/no-to-seismic-surveys/>* **More on capacity initiatives, knowledge gaps, and ocean defenders:** - Guilhon, M., M. Vierros, H. Harden-Davies, D. Amon, S. Cambronero-Solano, C. Gaebel, K. Hassanali, V. Lopes, A. McCarthy, A. Polejack, G. Sant, J.S. Veiga, A. Sekinairai, and S. Talma. (2025). [Measuring the success of ocean capacity initiatives.](https://doi.org/10.5670/oceanog.2025.122) *Oceanography* 38(1). - Saba, A.O., I.O. Elegbede, J.K. Ansong, V.O. Eyo, P.E. Akpan, T.O. Sogbanmu, M.F. Akinwunmi, N. Merolyne, A.H. Mohamed, O.A. Nubi, and A.O. Lawal-Are. 2025. [Building ocean science capacity in Africa: Impacts and challenges.](https://doi.org/10.5670/oceanog.2025.133) *Oceanography* 38(1). - Behl, M., Cooper, S., Garza, C., Kolesar, S. E., Legg, S., Lewis, J. C., White, L., & Jones, B. (2021). [Changing the culture of coastal, ocean, and marine sciences: strategies for individual and collective actions.](https://www.jstor.org/stable/27051390) *Oceanography*, *34*(3), 53–60. - The Ocean Defenders Project (2025). [Ocean Defenders: Protectors of our ocean environment and human rights.](https://oceandefendersproject.org/project-publications/ocean-defenders-protectors-of-our-ocean-environment-and-human-rights/) The Peopled Seas Initiative, Vancouver, Canada. - Belhabib, D. (2021) [Ocean science and advocacy work better when decolonized.](https://doi.org/10.1038/s41559-021-01477-1) *Nat Ecol Evol* 5, 709–710. - Kennedy, R. and Rotjan, R. (2023). [Mind the gap: comparing exploration effort with global biodiversity patterns and climate projects to determine ocean areas with greatest exploration needs.](https://doi.org/10.3389/fmars.2023.1219799) *Front. Mar. Sci.* (10). - Bell, K.L.C, Johannes, K.N., Kennedy, B.R.C., & Poulton, S.E. (2025) [How little we’ve seen: A visual coverage estimate of the deep seafloor.](https://www.science.org/doi/10.1126/sciadv.adp8602) *Science Advances*, Vol 11, Issue 19. ### 1.4 Ocean Datasets for Investigations and Storytelling Ocean datasets play a crucial role in both scientific investigations and storytelling by providing evidence-based insights into the health and dynamics of our global ocean. These datasets help scientists better understand the ocean, but beyond research, they serve an important role in ocean-related investigations. Ocean datasets can support communities, journalists, and activists in raising data-backed awareness about critical marine and environmental justice issues. By sharing this data in accessible ways, oceanographic narratives can amplify the voices of coastal communities, engage the public, and inspire action in support of more regenerative ocean futures. Numerous well-resourced journalistic and forensic organizations use ocean data to support their stories or reporting, such as Forensic Architecture, LIMINAL, Forensis, Earshot, Border Forensics, and others. In this guide, we will demonstrate how you can access datasets and conduct your own oceanic investigations. By the end, you will be able to illustrate a well-defined ocean or climate question using publicly available oceanographic datasets and media collections, which will enhance your evidence-based and visually engaging story. ## 2. Diversity of Data Types The following sub-sections serve as a collection of key ocean data types, organized by how they are collected and what they reveal. Each broad data source type (i.e., the overarching technique used to gather certain kinds of ocean data) begins with a bit of history and trade-offs, and then is further broken down into specific data products. For each data product, we share use cases, data collection methods, key considerations, and some databases (often from U.S. and European agencies), followed by resources to learn more. This structure is designed to help you understand how each dataset is produced, grasp the significance of the data, and know where to go for deeper investigations or analyses. There are many data types presented below which are organized in these four broad categories: *in situ* sensors; deep ocean observation, exploration, and research; mapping; and, satellite remote sensing. See Section 3 to review an example case study which demonstrates how some of these datasets may be used to support an investigative ocean and climate story.  *Illustration of a range of ocean sensor platforms—ships, profiling drifters, gliders, moored buoys, and satellites. Source: <https://www.ecmwf.int/en/about/media-centre/news/2021/world-meteorological-day-focuses-role-ocean-weather-and-climate>* ### 2.1 *In Situ* Sensors *In situ* sensing refers to the direct measurement of ocean properties using instruments that are physically located within the water. Sensors for measuring temperature, salinity, pressure, currents, and of biochemical concentrations are deployed on a range of platforms with various advantages and drawbacks. While satellites provide broad spatial coverage of the ocean’s surface, *in situ* platforms are essential for monitoring the ocean’s interior, tracking coastal change, and measuring water properties that cannot be detected remotely. - **History:** - Sailors have used thermometers to measure ocean temperature since at least as early as Captain James Cook’s [1772 voyage](https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/rog.20022) to the Antarctic Circle (another example of colonial science forming the foreground to Western ocean science). - The Nansen bottle (1896) and later the Niskin bottle (1966) enabled the capture of water samples at specific depths, which could then be pulled up and tested for temperature and salinity on the ship. - The first bathythermograph was developed in 1938 and featured a temperature sensor on a wire which recorded a temperature profile as it was lowered into the ocean. This was used by the US Navy in WWII to improve sonar accuracy, since temperature layers alter acoustic propagation. - Today, there are thousands of advanced sensors across the world’s oceans which transmit readings in real time via satellite. - **Trade-offs:** - *In situ* sensors can only measure the ocean properties at their exact location so great consideration is taken in their placement and timing. - Powering the instruments is a constant challenge and factors into decisions about sampling frequency. - The extreme pressure in the deep ocean limits the operating depth of some sensors. - Harsh operating conditions limit the lifespan of sensors and necessitates regular maintenance/replacement, often in remote locations at high costs. This leads to less accessible areas being undersampled. #### 2.1.1 Moorings and Fixed Platforms  *Example of a coastal data mooring. Source: Baily et al., 2019, <https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2019.00180/full>* - **About:** - A range of platforms anchored in place, collecting time-series data at a fixed location. - Used in the deep ocean (e.g., the TAO array across the Pacific Ocean), the continental shelf (e.g., NOAA’s National Data Buoy Network), and at the coastline (e.g., tide gauges). - **Use cases:** - Long-term climate monitoring of ocean heat content and circulation. - Data inputs for forecast models, improves accuracy of ocean and weather predictions. - Early warning systems for tsunamis and hurricane storm surge. Tracking tidal heights and local sea level rise. - Water quality monitoring for pollutants, algal blooms, and hypoxia. - **Data collection:** - Sensor packages measure temperature, salinity, pressure, biochemistry, and more. - Some moorings have Acoustic Doppler Current Profilers (ADCPs) to measure water current velocities throughout the water column. - **Key considerations:** - Data today is mostly broadcasted in near-real time, but there are some platforms that require physical retrieval before the data is downloaded. - Spatial coverage is extremely limited and focused around a small set of nations. - The diversity of databases and data types can pose a challenge for accessing and working with the data. - **Data sources:** - [US NOAA National Data Buoy Center (NDBC)](https://www.ndbc.noaa.gov/) - [EU Copernicus Marine Service In Situ dashboard](https://marineinsitu.eu/dashboard/) - [Global Tropical Moored Buoy Array](https://www.pmel.noaa.gov/gtmba/) - **Resources to learn more:** - [WHOI - Moorings & Buoys](https://www.whoi.edu/what-we-do/explore/instruments/instruments-moorings-buoys/) - [Ocean Observatories Initiative](https://oceanobservatories.org/) #### 2.1.2 Drifters and Floats  *Map of Argo float locations. Source: <https://argo.ucsd.edu/about/status/>* - **About:** - Unanchored and unpropelled instruments that drift with the currents and take ocean measurements. - Drifters stay at the surface and provide information about surface conditions, and surface currents are calculated from their GPS trajectory. - Floats profile the water column by adjusting their buoyancy to move up and down. The Argo program is the largest and most significant, with over 4,000 Argo floats profiling the world’s oceans. - Capable of providing global coverage at lower cost than moored sensors, especially in remote open-ocean regions. - **Use cases:** - Drifters: Mapping near-surface currents and tracking pollutants and marine debris transport. - Floats: Measuring subsurface temperature and salinity for climate studies. Some Argo floats also have biochemical sensors. - **Data collection:** - Drifters: GPS-tracked, measure SST, pressure, sometimes salinity, sometimes waves. - Argo floats: Profile down to 2,000 m every 10 days, transmitting data via satellite. - **Key considerations:** - Drifters and floats are always moving, so you can’t get a clean timeseries for a single location like you can with moorings. Additionally, Argo floats only take one profile every 10 days in order to preserve battery life. - Argo floats don’t generally operate near the coast on the continental shelf. - Some drifters/floats lack real-time telemetry (data transmission). - **Data sources:** - [Global Drifter Program](https://www.aoml.noaa.gov/phod/gdp/) - [Argo Program](https://argo.ucsd.edu/) - [Copernicus Marine Service drifters](https://data.marine.copernicus.eu/products?facets=featureTypes%7ETrajectory) - [SOCCOM Biogeochemical Floats](https://soccom.org/) - **Resources to learn more:** - [About Argo](https://argo.ucsd.edu/about/) #### 2.1.4 Autonomous Vehicles - ASVs and Gliders  *Illustration of a glider’s sawtooth propulsion pattern. Source: <https://earthzine.org/going-deep-to-go-far-how-dive-depth-impacts-seaglider-range/>* - **About:** - Autonomous Surface Vehicles (ASVs) and gliders are robotic platforms that enable long-duration, energy-efficient monitoring over vast areas. - Bridging the gap between targeted, expensive, ship-based measurements and low-cost, but uncontrolled, drifters/floats. - **Use cases:** - Significant overlap with the use cases for moored sensors and drifters/floats. - Targeted measurements in dangerous conditions like hurricanes. - Mapping surveys, autonomous of ship or in tandem with other vehicles. - **Data collection:** - Autonomous Surface Vehicles (ASVs) use solar panels, wind, or waves as a power source to supplement and recharge their batteries. - Gliders are underwater vehicles that create propulsion by adjusting their buoyancy and gliding horizontally while sinking/rising (similar to an airplane). This enables longer battery range than propellers or thrusters. - **Key considerations:** - Gliders and ASVs are often used in targeted studies rather than continuous global monitoring and thus have lower data availability. - Shorter mission durations than moorings or drifters/floats. - **Data sources:** - [NOAA Glider Data Assembly Center](https://gliders.ioos.us/) - [OceanGliders](https://www.oceangliders.org/) - **Resources to learn more:** - [National Oceanography Centre UK - Gliders](https://noc.ac.uk/facilities/marine-autonomous-robotic-systems/gliders) ### 2.2 Deep Ocean Observation, Exploration, and Research Deep ocean science is typically conducted to observe long-term changes at specific seafloor sites, to explore marine habitats that are unknown to science, and to conduct applied or experimental research on focused environmental questions. A range of deep ocean data collection tools include platforms or landers, cabled observatories, and deep submergence systems—such as human-occupied vehicles (HOVs), remotely-occupied vehicles (ROVs), and autonomous underwater vehicles (AUVs).  *Human-occupied vehicle (HOV) 'Alvin' being recovered in 2024 during an expedition to the East Pacific Rise hydrothermal vent fields. Photo credit: Mae Lubetkin* - **History** - 1872–1876: The *HMS Challenger* expedition, a milestone for deep ocean science but deeply tied to imperialism - Mid-20th century: Cold War military priorities further developed submersible vehicle capabilities, leading to *Trieste*’s 1960 Mariana Trench dive - 1964 onward: HOVs expanded access to the deep ocean, e.g., *Alvin* (US), *Nautile* (France), *Shinkai* (Japan), and *Mir* (Russia) - 1980s–2000s: ROVs and AUVs developed by industry (oil, mining, and defense) and scientific institutions, in parallel - 2000s–present: Cabled observatories (e.g., Ocean Networks Canada, DONET in Japan), public research campaigns (e.g., NOAA, IFREMER), and oceanographic instruments expanded reach and scope. - Today: Many regions still face barriers to participation and funding in deep ocean science (as outlined in the introduction). Meanwhile, deep submergence science in wealthy nations increasingly utilizes AI, autonomous systems, 4K and 3-D imaging techniques. - **Trade-offs:** - Provides direct access to deep ocean environments which are inaccessible by surface vessels or remote sensing - High spatial and contextual resolution: can capture detailed imagery, samples, and detailed *in situ* measurements - Resource-intensive: operations usually require ships, launch/recovery teams, and specialized personnel - Limited temporal and spatial coverage: data collection is episodic, site-specific, and dependent on expedition funding, schedules, and weather conditions at-sea - High costs and technical barriers mean deep ocean science is dominated by a few well-funded institutions or nations, with limited global access - Colonial legacies persist in relation to who sets research agendas, who makes funding decisions, and who benefits from collected data #### 2.2.1 Deep Submergence Systems (HOVs, ROVs, AUVs) - **About:** - Vehicles that operate in the deep ocean water column or along the seafloor, including: - Human-occupied vehicles (HOVs): carry scientists directly, typically 1-3 observers and a pilot - Remotely operated vehicles (ROVs): tethered and piloted from a surface vessel like a ship - Autonomous underwater vehicles (AUVs): untethered and pre-programmed - These systems can operate from hours to days and are depth-rated around 4000-6000 m, but some may reach full ocean depths (11 km) and others may work well in shallower waters. - **Use cases:** - High-resolution visual surveys - Precise targeted sampling with environmental context and imagery at diverse environments including hydrothermal vents, methane seeps, cold-water coral habitats, and others - Biogeographic habitat mapping - Wreck exploration and infrastructure inspection - Imagery of deep ocean environments can support visual storytelling, public engagement, and education - **Data collection:** - Data is streamed directly to the support vessel for tethered operations - While for untethered submersibles (HOVs and AUVs) most data is retrieved when the vehicle is recovered - All physical samples are retrieved and processed upon vehicle recovery - **Key considerations:** - Requires experienced pilots and operational support (expensive) - AUVs need detailed mission planning (mission failure could lead to vehicle loss) - Navigation and environmental risks must be managed carefully - **Data sources:** - SeaDataNet - [EU research vessel data](https://csr.seadatanet.org/), including cruise summary reports and more - EuroFleets - European initiative to compile [EU research cruise data](https://www.eurofleets.eu/data/) - Rolling Deck to Repository - [US research vessel data](https://www.rvdata.us/data), including: expedition summary, shiptrack navigation, scientific sampling event log, post-processed data - JAMSTEC Databases - [Japan research vessel data](https://www.jamstec.go.jp/e/database/), including: HOV *Shinkai 6500* and ROV *Kaiko* mission data, cruise reports, and dive logs - **Resources to learn more:** - [Woods Hole Oceanographic Institution - National Deep Submergence Facility](https://ndsf.whoi.edu/) - [Ocean Exploration Trust - Science and Technology](https://nautiluslive.org/science-tech)  *An imaging elevator equipped with two camera systems, lights, battery packs, and crates to store additional sampling tools to be used by an HOV during a dive in the same region. Source: Woods Hole Oceanographic Institution* #### 2.2.2 Landers and Elevators - **About:** - Landers are relatively simple systems that descend to the seafloor and remain stationary for the duration of their deployment. - They are sometimes referred to as 'elevators' since they descend to the seafloor then ascend back to the surface - There are no people on landers, but they typically carry sensors, cameras, samplers, and other instruments - Depending on power supply and scientific goals, they can spend hours to sometimes months on the seafloor - **Use cases:** - Collecting environmental data (e.g., conductivity, temperature, pH, oxygen) - Capturing imagery of habitats or operations - Deploying baited cameras or traps to study biodiversity - Using the platform to carry additional gear or instruments to the seafloor that a deep submergence system could not transport on its own due to space limitations - **Data collection:** - Typically data is retrieved when the lander is recovered back on deck - Some landers will transmit data acoustically or remotely from the seafloor - The frequency that imagery or other data are collected is pre-programmed before deployment - **Key considerations:** - Requires careful site selection, recovery planning, and often ship time - Currents can impact the intended landing location on the seafloor, sometimes drifting the platform or lander far off-site - Limited in capabilities, not as advanced as deep submergence vehicles, but also much cheaper and easier to custom build - **Data sources:** - Deep ocean lander data (e.g., imagery and environmental sensor data) would be found in the same databases and repositories listed in section 2.2.1 Deep Submergence Systems (HOVs, ROVs, AUVs). - **Resources to learn more:** - [Schmidt Ocean Institute - Elevators and Landers](https://schmidtocean.org/technology/elevators-landers/) - [The Deep Autonomous Profiler (DAP), a Platform for Hadal Profiling and Water Sample Collection (Muir et al., 2021)](https://journals.ametsoc.org/view/journals/atot/38/10/JTECH-D-20-0139.1.xml) - [Lander Lab: Technologies, Strategies and Use of Ocean Landers (Hardy, 2022)](https://magazines.marinelink.com/Magazines/MarineTechnology/202201/content/technologies-strategies-landers-594271)  *Map of Ocean Networks Canada NEPTUNE and VENUS Observatories near Vancouver Island, Canada. Each orange square represents a node or station along the cabled observatory where instruments or sensors are mounted. Source: <https://www.oceannetworks.ca/>* #### 2.2.3 Cabled Observatories - **About:** - Mostly permanent, wired infrastructure on the seafloor that transmit real-time power and data via fiber optic cables connected to shore stations - Similar in some ways to the data collection tools described in section 2.1 *In Situ* Sensors, but these networks are fixed in location and networked - People do not visit these observatories, instead they support a wide range of sensors (e.g., temperature, pressure, seismometers, hydrophones, cameras, samplers) - Can integrate with ROVs or AUV docking stations, and are also typically maintained and serviced by ROVs - Designed for continuous, high-frequency monitoring of deep ocean processes across years or decades - They connect highly diverse environments from hydrothermal vent regions to abyssal plains and continental shelves - **Use cases:** - Long-term and consistent monitoring of geophysical activity (e.g., earthquakes, hydrothermal vents) - Real-time data for early warning systems (e.g., tsunamis, gas releases) - To study oceanographic processes (e.g., currents, biogeochemical fluxes, and ecosystem change) - Supports public engagement and education through livestreams - **Data collection:** - Real-time data is livestreamed to shore stations and then available via online portals - Most are operated by national or international research infrastructures - **Key considerations:** - Extremely costly, high maintenance needs (ROVs are often used for annual servicing) - Site selection is key since they are fixed installations - **Data sources:** - Ocean Networks Canada - [Oceans 3.0 Data Portal](https://data.oceannetworks.ca/), including all datasets, dashboards, and visualizers (more info on [ONC data](https://www.oceannetworks.ca/data/)) - US Ocean Observatories Initiative - [OOI Data Portal](https://oceanobservatories.org/data-portal/), includes cable-linked arrays on East and West Coasts and deep Pacific - EU [EMSO ERIC Data Portal](https://data.emso.eu/home), real-time and archived data, tools and research environment to investigate seafloor observatories across European margins - **Resources to learn more:** - [Ocean Observatories Initiative](https://oceanobservatories.org/observatories/) - arrays, infrastructure, instruments - [Ocean Networks Canada](https://www.oceannetworks.ca/observatories/) - observatories - [Interactive map of ONC locations](https://www.arcgis.com/home/webmap/viewer.html?webmap=fcea4e5f087f41c58bcc5e51b13fffa1&extent=-158.3094,39.6681,-29.8133,75.4182) - [Regional Cabled Observatories](https://www.whoi.edu/what-we-do/explore/ocean-observatories/about-ocean-observatories/types-of-observatories/regional-cabled-observatories/) - summary by Woods Hole Oceanographic Institution ### 2.3 Mapping Marine hydrography or ocean scientific mapping involves the creation of high-resolution representations of the seafloor, water column, and other associated features or phenomena (e.g., fish migrations, vents or seeps bubbling up) using vessel-based sonar, autonomous vehicles, acoustic or optical tools. It is a type of remote sensing since the mapping instrument is not on the seafloor. Unlike satellite remote sensing, which observes only the ocean surface from space, hydrographic mapping is conducted from platforms within or on the ocean surface. These mapping systems can resolve fine-scale topography (seafloor bathymetry), subsurface geologic layers, water column imaging, and habitat mapping that integrates both physical and biological data. Laser and 3-D reconstruction are other forms of high-resolution mapping. - **History:** - 1870s–1900s: Early bathymetric charts created using lead lines, linked to colonial navigation and maritime claims as well as scientific knowledge creation - 1920s–1940s: Echo sounding developed for military and commercial navigation, later repurposed for seafloor mapping - 1950s–1970s: Multibeam sonar developed, enabling wider swath coverage (i.e. can map wider seafloor area, not just single points) and broader seafloor topography or bathymetry mapping - 2010s–present: Autonomous vehicles, numerous specialized sonar systems, and 3-D photogrammetry advance deep mapping capabilities - Today: Mapping remains uneven globally—nations with limited funding or access to ships and processing capacity are underrepresented and do not have detailed seafloor maps of the their waters - **Trade-offs:** - High-resolution, fine-scale maps of seafloor and water column features - Enables geologic, biologic, and habitat-based spatial analysis - Requires significant ship time, technical expertise, and post-processing - Data coverage is patchy, most of the seafloor remains unmapped - High cost and national interests impact where mapping occurs and who benefits from the data  *Bathymetric mapping using a hull-mounted multibeam sonar system. Black lines indicate the ship’s track, while the coloration represents depth differences (red is shallow, purple is deep) used for visualizing the bathymetric or topographic features of the seafloor. Source: https://www.worldofitech.com/mapping-the-ocean-floor-water-bathymetry-data/>* #### 2.3.1 Bathymetric Mapping - **About:** - Measurement and charting of the depth and shape of the seafloor - Typically uses sonar-based systems (e.g., single-beam or multibeam echosounders), mounted on ships, AUVs, or towed platforms - Short-range systems (e.g., ROV-mounted sonar) provide highly detailed data over small areas (centimeters in resolution), while medium-range systems (e.g., hull-mounted multibeam on ships or AUVs) cover much larger swaths with lower resolution - **Use cases:** - Mapping underwater topography and geological features - Planning submersible dives and identifying hazards - Supporting infrastructure projects like cables or offshore wind farms - Creating base maps for habitat mapping or biogeographic studies (i.e. understanding what marine life lives where and how their habitats are linked to geologic features as well as currents and physical oceanographic phenomena) - **Data collection:** - By research vessels or autonomous vehicles using sonar systems - Key manufacturers include Kongsberg, Teledyne, R2Sonic, and Edgetech - Data is processed using specialized hydrographic software (e.g., QPS Qimera, CARIS, MB-System) - **Key considerations:** - Requires calibration (e.g., sound speed profiles) and correction for vessel motion - Deep ocean mapping can be slow and resource-intensive - Interpretation of raw bathymetry data requires trained analysts and geospatial tools, it is not yet fully automated - **Data sources:** - European Marine Observation and Data Network - [EMODnet Bathymetry](https://emodnet.ec.europa.eu/en/bathymetry), multibeam datasets and other maps with a built in visualizer - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access), archive of US and global bathymetric surveys with visualizer - General Bathymetric Chart of the Oceans - [GEBCO Gridded Bathymetry Data](https://www.gebco.net/data-products/gridded-bathymetry-data), global map interface of compiled bathymetry - Global Multi-Resolution Topography - [GMRT data](https://www.gmrt.org/), global compilation of multibeam data (includes graphical map tool) - **Resources to learn more:** - [NOAA - What is bathymetry?](https://oceanservice.noaa.gov/facts/bathymetry.html) - [Seabed 2030](https://seabed2030.org/) – global initiative to map the entire seafloor by 2030 - [Using open-access mapping interfaces to advance deep ocean understanding](https://link.springer.com/article/10.1007/s40012-025-00410-2) (Johannes, 2025) - [GeoMapApp](https://www.geomapapp.org/) - free map-based application for browsing, visualizing and analyzing a diverse suite of curated global and regional geoscience data sets  *Plumes of bubbles emanating from the seafloor, indicating that there were methane gas seep ecosystems in this region. Sound waves reflect strongly off the gas bubbles and are visible in the water column data. Source: <https://nautiluslive.org/blog/2018/08/08/more-just-bathymetry-seafloor-mapping-tool-exploration>* #### 2.3.2 Water Column Mapping - **About:** - Acoustic systems—usually multibeam echosounders or special water column sonars—to detect and visualize features suspended in the ocean between the surface and the seafloor - Key for detecting gas plumes, biological layers (schools of fish or migrations in the twilight zone), suspended sediments, etc. - **Use cases:** - Observing midwater scattering layers (biological migrations) - Detecting hydrothermal plumes amd tracking gas plumes from methane seeps - **Data collection:** - Most multibeam systems include water column data modes, so it is often collected in tandem with bathymetry - From ships, AUVs, and ROVs - Data must be interpreted alongside oceanographic profiles (e.g., CTD casts) and often requires manual cleaning to reduce noise - **Key considerations:** - Processing and interpreting water column data is a bit tedious not yet standardized - Detection is sensitive to sonar frequency, range, and sea conditions - Validation with ground-truth sampling (e.g., bottle casts, nets, sensors) is helpful - **Data sources:** - European Marine Observation and Data Network - [EMODnet Physics](https://emodnet.ec.europa.eu/en/physics), includes some water column data layers - NOAA National Centers for Environmental Information - [NCEI water column sonar data](https://www.ncei.noaa.gov/maps/water-column-sonar/) - **Resources to learn more:** - [OET - More than just Bathymetry](https://nautiluslive.org/blog/2018/08/08/more-just-bathymetry-seafloor-mapping-tool-exploration) - Seafloor Mapping as a Tool for Exploration - [Migration in the Ocean Twilight Zone](https://twilightzone.whoi.edu/explore-the-otz/migration/) - often monitored with water column data #### 2.3.3 Seafloor Backscatter - **About:** - Analyzing the intensity of sound that is reflected or ‘scattered back’ from the seafloor when using sonar systems - Provides information about seafloor texture, hardness, and composition (e.g., sand, rock, mud) - Often conducted simultaneously with bathymetric mapping during ship-based or AUV surveys - **Use cases:** - Seafloor habitat classification or substrate mapping - Detecting anthropogenic objects or features (e.g., cables, wrecks) - Complements bathymetry for geologic or habitat models - **Data collection:** - Similar to bathymetry and water column mapping, backscatter data is collected using the same sonar systems and processed using similar software - **Key considerations:** - Requires calibration and post-processing to produce usable mosaics - Interpretation of sediment type from backscatter typically should be verified by ground-truth sampling (e.g., grabs, cores with ROVs or HOVs) - **Data sources:** - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access) - **Resources to learn more:** - NOAA - [How does backscatter help us understand the sea floor?](https://oceanservice.noaa.gov/facts/backscatter.html)  *Sediment layers seen in the sub-bottom profiler data collected in 2021 at the New England and Corner Rise Seamounts expedition on the NOAA Ship 'Okeanos Explorer'. Source: <https://oceanexplorer.noaa.gov/technology/sub-bottom-profiler/sub-bottom-profiler.html>* #### 2.3.4 Sub-bottom Profiling - **About:** - Uses low-frequency acoustic pulses to penetrate below the seabed and image sediment layers or other buried geologic features - Reveals vertical structures beneath the seafloor - Typically deployed from research vessels or towed systems - **Use cases:** - Studying sedimentation and geological processes - Locating subseafloor gas pockets or archaeological sites - For infrastructure planning or hazard assessment (e.g., submarine landslides) - **Data collection:** - Chirp profilers (high resolution, shallow penetration) and boomer/sparker systems (deeper penetration) are used - Operated from vessels with sonar equipment often collected simultaneously while collecting bathymetry and other mapping data, even if the sonar systems are different - **Key considerations:** - Resolution and penetration are inversely related (deeper = less detail) - Can be noisy and hard to interpret without ground-truthing (e.g., sediment cores) - **Data sources:** - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access) - European Marine Observation and Data Network - [EMODnet Geology](https://emodnet.ec.europa.eu/en/geology), includes sub-bottom and other forms of seafloor geological data - **Resources to learn more:** - NOAA - [Sub-Bottom Profiler](https://oceanexplorer.noaa.gov/technology/sub-bottom-profiler/sub-bottom-profiler.html)  *3-D reconstructed seafloor lava flows and hydrothermal vent field from the East Pacific Rise. This 3-D model was produced using downward facing video imagery and photogrammetry techniques. Credit: Mae Lubetkin* #### 2.3.5 Photogrammetry and 3-D Reconstruction - **About:** - Stitching together overlapping images or video frames from subsea camera systems often mounted on ROVs or AUVs - To create detailed mosaics or 3-D models of seafloor features - Uses optical data, offering true-color, high-resolution imagery (unlike acoustic mapping techniques described above) - **Use cases:** - Mapping hydrothermal vent fields, coral reefs, archaeological sites, etc. - Change detection in dynamic environments (e.g., volcanic or vent habitats, biological growth or loss) - Public engagement and educational tools - **Data collection:** - Collected by vehicle-mounted cameras with precise navigation and positioning - Software like Agisoft Metashape or custom photogrammetry pipelines are used for processing (which is easier now than ever before, becoming much more common in ocean sciences) - **Key considerations:** - Requires good lighting and water clarity - Processing is computationally intensive, and vehicle navigation data helps with plotting 3-D reconstructions onto broader bathymetric maps - Can be limited to small survey areas due to time constraints and battery limitations - **Data sources:** - Monterey Bay Aquarium Research Institute - [MBARI Sketchfab](https://sketchfab.com/mbari) - 3-D models of seafloor sites can be found in academic papers or at individual institutions or government agencies data repositories - **Resources to learn more:** - [Seafloor Futures](https://garden.ocean-archive.org/seafloor-futures/) (Lubetkin, 2024) - [Realtime Underwater Modeling and Immersion](https://nautiluslive.org/tech/realtime-underwater-modeling-and-immersion) - Ocean Exploration Trust - [Underwater 3-D Reconstruction from Video or Still Imagery: Matisse and 3-D Metrics Processing and Exploitation Software](https://www.mdpi.com/2077-1312/11/5/985) (Arnaubec et al., 2023) - [Seeing the Sea in 3-D](https://schmidtocean.org/cruise-log-post/seeing-the-sea-in-3d/) - Schmidt Ocean Institute ### 2.4 Satellite Remote Sensing Satellite data provides the most familiar, spatially complete picture of the ocean. This bird’s-eye perspective is invaluable for understanding large-scale phenomena like currents, sea surface temperature patterns, and phytoplankton blooms, providing visual evidence that can enhance storytelling. However, there are unique considerations since, unlike the use of satellite imagery on land, most of our understanding of the ocean does not come from the visual spectrum. In this section, we’ll introduce three of the most important types of satellite ocean data and discuss the use cases for each. - **History:** - 1978: NASA launched Seasat, the first satellite designed for ocean research. - Significant expansion in the 1990s with missions including TOPEX/Poseidon (ocean altimetry), AVHRR (high-resolution sea surface temperature), and SeaWiFS (ocean biology). - Modern constellations are operated by NASA, NOAA, ESA, EUMETSAT, CNES, ISRO, and others. - **Trade-offs:** - Excellent spatial coverage that’s impossible to achieve with ships or buoys. - Very high costs, these platforms are operated by government agencies. - Only seeing the very surface of the ocean, no subsurface data. - Limited horizontal resolution (spatial detail) and temporal resolution (orbital repeat time).  *Gulf of Mexico SST on a cloud-free data. Source: <https://marine.rutgers.edu/cool/data/satellites/imagery/>* #### 2.4.1 Radiometry - Sea Surface Temperature (SST) - **About:** - Sea surface temperature (SST) is the oldest and most extensive application of satellite oceanography. - **Use cases:** - Tracking climate change, El Niño, and marine heat waves. - Key input for weather models (e.g., very important for hurricane forecasting). - Mapping ocean eddies, currents, and upwelling, which are critical to fisheries. - **Data collection:** - Two separate types of sensors measure SST: Infrared and microwave. - IR sensors have higher spatial resolution ([1-4 km](https://coastwatch.noaa.gov/cwn/product-families/sea-surface-temperature.html)) and finer temporal coverage but cannot “see” through clouds, which block over 70% of the ocean at any given time. - Microwave sensors can see through most non-precipitating clouds but have a lower spatial resolution (about 25 km) and don’t work near the coastline. - Measures temperature of the top ~1 mm of the ocean - Blended products: Combine multiple sensors for better coverage (e.g., GHRSST L4)  *Example SST data at different processing levels. (Merchant et al. 2019): <https://www.nature.com/articles/s41597-019-0236-x>* - **Key considerations:** - Make yourself aware of the different processing levels when accessing data. Level 4 (L4) will be the easiest to work with but may not be fully accurate. - L2: Data along the original orbital track. - L3: Gridded data, sometimes averaged over time. - L4: Cloud-free, gaps are filled by various methods depending on the source. - Temporal resolution depends on the satellite orbit. There are good options that blend multiple satellites. - **Data sources:** - [EU Copernicus Marine Service](https://marine.copernicus.eu/) - [US NASA Physical Oceanography DAAC](https://podaac.jpl.nasa.gov/) - [NOAA CoastWatch](https://coastwatch.noaa.gov/cw_html/cwViewer.html) - Graphical Interface - **Resources to learn more:** - [Group for High Resolution Sea Surface Temperature (GHRSST)](https://www.ghrsst.org/ghrsst-data-services/for-sst-data-users/) #### 2.4.2 Radar Altimetry - Sea Surface Height (SSH) - **About:** - Measures ocean surface height by sending radio pulses and measuring return time. - SSH can tell us the strength of large scale currents like the Gulf Stream, as the slope of the sea surface is used to calculate the “geostrophic current”. - **Use cases:** - Key to understanding ocean circulation and long-term sea level rise. - **Data collection:** - Radar altimeters on satellites measure SSH directly (e.g., Jason-3, Sentinel-6) and then the geostrophic currents are calculated in a post-processing step. - Spatial resolution is significantly worse than SST (25+ km) - The recent SWOT satellite is a new type of altimeter with much higher resolution but has very limited coverage since there is only one currently in orbit. - **Key considerations:** - SSH is useful for large-scale ocean currents but not coastal tidal currents. - Similar to SST, be careful about processing level and look for re-gridded datasets. - Can generally see through clouds, so gaps are not a significant issue. - **Data sources:** - [EU Copernicus Marine Service](https://data.marine.copernicus.eu/products?facets=mainVariables%7ESea+surface+height--sources%7ESatellite+observations) and [Aviso](https://www.aviso.altimetry.fr/en/data/products/sea-surface-height-products/global.html) - [US NASA PODAAC](https://podaac.jpl.nasa.gov/NASA-SSH) - [Copernicus MyOcean Pro](https://data.marine.copernicus.eu/viewer/expert) and [Aviso](https://seewater.aviso.altimetry.fr/) - Graphical Interfaces - **Resources to learn more:** - [NASA JPL - What is Ocean Surface Topography?](https://podaac.jpl.nasa.gov/OceanSurfaceTopography)  *Global map of marine Chlorophyll concentration. Source: <https://sos.noaa.gov/catalog/datasets/biosphere-marine-chlorophyll-concentration/>* #### 2.4.3 Optical - “Ocean Color” - **About:** - Ocean color sensors measure the reflectance of sunlight from the ocean surface to infer biological and chemical properties, such as algal concentration, suspended sediments, and water clarity. - **Use cases:** - Tracking phytoplankton blooms and changes in marine ecosystems. - Useful for monitoring water quality, including coastal sediment and oil spills. - **Data collection:** - Sensors measure light reflected from the ocean at different wavelengths (e.g., MODIS, VIIRS, Sentinel-3 OLCI) and then apply algorithms in order to calculate variables such as Chlorophyll-a concentration. - **Key considerations:** - Ocean color data is significantly affected by cloud cover, aerosols, and atmospheric correction errors. - **Data sources:** - [EU Copernicus Marine Service](https://data.marine.copernicus.eu/products?facets=mainVariables%7EPlankton--sources%7ESatellite+observations) - [US NASA Ocean Color Web](https://oceancolor.gsfc.nasa.gov/data/find-data/) - [NASA Worldview](https://worldview.earthdata.nasa.gov/) - Graphical Interface - **Resources to learn more:** - [IOCCG (International Ocean-Color Coordinating Group)](https://ioccg.org/) ### 2.5 Additional databases and scientific support The four sub-sections above (*In Situ* Sensors; Deep Ocean Observation, Exploration, and Research Systems; Mapping; Satellite Remote Sensing) cover the main areas of ocean scientific data types and collection methods. There are some datasets that are not discussed in this guide since they are likely less useful for investigative storytelling or require technical skills to access and interpret the data. Below are some additional databases and information on contacting scientists to support your investigation. While in section 3, we outline a case study using real data to tell an ocean story. #### 2.5.1 Additional databases, collections, and visualizers Sites that either did not fit into one of the sub-sections above, or that contain information which is generated after scientific studies occur: - [PANGAEA](https://pangaea.de/) - data publisher for earth and environmental science (across disciplines) - [International Seabed Authority DeepData](https://www.isa.org.jm/deepdata-database/) - database hosting all data related to international deep-seabed activities, particularly those collected by contractors (i.e. nations or entities) during their exploration activities and other relevant environmental and resources-related data. Includes a dashboard and map to search for basic stats and information about what contractors have done during deep seabed mining exploration cruises. - [Marine Geoscience Data System](https://www.notion.so/Ocean-Datasets-for-Investigations-1caf92221af780c68873c2aecf9b3479?pvs=21) - geology and geophysical research data across collections - [USGS Earthquake Hazards Program](https://www.usgs.gov/programs/earthquake-hazards/earthquakes) - interactive map with magnitudes and additional information (earthquakes can occur on land and in the ocean) - [WoRMS – World Register of Marine Species](https://www.marinespecies.org/) - comprehensive taxonomic list of marine organism names - [OBIS – Ocean Biodiversity Information System](https://obis.org/) - global open-access data and information on marine biodiversity - [Windy](https://www.windy.com/) - animated weather maps, radar, waves and spot forecasts #### 2.5.2 Scientific support All of the datasets and databases we outlined above are free and open to the public. We hope that we outlined enough context and links to user-friendly platforms to access the data so that you feel empowered to conduct your own investigations with ocean datasets. That said, some data might be more challenging to work with depending on prior experience and computing skills, among other factors. When in doubt, you can always contact an ocean scientist to ask questions or seek support. Depending on your investigation or story, you will need to contact a specific type of ocean scientist since each has their own specialty. You can start by searching for and contacting scientists at nearby universities or research institutes. **Ocean scientists and their specializations:** - **Physical Oceanographers -** Study ocean currents, tides, waves, and ocean-atmosphere interactions. They can help explain phenomena like sea level rise or how ocean circulation affects weather and climate. - **Chemical Oceanographers -** Focus on the chemical composition of seawater and how it changes over time. Useful for stories involving ocean acidification, pollution, nutrient cycling, or chemical runoff impacts. - **Biological Oceanographers or Marine Biologists -** Study marine organisms and their interactions with the ocean environment. They are ideal sources for stories on biodiversity, fisheries, invasive species, and ecosystem health. - **Geological Oceanographers or Marine Geologists -** Study the structure and composition of the ocean floor. They can provide insights into underwater earthquakes, tsunamis, deep-sea mining, or the formation of underwater features. - **Climate Scientists with Ocean Expertise -** Examine how oceans influence and respond to climate change. They are helpful for broader climate stories that involve ocean heat content, carbon storage, or long-term trends in ocean conditions. - **Marine Ecologists -** Study relationships among marine organisms and their environment. They can clarify ecosystem-level impacts, like those from overfishing, coral bleaching, or marine protected areas. - **Fisheries Scientists -** Specialize in fish populations, fishing practices, and resource management. Helpful for reporting on commercial fishing, stock assessments, or policy/regulation issues. - **Ocean Data Scientists -** Work with large marine datasets and modeling, can assist with interpreting satellite data, ocean models, or big datasets. - **Marine Policy Experts and Ocean Economists -** Focus on the intersection of ocean science, law, and economics. Helpful for coverage of marine regulations, governance issues, or the ‘blue economy.’ - **Marine Technologists or Ocean Engineers -** Design and use tools like underwater drones, sensors, and buoys. They can help explain how ocean data is collected and what the limitations of certain technologies might be. ## 3. Case Study: Gulf of Maine Ocean Warming  As ocean scientists from the Northeastern United States, we have each witnessed how rapid ocean changes are affecting the ecosystems and communities around us. For this example case study, we focus on the Gulf of Maine—a region close to home. When telling ocean stories, it is helpful to have either first-hand or personal connections to the coastal or oceanic region you are investigating. ### 3.1 Motivation The Gulf of Maine is warming [faster than 99% of the global ocean](https://eos.org/features/why-is-the-gulf-of-maine-warming-faster-than-99-of-the-ocean), making it a key site to investigate local impacts of climate change on marine environments and coastal livelihoods. Stories of changing fish stocks and stressed fisheries are already discussed in communities within the Northeastern region. Before getting into the data, we will think through the historical and ecological context of the Gulf of Maine and its fisheries. For centuries, the regional identity has been deeply linked to the ocean. Indigenous [Wabanaki peoples](https://www.wabanakialliance.com/wabanaki-history/)—including the Abenaki, Mi'kmaq, Maliseet, Passamaquoddy, and Penobscot nations—relied on these coastal waters for food as well as cultural practices and trade. They managed their coastal and marine environments with ocean knowledge developed across generations. When European colonization began, [intensive cod fishing fueled transatlantic trade and early settlements](https://www.markkurlansky.com/books/cod-a-biography-of-the-fish-that-changed-the-world/). Europeans considered the cod fisheries to be so abundant that they were endless. The overfishing by settlers caused a massive collapse in cod stocks by the 1950s. Now, other important local fisheries like the American lobster are being impacted by the combination of ocean warming and historic overfishing. Harmful algal blooms have also increased in frequency which indicate that the broader Gulf ecosystems are under stress. In the following sections, we guide you through using publicly available ocean datasets to investigate the scientific questions behind the Gulf of Maine and its warming waters. By accessing current and archival datasets you will be able to visually show the seawater temperatures going up and connect that to other environmental stories or investigations about the Gulf. ### 3.2 Data acquisition In order to investigate warming in the Gulf of Maine, we will analyze surface temperatures from two different datasets: a local *in situ* temperature sensor and the global-average SST. With the global SST as our baseline, we’ll be able to determine how much faster the Gulf of Maine is warming compared to the rest of the world. This analysis involves database downloads, data post-processing/analysis, and data visualization. If you don’t have experience with coding and want to get started with the Python programming language, see the appendix for tips on getting setup. Otherwise, you can always consider contacting a scientist to support you with your investigation (see section 2.5.2 Scientific support). #### 3.2.1 Gulf of Maine buoy temperature dataset First, we’ll go to the [National Data Buoy Center (NDBC)](https://www.ndbc.noaa.gov/) website and look for a buoy in the Gulf of Maine with a long historical record of temperature measurements. Clicking on the “Historical Data & Climatic Summaries” link at the bottom of [Station 44007’s page](https://www.ndbc.noaa.gov/station_page.php?station=44007) reveals annual text files going back to 1982.  *Screenshot from the NDBC website showing potential buoys to use for Gulf of Maine case study.* The task now is to process all of this data into a more usable format. We’ll do this with a python script using the [pandas](https://pandas.pydata.org/) data analysis library. 1. Loop through the years 1982-2024 and create the dataset url for each year, using the NDBC website to deduce the url structure. 2. Load the text data directly from each url via `pandas.read_csv()` 3. Convert the year, month, day, hour columns into a single pandas datetime column. 4. Combine all of the data into a single dataframe. 5. Save our data to a new CSV file.  *An example of what the available buoy data looks like for the year 1985. The highlighted sections show the parts of the dataset that we’re interested in: the date/time and the water temperature.* #### 3.2.2 Global mean SST dataset Next, we want a corresponding dataset for the globally-averaged SST, in order to determine whether the Gulf of Maine is warming faster or slower than the average. The [Climate Reanalyzer](https://climatereanalyzer.org/clim/sst_daily/?dm_id=world2) displays globally-averaged SST from the [NOAA 1/4° Daily Optimum Interpolation Sea Surface Temperature (OISST)](https://www.ncei.noaa.gov/products/optimum-interpolation-sst), a long term Climate Data Record that incorporates observations from different platforms (satellites, ships, buoys and Argo floats) into a regular global grid. 1/4° refers to the grid resolution—about 25 km. There is an option to download the underlying data from the Climate Reanalyzer website, which will save us a lot of time vs. trying to access decades of data and doing the global averaging ourselves. The data is available as a JSON file, which is a different text file format that will require a more custom approach for converting into a pandas dataframe.  *Screenshot of the [Climate Reanalyzer](https://climatereanalyzer.org/clim/sst_daily/?dm_id=world2) website. In the dropdown menu, we want to download the JSON data.* One key concept to note is how this data handles dates. Each year includes a list of 366 temperatures, without any explicit list of the corresponding dates. This is using the format of “day of year” and we see that the last temperature is “null” for non-leap years. When processing this data, we need to take this into account and ignore the null final value. Similar to the buoy data, we’ll re-format this data in a pandas dataframe and save to a new CSV file.  *A look inside the globally-averaged SST JSON file. The data is arranged as a list of years where each year has a list of 366 temperatures.* ### 3.3 Climatological data analysis A standard method for analyzing climate change anomalies is to first remove the climatological “seasonal” signal from the data. This will allow us to show, for each data point, how much warmer or colder it was than the average temperature for that day of the year. The first step is choosing which time period we’ll use for our climatology “baseline”. Here we’ve chosen 1991 to 2020 since it is fully covered by our data and matches the climatology period used by the Climate Reanalyzer website. Next, we’ll use some built-in pandas methods to get the climatological average temperature for each day and then map that to each datapoint in our timeseries. The following code snippet shows the steps used for both the buoy data and the global SST: ```python # Select just the data in the range of the climatology period df_clim = df[(df.index.year >= 1991) & (df.index.year <= 2020)].copy() # Assign the day of year (1-366) to each data point in the timeseris df_clim["day_of_year"] = df_clim.index.dayofyear # Take the mean for each day_of_year df_clim = df_clim.groupby("day_of_year")["temp"].mean() # New variable in df: the climatological temperature for that day df["day_of_year"] = df.index.dayofyear df["climatology_value"] = df["day_of_year"].map(df_clim) # Temperature anomaly is observed temperature minus climatological temperature df["anomaly"] = df["temp"] - df["climatology_value"] ```  *Our resulting dataframe includes new columns for climatology and temperature anomaly.* ### 3.4 Analyzing and Visualizing the results First, we’ll simply plot the full temperature timeseries and see what we find.  The warming signal is instantly apparent in the global SST data because the seasonal signal is so small. The Gulf of Maine, however, varies by more than 15° C throughout the year so any long term changes are difficult to see in this format. Plotting the climatology signal illustrates this point (pay attention to the y-axis).  Next we’ll view our temperature anomaly data (observed temperature minus climatology). As expected, there is more noise in the buoy data since it’s taken from a single point and any given day can vary by as much as 4 °C from climatology. The globally-averaged temperature has much less variance.  For the final version of our plot, we’re incorporate 3 changes: 1. Fit a simple linear regression using [numpy’s polyfit](https://numpy.org/doc/stable/reference/generated/numpy.polyfit.html) in order to quantify the average rate of warming for the two datasets. 2. Plot the monthly averages instead of the daily values in order to simplify the visual clutter. 3. Use the same y-axis range for the two plots for direct visual comparison.  Comparing our warming rate calculations against the published literature finds good agreement: - Gulf of Maine SST: our rate of 0.496°C/decade is within 5% of the 0.47°C/decade reported by the [Gulf of Maine Research Institute](https://gmri.org/stories/2024-gulf-of-maine-warming-update/). This is likely due to differences in methods—we used a single buoy and they used the OISST data averaged across the entire Gulf. - For global SST, our rate of 0.188 °C/decade is within 5% of the 0.18 °C/decade (over the past 50 years) published by [Samset et al. (2023)](https://www.nature.com/articles/s43247-023-01061-4). These final plots provide simple visual evidence of the Gulf of Maine’s rapid warming over the past 40 years. We showed the data transform from text files, to noisy timeseries, and finally to expert-validated trend lines. By removing the strong seasonal signal and focusing on the anomalies, we can clearly see the long-term warming trend in both the Gulf of Maine buoy data and the global mean SST. Finally, note that the linear regression is useful for quantifying the recent warming in an easily understandable number but is not necessarily a predictor of future warming. The [Maine Climate Science Dashboard](https://climatecouncil.maine.gov/future/climate/science-dashboard) shows the potential for human emissions to either accelerate or slow down this rapid warming.  *The [Maine Climate Science Dashboard](https://climatecouncil.maine.gov/future/climate/science-dashboard) combines historical water temperature measurements with different climate scenario forecasts.* ## 4. Conclusion Our investigation into Gulf of Maine temperatures, using readily available public datasets, highlights one local manifestation of global climate change. This rapid warming isn't merely an abstract data point, it continues to have profound implications for the region’s biodiversity and the human communities who rely on the ocean. Marine species are highly sensitive to temperature changes, and the Gulf of Maine has been experiencing a noteworthy decline in native species and [increase in warmer-water species](https://online.ucpress.edu/elementa/article/9/1/00076/118284/Climate-impacts-on-the-Gulf-of-Maine-ecosystemA). The next steps in this story might look to other data sources to explore: Why is the Gulf of Maine warming so quickly? and What will the region look like in the future? or How exactly are local fisheries affected by warming waters? This case study is one example of how to find the connections between global environmental change, local ocean data, and tangible human impacts. This process offers a template for investigating similar stories in your own regions: 1. **Start with a local observation or community concern:** What are people witnessing or experiencing in your local environment? 2. **Explore the scientific context:** Consult with scientists, read relevant research, and understand the underlying environmental drivers. 3. **Seek out publicly available data:** As shown in section 2, there is a large assortment of high-quality public ocean datasets that can be used to investigate countless questions. 4. **Connect the data back to human issues:** How do the environmental changes revealed by the data affect local cultures, livelihoods, health, and economies? The key thing to remember is that there are multiple angles to uncover and expose often-invisible impacts to the ocean. Datasets provide one lens to report on climate and environmental changes, but these stories impact communities and are thus both political and social. Just as ocean science has changed and begun to decolonize, it's crucial to investigate and tell stories that reflect diverse experiences. Ocean data can help highlight intersecting issues—such as deep seabed mining, marine health, and colonial continuums—with evidence-based information and compelling visualizations. We hope this guide offers a practical starting point for navigating ocean science, accessing and interpreting data, and connecting your investigation to real-world consequences that are planetary in scale yet intimately local. ```cik-note ``` >**APPENDIX: Getting started with python** > >If you have not done any coding before, the initial task of setting up your coding environment can be a challenging hurdle. There are multiple options for code editors/IDEs (integrating development environment), ways of handling dependencies (the external packages you install to give you advanced functionality), and other decisions that are outside the scope of this article. Luckily, once you’ve chosen your tools, there are good resources online so here are a few recommendations and then you can seek out more detailed tutorials: > >1. Use Visual Studio Code as your code editor (the application where you will write and run code). This is the most popular option and there is an extensive ecosystem of 3rd party plug-ins and help resources. <https://code.visualstudio.com/> > >2. Use [conda](https://docs.conda.io/projects/conda/en/latest/user-guide/getting-started.html) for package management. Your computer’s operating system may come with a version of python pre-installed but it's not a good idea to install packages onto this global location. Instead, create separate conda "environments" for different projects. This will allow you to experiment in a safe and organized way. Here is a helpful article on the VS Code website: <https://code.visualstudio.com/docs/python/environments>. For example, to create a new environment that we'll name "ocean-study" and install the "matplotlib" plotting package would look like this: > > ```bash > conda create -n ocean-study > conda activate ocean-study > conda install matplotlib > ``` > Now, in VS Code, just make sure your Python Interpreter is using this environment (it will look something like `~/miniconda3/envs/ocean-study/bin/python` and you will be able to use the matplotlib package in your code. > >3. Finally, consider using Jupyter Notebooks for exploratory coding where you’re loading datasets and making plots. Notebooks have the file extension `.ipynb` and allow you to run chunks of code independently in code "cells" and view the output right below. You can also use Markdown text cells to write notes and explanations for yourself and collaborators. Instructions on using VS Code: <https://code.visualstudio.com/docs/datascience/jupyter-notebooks> > <hr class="thick"> #### **Mae Lubetkin** is an ocean scientist, transmedia artist, and writer based in Paris and at sea. Their practice-led research remaps our relations to bodies of water and digital worlds by means of investigation, counter-narrative, and memory. With a background in marine geology and subsea imaging, their artistic practice is in dialogue with Science while situated in queer, intersectional, anti-extractivist, and decolonial frameworks. Guided by wet-techno-critical studies and thinking with other-than-human worlds, they compose environmental traces in installations and digital outputs. Their core practice is in solidarity with submerged, ancient, ephemeral and imaginary environments. **Dr. Kevin Rosa** is an oceanographer and the founder of Current Lab, a startup specializing in computational ocean forecasting. He holds a B.A. in Physics and a PhD in Physical Oceanography from the University of Rhode Island, with a focus on ocean physics and hydrodynamic modeling. <hr class="thick"> *Published in June 2025* |
|
None
String updated in the repository |
# Credits and Licensing
 ## Credits **Contributing Authors:** Ankita Anand, Deepak Adhikari, Yung Au, Manuel Beltrán, Annabel Church, Xavier Coadic, Hang Do Thi Duc, Ricardo Ginés, Alison Killing, Jess Lempit, Hanna Liubakova, Offray Luna-Cárdenas, Amber Macintyre, Filip Milosevic, Matt Mitchell, Bianca Mondo, Brad Murray, Miguel Pinheiro, Megha Rajagopalan, Wael Eskandar, Laura Ranca, Nayantara Ranganathan, Sajad Rasool, Mario Rautner, Gabi Sobliye, Nuria Tesón, Carolyn Thompson, Marek Tuszynski, Jose Felix Farachala Valle, Chris Walker, Johanna Wild, Lylla Younes, Nicola Bruno, Tetyana Bohdanova, Kevin Rosa, Mae Lubetkin, Stavros Malichudis. **Editors:** Natalie Holmes, Christy Lange, Tyler McBrien, Karolle Rabarison, Laura Ranca, Michael Runyan. **Illustrator:** Ann Kiernan. **Design:** Tactical Tech's design team - Yiorgos Bagakis, Cade Diehm, Ida Flik, Philipp Dollinger. **Developers:** Jacopo Anderlini, Laurent Delleré, Wael Eskandar, Zach Green, Danja Vasiliev, Chris Walker. **Communications:** Ana Maria Salinas, Daisy Kidd, Sasha Ockenden. **Translations - French:** Xavier Coadic. **Translations - Spanish:** Casa Digital, Datos Protegidos, Cooperativa Sulá Batsú, Bruno Álvarez Herrero, Jacinto Pariente, Carolina Soligo, José Monserrat Vicent. **Translations - Portuguese:** Biti Averbach, Eduardo Macedo, Thiago Antônio Vieira, Yael Berman, Celso Bessa. **Project Team Behind The Kit:** Wael Eskandar, Christy Lange, Matt Mitchell, Laura Ranca, Marek Tuszynski, Chris Walker, Lieke Ploeger **Special Thanks:** This kit would not have been possible without our collaboration with [Share Lab](https://labs.rs/en/) and [Share Foundation](https://www.sharefoundation.info/en/) in co-hosting the 2017 Data Investigation Camp and the 2018 Citizen Investigation Kit Residency. The Kit would also not have been possible without the hard work of Exposing the Invisible's former project lead Gabi Sobliye. Special, special thanks to the Perast Group! ### ### A publication by: [](https://tacticaltech.org/) [](https://exposingtheinvisible.org/) ### This Kit was possible thanks to the support of:   and Tactical Tech's other funders. ## Licensing **CC BY-SA 4.0** The Kit is licensed under a [Creative Commons Attribution-ShareAlike 4.0 International license](https://creativecommons.org/licenses/by-sa/4.0/)  <hr class="thick"> [Get started with The Kit](https://kit.exposingtheinvisible.org/)
# Credits and Licensing
 ## Credits **Contributing Authors:** Ankita Anand, Deepak Adhikari, Yung Au, Manuel Beltrán, Annabel Church, Xavier Coadic, Hang Do Thi Duc, **Editors:** Natalie Holmes, Christy Lange, Tyler McBrien, Karolle Rabarison, Laura Ranca, Michael Runyan. **Illustrator:** Ann Kiernan. **Design:** Tactical Tech's design team - Yiorgos Bagakis, Cade Diehm, Ida Flik, Philipp Dollinger. **Developers:** Jacopo Anderlini, Laurent Delleré, Wael Eskandar, Zach Green, Danja Vasiliev, Chris Walker. **Communications:** Ana Maria Salinas, Daisy Kidd, Sasha Ockenden. **Translations - French:** Xavier Coadic. **Translations - Spanish:** Casa Digital, Datos Protegidos, Cooperativa Sulá Batsú, Bruno Álvarez Herrero, Jacinto Pariente, Carolina Soligo, José Monserrat Vicent. **Translations - Portuguese:** Biti Averbach, Eduardo Macedo, Thiago Antônio Vieira, Yael Berman, Celso Bessa. **Project Team Behind The Kit:** Wael Eskandar, Christy Lange, Matt Mitchell, Laura Ranca, Marek Tuszynski, Chris Walker, Lieke Ploeger **Special Thanks:** This kit would not have been possible without our collaboration with [Share Lab](https://labs.rs/en/) and [Share Foundation](https://www.sharefoundation.info/en/) in co-hosting the 2017 Data Investigation Camp and the 2018 Citizen Investigation Kit Residency. The Kit would also not have been possible without the hard work of Exposing the Invisible's former project lead Gabi Sobliye. Special, special thanks to the Perast Group! ### ### A publication by: [](https://tacticaltech.org/) [](https://exposingtheinvisible.org/) ### This Kit was possible thanks to the support of:   and Tactical Tech's other funders. ## Licensing **CC BY-SA 4.0** The Kit is licensed under a [Creative Commons Attribution-ShareAlike 4.0 International license](https://creativecommons.org/licenses/by-sa/4.0/)  <hr class="thick"> [Get started with The Kit](https://kit.exposingtheinvisible.org/) |
|
None
Resource updated |
The “
source/en/the-kit-en.json” file was changed.
4 months ago
|
|
None
String updated in the repository |
Ocean Datasets for Investigations
=============================== By Mae Lubetkin and Kevin Rosa  ```cik-in-short ``` **In short:** Learn how to identify and use Ocean datasets as a tool for revealing the unseen or underreported dynamics of the world's most significant bodies of water. --- ## 1. Introduction: Ocean Science, Data, Storytelling Given our current planetary condition, many of the world's most pressing stories are linked to the ocean. Covering 71% of Earth's surface and interacting constantly with the atmosphere, the ocean is our greatest carbon sink and an essential indicator of climate change. Despite its critical role in maintaining a habitable planet and supporting coastal livelihoods, the ocean is often invisible to the lived experience of most individuals, particularly those far from its shores. There are so many ways to tell stories about the ocean, and countless diverse perspectives from which to understand it. Now, more than ever, we need to incorporate cross-cultural and trans-disciplinary strategies for investigative projects, particularly those concerning the ocean. This guide presents ocean datasets as a tool for revealing the unseen or underreported dynamics of the world's most significant bodies of water. For informed investigations and impactful storytelling, oceanographic datasets can be an essential resource for journalists, activists, and anyone interested in data-driven methods to communicate the climate crisis, environmental change, natural disasters, extractivism, and associated ocean justice issues. From bathymetric maps, subsea imagery, and 3-D habitat models, to satellite-derived and *in situ* real-time monitoring data – a vast amount of oceanographic media and data is publicly available. In this Exposing the Invisible Guide, we begin with an introduction on the broader scientific history and context within which ocean data is collected, stored, and made accessible. Section two outlines the diversity of datasets, including some history, trade-offs, use cases, data collection methods, data sources, and resources to learn more about each data type that we present. Section three offers a specific application of ocean data in a case study, including: steps explaining why the data is useful for supporting this particular story; how to source and present the data; and, finally, how to bring it into a meaningful investigative report, journalistic piece, or other storytelling format. The guide concludes with a summarized approach for using ocean datasets in investigations and outlines strategies for identifying the right ocean scientists who could support you and your investigation.  *Boulder resting on the seafloor offshore the Revillagigedo Archipelago in the Pacific Ocean. Credit: Ocean Exploration Trust.* ### 1.1 Ocean Science: History and Context Ocean science generally refers to the observation and investigation of biological, geological, physical, and chemical processes that shape and constitute global marine environments. This broad disciplinary field includes numerous sub-disciplines that focus on intricate, detailed studies of specific scientific questions concerning the ocean. Ocean scientists monitor habitat change, measure biodiversity, investigate geological phenomena, and study human impacts on ocean systems (i.e., global warming, pollution, overfishing, and extractive projects). Interactions between marine ecosystems and human activities, as well as atmospheric and coastal processes, are all carefully investigated by ocean scientists today. Despite niche specializations, there are increasingly multidisciplinary projects that involve diverse experts in order to more comprehensively understand the interconnections between oceanic processes and phenomena. Collectively, this research improves our baseline knowledge of the ocean, which can then support preservation strategies while maintaining sustainable and regenerative relationships with diverse marine ecosystems. Although ‘contemporary’ ocean science has deep roots in European colonialism and imperial exploration, ocean knowledge systems long predate Western scientific inquiry. Indigenous and coastal communities across Oceania, as well as the Atlantic and Indian Oceans, carefully studied and navigated the seas for thousands of years. These forms of ocean science are less known or dominant on a global scale, but they nevertheless involve highly sophisticated techniques for understanding the stars, ocean swells, winds, and currents. Navigators across Oceania used interconnected and embodied forms of ocean science, on their own terms, to journey vast distances across the seas with tremendous precision. While other coastal Indigenous peoples around the world developed specific place-based systems of knowledge, including both land and marine management practices that viewed these spaces as highly linked. Most of these communities understood the ocean not as a monstrous or alien-filled void (as European explorers often depicted it), but as a vast world interconnected with our own.  *Rebbilib (navigational chart) by a Marshall Islands artist in the 19th to early 20th century. These stick charts were used by Marshall Islander navigators during long ocean voyages. Credit: Gift of the Estate of Kay Sage Tanguy, 1963. Source: <https://www.metmuseum.org/art/collection/search/311297>* When European seafarers began worldwide exploration voyages in the 15th-16th centuries, they dismissed or actively erased these Indigenous ocean knowledge systems. European or 'Western' scientific models were linked to a colonial mindset that often viewed the natural world as a space to examine in order to master and own its elements, rendering them as 'resources'. Notions of relationality were strongly opposed to the point that scientists considered their surroundings as objects to be studied rather than elements to relate to or work with. At the core, these opposing ocean scientific methods or knowledge systems reflected the specific values and worldviews of each culture, respectively. The Challenger Expedition (1872–1876) was the first European-led systematic scientific exploration of the global oceans. In some ways, it was groundbreaking. However, it also played a key role in the broader colonial project, which aimed to map, control, and extract 'resources' from around the world. Today, Western or 'contemporary' ocean science continues to use investigatory methods that stem from European knowledge systems. Oceanographic research often occurs in the context of economic or territorial expansion and military-supported science projects. Nevertheless, these methods are beginning to open up to other forms of knowledge creation that move beyond the geopolitical interests of wealthy nations.  *Map of ocean currents created by John Nelson using the "WGS 1984 Spilhaus Ocean Map in Square" projected coordinate system in ArcGIS. The Spilhaus Projection (developed by oceanographer Athelstan Spilhaus in 1942) reflects aspects of decolonial cartography by shifting focus from land-centered perspectives to an oceanic worldview. Source <https://storymaps.arcgis.com/stories/756bcae18d304a1eac140f19f4d5cb3d>* Thanks to the enduring activism and decolonizing work led by many Indigenous and coastal communities, there is increasing recognition of the need to reclaim ocean science by amplifying the knowledge and perspectives of those who have long understood the seas. Ocean science is just beginning this deep process of decolonization, which first seeks to acknowledge and reckon with the violent and ongoing impacts of colonization. Decolonizing ocean science requires a fundamental shift in who holds agency and sovereignty over their own ocean waters. This also relates to international waters and who is included, excluded, or undervalued throughout negotiations concerning the legal and regulatory structures governing global oceans. Today, many ocean scientific projects are co-designed and co-led by ocean knowledge holders from diverse backgrounds. Collecting ocean datasets requires a team of experts who follow cultural protocols, ensure environmental safety, and apply diverse scientific methods, all while striving for more relational practices. ### 1.2 Ocean Data Collection Today Today, ocean datasets are collected by ocean scientists in collaboration with ocean engineers. These datasets are gathered from several sources to understand the global ocean and its role in maintaining Earth's habitability and critical planetary cycles. Ocean engineers develop the tools, platforms, and instruments that are required for data collection, such as underwater vehicles, satellite-mounted sensors, and buoys. By designing technologies that can operate in diverse and sometimes extreme conditions, these engineers support and expand ocean scientific capabilities. Together, ocean scientists and engineers advance our understanding of the planet for both research and conservation. There is a considerable variety of ocean data types, tools for data collection, and associated databases to store these recorded entities. This diversity of datasets is outlined in section 2. Like most scientific fields, funding can be secured from public governmental bodies or private sources. The ocean datasets we focus on here are publicly accessible and typically funded by governments via taxpayer contributions. This means that ocean datasets are for the people and should be accessible. Unfortunately, many public ocean datasets are stocked in complex databases and require specialized software, programming experience, or extensive knowledge to access. That said, there are plenty of datasets that can be retrieved and visualized more easily, with little to no background knowledge. There are also some ocean datasets that can be accessed with helpful tips and instructions, which is what we will focus on here. The Exposing the Invisible Toolkit is designed as a self-learning resource, we hope that this guide will support future investigations and make ocean datasets more accessible to communities and investigators around the world. ### 1.3 Data Gaps, Capacity Initiatives, Ocean Defenders Conducting ocean science can be a costly endeavor. Depending on the environment, scientific goals, and technical requirements, some ocean scientific work can only be conducted by wealthy nations or private organizations. Typically, this kind of science takes place at sea or uses remote sensing techniques. For example, deep ocean exploration and research requires a ship, vehicles, or platforms to deploy down to the deep ocean, technical and computing systems aboard to process the data, and a diverse team of experts to manage these operations. In contrast, satellite remote sensing used for ocean research typically covers the entire Earth surface. Publicly funded satellite-derived ocean datasets can be useful across territorial waters, throughout international seas, and are accessible regardless of one's nationality. Near-shore ocean science and *in situ* coastal monitoring efforts are more financially affordable, especially as diverse knowledge systems merge with lower-cost technologies and capacity initiatives. In this context, capacity refers to the skills, resources, and knowledge needed to effectively conduct ocean science. As ocean science undergoes a process of decolonization, this emphasis on capacity building, development, and sharing is also strengthening. Additionally, several international ocean law and policy frameworks specifically aim to increase ocean science capacity. Ocean defenders are also central to these efforts. As groups, individuals, or organizations dedicated to protecting marine environments, defenders play a key role in advocating for capacity building within and outside scientific structures. Many defenders are fisherpeople, coastal communities, or groups directly affected by changes in climate and ocean health. Beyond advocating for sustainable and generative oceanic futures, they also fight to overcome political resistance and funding barriers. Ocean defenders, like land defenders, face challenging or dangerous obstacles while pushing for local and global ocean preservation. Ocean science and policy clearly needs collaborative approaches that bring multiple knowledge systems forward while prioritizing those most impacted by climate change, pollution, and other threats to both marine habitats and coastal livelihoods.  *Ocean-defending artisanal fishers and their supporters in South Africa celebrate upon receiving the news that Shell’s permit to conduct a seismic survey on the Wild Coast had been set aside by the Makhanda High Court, in September 2022. Photo credit: Taryn Pereira. Source: <https://oceandefendersproject.org/case-study/no-to-seismic-surveys/>* **More on capacity initiatives, knowledge gaps, and ocean defenders:** - Guilhon, M., M. Vierros, H. Harden-Davies, D. Amon, S. Cambronero-Solano, C. Gaebel, K. Hassanali, V. Lopes, A. McCarthy, A. Polejack, G. Sant, J.S. Veiga, A. Sekinairai, and S. Talma. (2025). [Measuring the success of ocean capacity initiatives.](https://doi.org/10.5670/oceanog.2025.122) *Oceanography* 38(1). - Saba, A.O., I.O. Elegbede, J.K. Ansong, V.O. Eyo, P.E. Akpan, T.O. Sogbanmu, M.F. Akinwunmi, N. Merolyne, A.H. Mohamed, O.A. Nubi, and A.O. Lawal-Are. 2025. [Building ocean science capacity in Africa: Impacts and challenges.](https://doi.org/10.5670/oceanog.2025.133) *Oceanography* 38(1). - Behl, M., Cooper, S., Garza, C., Kolesar, S. E., Legg, S., Lewis, J. C., White, L., & Jones, B. (2021). [Changing the culture of coastal, ocean, and marine sciences: strategies for individual and collective actions.](https://www.jstor.org/stable/27051390) *Oceanography*, *34*(3), 53–60. - The Ocean Defenders Project (2025). [Ocean Defenders: Protectors of our ocean environment and human rights.](https://oceandefendersproject.org/project-publications/ocean-defenders-protectors-of-our-ocean-environment-and-human-rights/) The Peopled Seas Initiative, Vancouver, Canada. - Belhabib, D. (2021) [Ocean science and advocacy work better when decolonized.](https://doi.org/10.1038/s41559-021-01477-1) *Nat Ecol Evol* 5, 709–710. - Kennedy, R. and Rotjan, R. (2023). [Mind the gap: comparing exploration effort with global biodiversity patterns and climate projects to determine ocean areas with greatest exploration needs.](https://doi.org/10.3389/fmars.2023.1219799) *Front. Mar. Sci.* (10). - Bell, K.L.C, Johannes, K.N., Kennedy, B.R.C., & Poulton, S.E. (2025) [How little we’ve seen: A visual coverage estimate of the deep seafloor.](https://www.science.org/doi/10.1126/sciadv.adp8602) *Science Advances*, Vol 11, Issue 19. ### 1.4 Ocean Datasets for Investigations and Storytelling Ocean datasets play a crucial role in both scientific investigations and storytelling by providing evidence-based insights into the health and dynamics of our global ocean. These datasets help scientists better understand the ocean, but beyond research, they serve an important role in ocean-related investigations. Ocean datasets can support communities, journalists, and activists in raising data-backed awareness about critical marine and environmental justice issues. By sharing this data in accessible ways, oceanographic narratives can amplify the voices of coastal communities, engage the public, and inspire action in support of more regenerative ocean futures. Numerous well-resourced journalistic and forensic organizations use ocean data to support their stories or reporting, such as Forensic Architecture, LIMINAL, Forensis, Earshot, Border Forensics, and others. In this guide, we will demonstrate how you can access datasets and conduct your own oceanic investigations. By the end, you will be able to illustrate a well-defined ocean or climate question using publicly available oceanographic datasets and media collections, which will enhance your evidence-based and visually engaging story. ## 2. Diversity of Data Types The following sub-sections serve as a collection of key ocean data types, organized by how they are collected and what they reveal. Each broad data source type (i.e., the overarching technique used to gather certain kinds of ocean data) begins with a bit of history and trade-offs, and then is further broken down into specific data products. For each data product, we share use cases, data collection methods, key considerations, and some databases (often from U.S. and European agencies), followed by resources to learn more. This structure is designed to help you understand how each dataset is produced, grasp the significance of the data, and know where to go for deeper investigations or analyses. There are many data types presented below which are organized in these four broad categories: *in situ* sensors; deep ocean observation, exploration, and research; mapping; and, satellite remote sensing. See Section 3 to review an example case study which demonstrates how some of these datasets may be used to support an investigative ocean and climate story.  *Illustration of a range of ocean sensor platforms—ships, profiling drifters, gliders, moored buoys, and satellites. Source: <https://www.ecmwf.int/en/about/media-centre/news/2021/world-meteorological-day-focuses-role-ocean-weather-and-climate>* ### 2.1 *In Situ* Sensors *In situ* sensing refers to the direct measurement of ocean properties using instruments that are physically located within the water. Sensors for measuring temperature, salinity, pressure, currents, and of biochemical concentrations are deployed on a range of platforms with various advantages and drawbacks. While satellites provide broad spatial coverage of the ocean’s surface, *in situ* platforms are essential for monitoring the ocean’s interior, tracking coastal change, and measuring water properties that cannot be detected remotely. - **History:** - Sailors have used thermometers to measure ocean temperature since at least as early as Captain James Cook’s [1772 voyage](https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/rog.20022) to the Antarctic Circle (another example of colonial science forming the foreground to Western ocean science). - The Nansen bottle (1896) and later the Niskin bottle (1966) enabled the capture of water samples at specific depths, which could then be pulled up and tested for temperature and salinity on the ship. - The first bathythermograph was developed in 1938 and featured a temperature sensor on a wire which recorded a temperature profile as it was lowered into the ocean. This was used by the US Navy in WWII to improve sonar accuracy, since temperature layers alter acoustic propagation. - Today, there are thousands of advanced sensors across the world’s oceans which transmit readings in real time via satellite. - **Trade-offs:** - *In situ* sensors can only measure the ocean properties at their exact location so great consideration is taken in their placement and timing. - Powering the instruments is a constant challenge and factors into decisions about sampling frequency. - The extreme pressure in the deep ocean limits the operating depth of some sensors. - Harsh operating conditions limit the lifespan of sensors and necessitates regular maintenance/replacement, often in remote locations at high costs. This leads to less accessible areas being undersampled. #### 2.1.1 Moorings and Fixed Platforms  *Example of a coastal data mooring. Source: Baily et al., 2019, <https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2019.00180/full>* - **About:** - A range of platforms anchored in place, collecting time-series data at a fixed location. - Used in the deep ocean (e.g., the TAO array across the Pacific Ocean), the continental shelf (e.g., NOAA’s National Data Buoy Network), and at the coastline (e.g., tide gauges). - **Use cases:** - Long-term climate monitoring of ocean heat content and circulation. - Data inputs for forecast models, improves accuracy of ocean and weather predictions. - Early warning systems for tsunamis and hurricane storm surge. Tracking tidal heights and local sea level rise. - Water quality monitoring for pollutants, algal blooms, and hypoxia. - **Data collection:** - Sensor packages measure temperature, salinity, pressure, biochemistry, and more. - Some moorings have Acoustic Doppler Current Profilers (ADCPs) to measure water current velocities throughout the water column. - **Key considerations:** - Data today is mostly broadcasted in near-real time, but there are some platforms that require physical retrieval before the data is downloaded. - Spatial coverage is extremely limited and focused around a small set of nations. - The diversity of databases and data types can pose a challenge for accessing and working with the data. - **Data sources:** - [US NOAA National Data Buoy Center (NDBC)](https://www.ndbc.noaa.gov/) - [EU Copernicus Marine Service In Situ dashboard](https://marineinsitu.eu/dashboard/) - [Global Tropical Moored Buoy Array](https://www.pmel.noaa.gov/gtmba/) - **Resources to learn more:** - [WHOI - Moorings & Buoys](https://www.whoi.edu/what-we-do/explore/instruments/instruments-moorings-buoys/) - [Ocean Observatories Initiative](https://oceanobservatories.org/) #### 2.1.2 Drifters and Floats  *Map of Argo float locations. Source: <https://argo.ucsd.edu/about/status/>* - **About:** - Unanchored and unpropelled instruments that drift with the currents and take ocean measurements. - Drifters stay at the surface and provide information about surface conditions, and surface currents are calculated from their GPS trajectory. - Floats profile the water column by adjusting their buoyancy to move up and down. The Argo program is the largest and most significant, with over 4,000 Argo floats profiling the world’s oceans. - Capable of providing global coverage at lower cost than moored sensors, especially in remote open-ocean regions. - **Use cases:** - Drifters: Mapping near-surface currents and tracking pollutants and marine debris transport. - Floats: Measuring subsurface temperature and salinity for climate studies. Some Argo floats also have biochemical sensors. - **Data collection:** - Drifters: GPS-tracked, measure SST, pressure, sometimes salinity, sometimes waves. - Argo floats: Profile down to 2,000 m every 10 days, transmitting data via satellite. - **Key considerations:** - Drifters and floats are always moving, so you can’t get a clean timeseries for a single location like you can with moorings. Additionally, Argo floats only take one profile every 10 days in order to preserve battery life. - Argo floats don’t generally operate near the coast on the continental shelf. - Some drifters/floats lack real-time telemetry (data transmission). - **Data sources:** - [Global Drifter Program](https://www.aoml.noaa.gov/phod/gdp/) - [Argo Program](https://argo.ucsd.edu/) - [Copernicus Marine Service drifters](https://data.marine.copernicus.eu/products?facets=featureTypes%7ETrajectory) - [SOCCOM Biogeochemical Floats](https://soccom.org/) - **Resources to learn more:** - [About Argo](https://argo.ucsd.edu/about/) #### 2.1.4 Autonomous Vehicles - ASVs and Gliders  *Illustration of a glider’s sawtooth propulsion pattern. Source: <https://earthzine.org/going-deep-to-go-far-how-dive-depth-impacts-seaglider-range/>* - **About:** - Autonomous Surface Vehicles (ASVs) and gliders are robotic platforms that enable long-duration, energy-efficient monitoring over vast areas. - Bridging the gap between targeted, expensive, ship-based measurements and low-cost, but uncontrolled, drifters/floats. - **Use cases:** - Significant overlap with the use cases for moored sensors and drifters/floats. - Targeted measurements in dangerous conditions like hurricanes. - Mapping surveys, autonomous of ship or in tandem with other vehicles. - **Data collection:** - Autonomous Surface Vehicles (ASVs) use solar panels, wind, or waves as a power source to supplement and recharge their batteries. - Gliders are underwater vehicles that create propulsion by adjusting their buoyancy and gliding horizontally while sinking/rising (similar to an airplane). This enables longer battery range than propellers or thrusters. - **Key considerations:** - Gliders and ASVs are often used in targeted studies rather than continuous global monitoring and thus have lower data availability. - Shorter mission durations than moorings or drifters/floats. - **Data sources:** - [NOAA Glider Data Assembly Center](https://gliders.ioos.us/) - [OceanGliders](https://www.oceangliders.org/) - **Resources to learn more:** - [National Oceanography Centre UK - Gliders](https://noc.ac.uk/facilities/marine-autonomous-robotic-systems/gliders) ### 2.2 Deep Ocean Observation, Exploration, and Research Deep ocean science is typically conducted to observe long-term changes at specific seafloor sites, to explore marine habitats that are unknown to science, and to conduct applied or experimental research on focused environmental questions. A range of deep ocean data collection tools include platforms or landers, cabled observatories, and deep submergence systems—such as human-occupied vehicles (HOVs), remotely-occupied vehicles (ROVs), and autonomous underwater vehicles (AUVs).  *Human-occupied vehicle (HOV) 'Alvin' being recovered in 2024 during an expedition to the East Pacific Rise hydrothermal vent fields. Photo credit: Mae Lubetkin* - **History** - 1872–1876: The *HMS Challenger* expedition, a milestone for deep ocean science but deeply tied to imperialism - Mid-20th century: Cold War military priorities further developed submersible vehicle capabilities, leading to *Trieste*’s 1960 Mariana Trench dive - 1964 onward: HOVs expanded access to the deep ocean, e.g., *Alvin* (US), *Nautile* (France), *Shinkai* (Japan), and *Mir* (Russia) - 1980s–2000s: ROVs and AUVs developed by industry (oil, mining, and defense) and scientific institutions, in parallel - 2000s–present: Cabled observatories (e.g., Ocean Networks Canada, DONET in Japan), public research campaigns (e.g., NOAA, IFREMER), and oceanographic instruments expanded reach and scope. - Today: Many regions still face barriers to participation and funding in deep ocean science (as outlined in the introduction). Meanwhile, deep submergence science in wealthy nations increasingly utilizes AI, autonomous systems, 4K and 3-D imaging techniques. - **Trade-offs:** - Provides direct access to deep ocean environments which are inaccessible by surface vessels or remote sensing - High spatial and contextual resolution: can capture detailed imagery, samples, and detailed *in situ* measurements - Resource-intensive: operations usually require ships, launch/recovery teams, and specialized personnel - Limited temporal and spatial coverage: data collection is episodic, site-specific, and dependent on expedition funding, schedules, and weather conditions at-sea - High costs and technical barriers mean deep ocean science is dominated by a few well-funded institutions or nations, with limited global access - Colonial legacies persist in relation to who sets research agendas, who makes funding decisions, and who benefits from collected data #### 2.2.1 Deep Submergence Systems (HOVs, ROVs, AUVs) - **About:** - Vehicles that operate in the deep ocean water column or along the seafloor, including: - Human-occupied vehicles (HOVs): carry scientists directly, typically 1-3 observers and a pilot - Remotely operated vehicles (ROVs): tethered and piloted from a surface vessel like a ship - Autonomous underwater vehicles (AUVs): untethered and pre-programmed - These systems can operate from hours to days and are depth-rated around 4000-6000 m, but some may reach full ocean depths (11 km) and others may work well in shallower waters. - **Use cases:** - High-resolution visual surveys - Precise targeted sampling with environmental context and imagery at diverse environments including hydrothermal vents, methane seeps, cold-water coral habitats, and others - Biogeographic habitat mapping - Wreck exploration and infrastructure inspection - Imagery of deep ocean environments can support visual storytelling, public engagement, and education - **Data collection:** - Data is streamed directly to the support vessel for tethered operations - While for untethered submersibles (HOVs and AUVs) most data is retrieved when the vehicle is recovered - All physical samples are retrieved and processed upon vehicle recovery - **Key considerations:** - Requires experienced pilots and operational support (expensive) - AUVs need detailed mission planning (mission failure could lead to vehicle loss) - Navigation and environmental risks must be managed carefully - **Data sources:** - SeaDataNet - [EU research vessel data](https://csr.seadatanet.org/), including cruise summary reports and more - EuroFleets - European initiative to compile [EU research cruise data](https://www.eurofleets.eu/data/) - Rolling Deck to Repository - [US research vessel data](https://www.rvdata.us/data), including: expedition summary, shiptrack navigation, scientific sampling event log, post-processed data - JAMSTEC Databases - [Japan research vessel data](https://www.jamstec.go.jp/e/database/), including: HOV *Shinkai 6500* and ROV *Kaiko* mission data, cruise reports, and dive logs - **Resources to learn more:** - [Woods Hole Oceanographic Institution - National Deep Submergence Facility](https://ndsf.whoi.edu/) - [Ocean Exploration Trust - Science and Technology](https://nautiluslive.org/science-tech)  *An imaging elevator equipped with two camera systems, lights, battery packs, and crates to store additional sampling tools to be used by an HOV during a dive in the same region. Source: Woods Hole Oceanographic Institution* #### 2.2.2 Landers and Elevators - **About:** - Landers are relatively simple systems that descend to the seafloor and remain stationary for the duration of their deployment. - They are sometimes referred to as 'elevators' since they descend to the seafloor then ascend back to the surface - There are no people on landers, but they typically carry sensors, cameras, samplers, and other instruments - Depending on power supply and scientific goals, they can spend hours to sometimes months on the seafloor - **Use cases:** - Collecting environmental data (e.g., conductivity, temperature, pH, oxygen) - Capturing imagery of habitats or operations - Deploying baited cameras or traps to study biodiversity - Using the platform to carry additional gear or instruments to the seafloor that a deep submergence system could not transport on its own due to space limitations - **Data collection:** - Typically data is retrieved when the lander is recovered back on deck - Some landers will transmit data acoustically or remotely from the seafloor - The frequency that imagery or other data are collected is pre-programmed before deployment - **Key considerations:** - Requires careful site selection, recovery planning, and often ship time - Currents can impact the intended landing location on the seafloor, sometimes drifting the platform or lander far off-site - Limited in capabilities, not as advanced as deep submergence vehicles, but also much cheaper and easier to custom build - **Data sources:** - Deep ocean lander data (e.g., imagery and environmental sensor data) would be found in the same databases and repositories listed in section 2.2.1 Deep Submergence Systems (HOVs, ROVs, AUVs). - **Resources to learn more:** - [Schmidt Ocean Institute - Elevators and Landers](https://schmidtocean.org/technology/elevators-landers/) - [The Deep Autonomous Profiler (DAP), a Platform for Hadal Profiling and Water Sample Collection (Muir et al., 2021)](https://journals.ametsoc.org/view/journals/atot/38/10/JTECH-D-20-0139.1.xml) - [Lander Lab: Technologies, Strategies and Use of Ocean Landers (Hardy, 2022)](https://magazines.marinelink.com/Magazines/MarineTechnology/202201/content/technologies-strategies-landers-594271)  *Map of Ocean Networks Canada NEPTUNE and VENUS Observatories near Vancouver Island, Canada. Each orange square represents a node or station along the cabled observatory where instruments or sensors are mounted. Source: <https://www.oceannetworks.ca/>* #### 2.2.3 Cabled Observatories - **About:** - Mostly permanent, wired infrastructure on the seafloor that transmit real-time power and data via fiber optic cables connected to shore stations - Similar in some ways to the data collection tools described in section 2.1 *In Situ* Sensors, but these networks are fixed in location and networked - People do not visit these observatories, instead they support a wide range of sensors (e.g., temperature, pressure, seismometers, hydrophones, cameras, samplers) - Can integrate with ROVs or AUV docking stations, and are also typically maintained and serviced by ROVs - Designed for continuous, high-frequency monitoring of deep ocean processes across years or decades - They connect highly diverse environments from hydrothermal vent regions to abyssal plains and continental shelves - **Use cases:** - Long-term and consistent monitoring of geophysical activity (e.g., earthquakes, hydrothermal vents) - Real-time data for early warning systems (e.g., tsunamis, gas releases) - To study oceanographic processes (e.g., currents, biogeochemical fluxes, and ecosystem change) - Supports public engagement and education through livestreams - **Data collection:** - Real-time data is livestreamed to shore stations and then available via online portals - Most are operated by national or international research infrastructures - **Key considerations:** - Extremely costly, high maintenance needs (ROVs are often used for annual servicing) - Site selection is key since they are fixed installations - **Data sources:** - Ocean Networks Canada - [Oceans 3.0 Data Portal](https://data.oceannetworks.ca/), including all datasets, dashboards, and visualizers (more info on [ONC data](https://www.oceannetworks.ca/data/)) - US Ocean Observatories Initiative - [OOI Data Portal](https://oceanobservatories.org/data-portal/), includes cable-linked arrays on East and West Coasts and deep Pacific - EU [EMSO ERIC Data Portal](https://data.emso.eu/home), real-time and archived data, tools and research environment to investigate seafloor observatories across European margins - **Resources to learn more:** - [Ocean Observatories Initiative](https://oceanobservatories.org/observatories/) - arrays, infrastructure, instruments - [Ocean Networks Canada](https://www.oceannetworks.ca/observatories/) - observatories - [Interactive map of ONC locations](https://www.arcgis.com/home/webmap/viewer.html?webmap=fcea4e5f087f41c58bcc5e51b13fffa1&extent=-158.3094,39.6681,-29.8133,75.4182) - [Regional Cabled Observatories](https://www.whoi.edu/what-we-do/explore/ocean-observatories/about-ocean-observatories/types-of-observatories/regional-cabled-observatories/) - summary by Woods Hole Oceanographic Institution ### 2.3 Mapping Marine hydrography or ocean scientific mapping involves the creation of high-resolution representations of the seafloor, water column, and other associated features or phenomena (e.g., fish migrations, vents or seeps bubbling up) using vessel-based sonar, autonomous vehicles, acoustic or optical tools. It is a type of remote sensing since the mapping instrument is not on the seafloor. Unlike satellite remote sensing, which observes only the ocean surface from space, hydrographic mapping is conducted from platforms within or on the ocean surface. These mapping systems can resolve fine-scale topography (seafloor bathymetry), subsurface geologic layers, water column imaging, and habitat mapping that integrates both physical and biological data. Laser and 3-D reconstruction are other forms of high-resolution mapping. - **History:** - 1870s–1900s: Early bathymetric charts created using lead lines, linked to colonial navigation and maritime claims as well as scientific knowledge creation - 1920s–1940s: Echo sounding developed for military and commercial navigation, later repurposed for seafloor mapping - 1950s–1970s: Multibeam sonar developed, enabling wider swath coverage (i.e. can map wider seafloor area, not just single points) and broader seafloor topography or bathymetry mapping - 2010s–present: Autonomous vehicles, numerous specialized sonar systems, and 3-D photogrammetry advance deep mapping capabilities - Today: Mapping remains uneven globally—nations with limited funding or access to ships and processing capacity are underrepresented and do not have detailed seafloor maps of the their waters - **Trade-offs:** - High-resolution, fine-scale maps of seafloor and water column features - Enables geologic, biologic, and habitat-based spatial analysis - Requires significant ship time, technical expertise, and post-processing - Data coverage is patchy, most of the seafloor remains unmapped - High cost and national interests impact where mapping occurs and who benefits from the data  *Bathymetric mapping using a hull-mounted multibeam sonar system. Black lines indicate the ship’s track, while the coloration represents depth differences (red is shallow, purple is deep) used for visualizing the bathymetric or topographic features of the seafloor. Source: https://www.worldofitech.com/mapping-the-ocean-floor-water-bathymetry-data/>* #### 2.3.1 Bathymetric Mapping - **About:** - Measurement and charting of the depth and shape of the seafloor - Typically uses sonar-based systems (e.g., single-beam or multibeam echosounders), mounted on ships, AUVs, or towed platforms - Short-range systems (e.g., ROV-mounted sonar) provide highly detailed data over small areas (centimeters in resolution), while medium-range systems (e.g., hull-mounted multibeam on ships or AUVs) cover much larger swaths with lower resolution - **Use cases:** - Mapping underwater topography and geological features - Planning submersible dives and identifying hazards - Supporting infrastructure projects like cables or offshore wind farms - Creating base maps for habitat mapping or biogeographic studies (i.e. understanding what marine life lives where and how their habitats are linked to geologic features as well as currents and physical oceanographic phenomena) - **Data collection:** - By research vessels or autonomous vehicles using sonar systems - Key manufacturers include Kongsberg, Teledyne, R2Sonic, and Edgetech - Data is processed using specialized hydrographic software (e.g., QPS Qimera, CARIS, MB-System) - **Key considerations:** - Requires calibration (e.g., sound speed profiles) and correction for vessel motion - Deep ocean mapping can be slow and resource-intensive - Interpretation of raw bathymetry data requires trained analysts and geospatial tools, it is not yet fully automated - **Data sources:** - European Marine Observation and Data Network - [EMODnet Bathymetry](https://emodnet.ec.europa.eu/en/bathymetry), multibeam datasets and other maps with a built in visualizer - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access), archive of US and global bathymetric surveys with visualizer - General Bathymetric Chart of the Oceans - [GEBCO Gridded Bathymetry Data](https://www.gebco.net/data-products/gridded-bathymetry-data), global map interface of compiled bathymetry - Global Multi-Resolution Topography - [GMRT data](https://www.gmrt.org/), global compilation of multibeam data (includes graphical map tool) - **Resources to learn more:** - [NOAA - What is bathymetry?](https://oceanservice.noaa.gov/facts/bathymetry.html) - [Seabed 2030](https://seabed2030.org/) – global initiative to map the entire seafloor by 2030 - [Using open-access mapping interfaces to advance deep ocean understanding](https://link.springer.com/article/10.1007/s40012-025-00410-2) (Johannes, 2025) - [GeoMapApp](https://www.geomapapp.org/) - free map-based application for browsing, visualizing and analyzing a diverse suite of curated global and regional geoscience data sets  *Plumes of bubbles emanating from the seafloor, indicating that there were methane gas seep ecosystems in this region. Sound waves reflect strongly off the gas bubbles and are visible in the water column data. Source: <https://nautiluslive.org/blog/2018/08/08/more-just-bathymetry-seafloor-mapping-tool-exploration>* #### 2.3.2 Water Column Mapping - **About:** - Acoustic systems—usually multibeam echosounders or special water column sonars—to detect and visualize features suspended in the ocean between the surface and the seafloor - Key for detecting gas plumes, biological layers (schools of fish or migrations in the twilight zone), suspended sediments, etc. - **Use cases:** - Observing midwater scattering layers (biological migrations) - Detecting hydrothermal plumes amd tracking gas plumes from methane seeps - **Data collection:** - Most multibeam systems include water column data modes, so it is often collected in tandem with bathymetry - From ships, AUVs, and ROVs - Data must be interpreted alongside oceanographic profiles (e.g., CTD casts) and often requires manual cleaning to reduce noise - **Key considerations:** - Processing and interpreting water column data is a bit tedious not yet standardized - Detection is sensitive to sonar frequency, range, and sea conditions - Validation with ground-truth sampling (e.g., bottle casts, nets, sensors) is helpful - **Data sources:** - European Marine Observation and Data Network - [EMODnet Physics](https://emodnet.ec.europa.eu/en/physics), includes some water column data layers - NOAA National Centers for Environmental Information - [NCEI water column sonar data](https://www.ncei.noaa.gov/maps/water-column-sonar/) - **Resources to learn more:** - [OET - More than just Bathymetry](https://nautiluslive.org/blog/2018/08/08/more-just-bathymetry-seafloor-mapping-tool-exploration) - Seafloor Mapping as a Tool for Exploration - [Migration in the Ocean Twilight Zone](https://twilightzone.whoi.edu/explore-the-otz/migration/) - often monitored with water column data #### 2.3.3 Seafloor Backscatter - **About:** - Analyzing the intensity of sound that is reflected or ‘scattered back’ from the seafloor when using sonar systems - Provides information about seafloor texture, hardness, and composition (e.g., sand, rock, mud) - Often conducted simultaneously with bathymetric mapping during ship-based or AUV surveys - **Use cases:** - Seafloor habitat classification or substrate mapping - Detecting anthropogenic objects or features (e.g., cables, wrecks) - Complements bathymetry for geologic or habitat models - **Data collection:** - Similar to bathymetry and water column mapping, backscatter data is collected using the same sonar systems and processed using similar software - **Key considerations:** - Requires calibration and post-processing to produce usable mosaics - Interpretation of sediment type from backscatter typically should be verified by ground-truth sampling (e.g., grabs, cores with ROVs or HOVs) - **Data sources:** - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access) - **Resources to learn more:** - NOAA - [How does backscatter help us understand the sea floor?](https://oceanservice.noaa.gov/facts/backscatter.html)  *Sediment layers seen in the sub-bottom profiler data collected in 2021 at the New England and Corner Rise Seamounts expedition on the NOAA Ship 'Okeanos Explorer'. Source: <https://oceanexplorer.noaa.gov/technology/sub-bottom-profiler/sub-bottom-profiler.html>* #### 2.3.4 Sub-bottom Profiling - **About:** - Uses low-frequency acoustic pulses to penetrate below the seabed and image sediment layers or other buried geologic features - Reveals vertical structures beneath the seafloor - Typically deployed from research vessels or towed systems - **Use cases:** - Studying sedimentation and geological processes - Locating subseafloor gas pockets or archaeological sites - For infrastructure planning or hazard assessment (e.g., submarine landslides) - **Data collection:** - Chirp profilers (high resolution, shallow penetration) and boomer/sparker systems (deeper penetration) are used - Operated from vessels with sonar equipment often collected simultaneously while collecting bathymetry and other mapping data, even if the sonar systems are different - **Key considerations:** - Resolution and penetration are inversely related (deeper = less detail) - Can be noisy and hard to interpret without ground-truthing (e.g., sediment cores) - **Data sources:** - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access) - European Marine Observation and Data Network - [EMODnet Geology](https://emodnet.ec.europa.eu/en/geology), includes sub-bottom and other forms of seafloor geological data - **Resources to learn more:** - NOAA - [Sub-Bottom Profiler](https://oceanexplorer.noaa.gov/technology/sub-bottom-profiler/sub-bottom-profiler.html)  *3-D reconstructed seafloor lava flows and hydrothermal vent field from the East Pacific Rise. This 3-D model was produced using downward facing video imagery and photogrammetry techniques. Credit: Mae Lubetkin* #### 2.3.5 Photogrammetry and 3-D Reconstruction - **About:** - Stitching together overlapping images or video frames from subsea camera systems often mounted on ROVs or AUVs - To create detailed mosaics or 3-D models of seafloor features - Uses optical data, offering true-color, high-resolution imagery (unlike acoustic mapping techniques described above) - **Use cases:** - Mapping hydrothermal vent fields, coral reefs, archaeological sites, etc. - Change detection in dynamic environments (e.g., volcanic or vent habitats, biological growth or loss) - Public engagement and educational tools - **Data collection:** - Collected by vehicle-mounted cameras with precise navigation and positioning - Software like Agisoft Metashape or custom photogrammetry pipelines are used for processing (which is easier now than ever before, becoming much more common in ocean sciences) - **Key considerations:** - Requires good lighting and water clarity - Processing is computationally intensive, and vehicle navigation data helps with plotting 3-D reconstructions onto broader bathymetric maps - Can be limited to small survey areas due to time constraints and battery limitations - **Data sources:** - Monterey Bay Aquarium Research Institute - [MBARI Sketchfab](https://sketchfab.com/mbari) - 3-D models of seafloor sites can be found in academic papers or at individual institutions or government agencies data repositories - **Resources to learn more:** - [Seafloor Futures](https://garden.ocean-archive.org/seafloor-futures/) (Lubetkin, 2024) - [Realtime Underwater Modeling and Immersion](https://nautiluslive.org/tech/realtime-underwater-modeling-and-immersion) - Ocean Exploration Trust - [Underwater 3-D Reconstruction from Video or Still Imagery: Matisse and 3-D Metrics Processing and Exploitation Software](https://www.mdpi.com/2077-1312/11/5/985) (Arnaubec et al., 2023) - [Seeing the Sea in 3-D](https://schmidtocean.org/cruise-log-post/seeing-the-sea-in-3d/) - Schmidt Ocean Institute ### 2.4 Satellite Remote Sensing Satellite data provides the most familiar, spatially complete picture of the ocean. This bird’s-eye perspective is invaluable for understanding large-scale phenomena like currents, sea surface temperature patterns, and phytoplankton blooms, providing visual evidence that can enhance storytelling. However, there are unique considerations since, unlike the use of satellite imagery on land, most of our understanding of the ocean does not come from the visual spectrum. In this section, we’ll introduce three of the most important types of satellite ocean data and discuss the use cases for each. - **History:** - 1978: NASA launched Seasat, the first satellite designed for ocean research. - Significant expansion in the 1990s with missions including TOPEX/Poseidon (ocean altimetry), AVHRR (high-resolution sea surface temperature), and SeaWiFS (ocean biology). - Modern constellations are operated by NASA, NOAA, ESA, EUMETSAT, CNES, ISRO, and others. - **Trade-offs:** - Excellent spatial coverage that’s impossible to achieve with ships or buoys. - Very high costs, these platforms are operated by government agencies. - Only seeing the very surface of the ocean, no subsurface data. - Limited horizontal resolution (spatial detail) and temporal resolution (orbital repeat time).  *Gulf of Mexico SST on a cloud-free data. Source: <https://marine.rutgers.edu/cool/data/satellites/imagery/>* #### 2.4.1 Radiometry - Sea Surface Temperature (SST) - **About:** - Sea surface temperature (SST) is the oldest and most extensive application of satellite oceanography. - **Use cases:** - Tracking climate change, El Niño, and marine heat waves. - Key input for weather models (e.g., very important for hurricane forecasting). - Mapping ocean eddies, currents, and upwelling, which are critical to fisheries. - **Data collection:** - Two separate types of sensors measure SST: Infrared and microwave. - IR sensors have higher spatial resolution ([1-4 km](https://coastwatch.noaa.gov/cwn/product-families/sea-surface-temperature.html)) and finer temporal coverage but cannot “see” through clouds, which block over 70% of the ocean at any given time. - Microwave sensors can see through most non-precipitating clouds but have a lower spatial resolution (about 25 km) and don’t work near the coastline. - Measures temperature of the top ~1 mm of the ocean - Blended products: Combine multiple sensors for better coverage (e.g., GHRSST L4)  *Example SST data at different processing levels. (Merchant et al. 2019): <https://www.nature.com/articles/s41597-019-0236-x>* - **Key considerations:** - Make yourself aware of the different processing levels when accessing data. Level 4 (L4) will be the easiest to work with but may not be fully accurate. - L2: Data along the original orbital track. - L3: Gridded data, sometimes averaged over time. - L4: Cloud-free, gaps are filled by various methods depending on the source. - Temporal resolution depends on the satellite orbit. There are good options that blend multiple satellites. - **Data sources:** - [EU Copernicus Marine Service](https://marine.copernicus.eu/) - [US NASA Physical Oceanography DAAC](https://podaac.jpl.nasa.gov/) - [NOAA CoastWatch](https://coastwatch.noaa.gov/cw_html/cwViewer.html) - Graphical Interface - **Resources to learn more:** - [Group for High Resolution Sea Surface Temperature (GHRSST)](https://www.ghrsst.org/ghrsst-data-services/for-sst-data-users/) #### 2.4.2 Radar Altimetry - Sea Surface Height (SSH) - **About:** - Measures ocean surface height by sending radio pulses and measuring return time. - SSH can tell us the strength of large scale currents like the Gulf Stream, as the slope of the sea surface is used to calculate the “geostrophic current”. - **Use cases:** - Key to understanding ocean circulation and long-term sea level rise. - **Data collection:** - Radar altimeters on satellites measure SSH directly (e.g., Jason-3, Sentinel-6) and then the geostrophic currents are calculated in a post-processing step. - Spatial resolution is significantly worse than SST (25+ km) - The recent SWOT satellite is a new type of altimeter with much higher resolution but has very limited coverage since there is only one currently in orbit. - **Key considerations:** - SSH is useful for large-scale ocean currents but not coastal tidal currents. - Similar to SST, be careful about processing level and look for re-gridded datasets. - Can generally see through clouds, so gaps are not a significant issue. - **Data sources:** - [EU Copernicus Marine Service](https://data.marine.copernicus.eu/products?facets=mainVariables%7ESea+surface+height--sources%7ESatellite+observations) and [Aviso](https://www.aviso.altimetry.fr/en/data/products/sea-surface-height-products/global.html) - [US NASA PODAAC](https://podaac.jpl.nasa.gov/NASA-SSH) - [Copernicus MyOcean Pro](https://data.marine.copernicus.eu/viewer/expert) and [Aviso](https://seewater.aviso.altimetry.fr/) - Graphical Interfaces - **Resources to learn more:** - [NASA JPL - What is Ocean Surface Topography?](https://podaac.jpl.nasa.gov/OceanSurfaceTopography)  *Global map of marine Chlorophyll concentration. Source: <https://sos.noaa.gov/catalog/datasets/biosphere-marine-chlorophyll-concentration/>* #### 2.4.3 Optical - “Ocean Color” - **About:** - Ocean color sensors measure the reflectance of sunlight from the ocean surface to infer biological and chemical properties, such as algal concentration, suspended sediments, and water clarity. - **Use cases:** - Tracking phytoplankton blooms and changes in marine ecosystems. - Useful for monitoring water quality, including coastal sediment and oil spills. - **Data collection:** - Sensors measure light reflected from the ocean at different wavelengths (e.g., MODIS, VIIRS, Sentinel-3 OLCI) and then apply algorithms in order to calculate variables such as Chlorophyll-a concentration. - **Key considerations:** - Ocean color data is significantly affected by cloud cover, aerosols, and atmospheric correction errors. - **Data sources:** - [EU Copernicus Marine Service](https://data.marine.copernicus.eu/products?facets=mainVariables%7EPlankton--sources%7ESatellite+observations) - [US NASA Ocean Color Web](https://oceancolor.gsfc.nasa.gov/data/find-data/) - [NASA Worldview](https://worldview.earthdata.nasa.gov/) - Graphical Interface - **Resources to learn more:** - [IOCCG (International Ocean-Color Coordinating Group)](https://ioccg.org/) ### 2.5 Additional databases and scientific support The four sub-sections above (*In Situ* Sensors; Deep Ocean Observation, Exploration, and Research Systems; Mapping; Satellite Remote Sensing) cover the main areas of ocean scientific data types and collection methods. There are some datasets that are not discussed in this guide since they are likely less useful for investigative storytelling or require technical skills to access and interpret the data. Below are some additional databases and information on contacting scientists to support your investigation. While in section 3, we outline a case study using real data to tell an ocean story. #### 2.5.1 Additional databases, collections, and visualizers Sites that either did not fit into one of the sub-sections above, or that contain information which is generated after scientific studies occur: - [PANGAEA](https://pangaea.de/) - data publisher for earth and environmental science (across disciplines) - [International Seabed Authority DeepData](https://www.isa.org.jm/deepdata-database/) - database hosting all data related to international deep-seabed activities, particularly those collected by contractors (i.e. nations or entities) during their exploration activities and other relevant environmental and resources-related data. Includes a dashboard and map to search for basic stats and information about what contractors have done during deep seabed mining exploration cruises. - [Marine Geoscience Data System](https://www.notion.so/Ocean-Datasets-for-Investigations-1caf92221af780c68873c2aecf9b3479?pvs=21) - geology and geophysical research data across collections - [USGS Earthquake Hazards Program](https://www.usgs.gov/programs/earthquake-hazards/earthquakes) - interactive map with magnitudes and additional information (earthquakes can occur on land and in the ocean) - [WoRMS – World Register of Marine Species](https://www.marinespecies.org/) - comprehensive taxonomic list of marine organism names - [OBIS – Ocean Biodiversity Information System](https://obis.org/) - global open-access data and information on marine biodiversity - [Windy](https://www.windy.com/) - animated weather maps, radar, waves and spot forecasts #### 2.5.2 Scientific support All of the datasets and databases we outlined above are free and open to the public. We hope that we outlined enough context and links to user-friendly platforms to access the data so that you feel empowered to conduct your own investigations with ocean datasets. That said, some data might be more challenging to work with depending on prior experience and computing skills, among other factors. When in doubt, you can always contact an ocean scientist to ask questions or seek support. Depending on your investigation or story, you will need to contact a specific type of ocean scientist since each has their own specialty. You can start by searching for and contacting scientists at nearby universities or research institutes. **Ocean scientists and their specializations:** - **Physical Oceanographers -** Study ocean currents, tides, waves, and ocean-atmosphere interactions. They can help explain phenomena like sea level rise or how ocean circulation affects weather and climate. - **Chemical Oceanographers -** Focus on the chemical composition of seawater and how it changes over time. Useful for stories involving ocean acidification, pollution, nutrient cycling, or chemical runoff impacts. - **Biological Oceanographers or Marine Biologists -** Study marine organisms and their interactions with the ocean environment. They are ideal sources for stories on biodiversity, fisheries, invasive species, and ecosystem health. - **Geological Oceanographers or Marine Geologists -** Study the structure and composition of the ocean floor. They can provide insights into underwater earthquakes, tsunamis, deep-sea mining, or the formation of underwater features. - **Climate Scientists with Ocean Expertise -** Examine how oceans influence and respond to climate change. They are helpful for broader climate stories that involve ocean heat content, carbon storage, or long-term trends in ocean conditions. - **Marine Ecologists -** Study relationships among marine organisms and their environment. They can clarify ecosystem-level impacts, like those from overfishing, coral bleaching, or marine protected areas. - **Fisheries Scientists -** Specialize in fish populations, fishing practices, and resource management. Helpful for reporting on commercial fishing, stock assessments, or policy/regulation issues. - **Ocean Data Scientists -** Work with large marine datasets and modeling, can assist with interpreting satellite data, ocean models, or big datasets. - **Marine Policy Experts and Ocean Economists -** Focus on the intersection of ocean science, law, and economics. Helpful for coverage of marine regulations, governance issues, or the ‘blue economy.’ - **Marine Technologists or Ocean Engineers -** Design and use tools like underwater drones, sensors, and buoys. They can help explain how ocean data is collected and what the limitations of certain technologies might be. ## 3. Case Study: Gulf of Maine Ocean Warming  As ocean scientists from the Northeastern United States, we have each witnessed how rapid ocean changes are affecting the ecosystems and communities around us. For this example case study, we focus on the Gulf of Maine—a region close to home. When telling ocean stories, it is helpful to have either first-hand or personal connections to the coastal or oceanic region you are investigating. ### 3.1 Motivation The Gulf of Maine is warming [faster than 99% of the global ocean](https://eos.org/features/why-is-the-gulf-of-maine-warming-faster-than-99-of-the-ocean), making it a key site to investigate local impacts of climate change on marine environments and coastal livelihoods. Stories of changing fish stocks and stressed fisheries are already discussed in communities within the Northeastern region. Before getting into the data, we will think through the historical and ecological context of the Gulf of Maine and its fisheries. For centuries, the regional identity has been deeply linked to the ocean. Indigenous [Wabanaki peoples](https://www.wabanakialliance.com/wabanaki-history/)—including the Abenaki, Mi'kmaq, Maliseet, Passamaquoddy, and Penobscot nations—relied on these coastal waters for food as well as cultural practices and trade. They managed their coastal and marine environments with ocean knowledge developed across generations. When European colonization began, [intensive cod fishing fueled transatlantic trade and early settlements](https://www.markkurlansky.com/books/cod-a-biography-of-the-fish-that-changed-the-world/). Europeans considered the cod fisheries to be so abundant that they were endless. The overfishing by settlers caused a massive collapse in cod stocks by the 1950s. Now, other important local fisheries like the American lobster are being impacted by the combination of ocean warming and historic overfishing. Harmful algal blooms have also increased in frequency which indicate that the broader Gulf ecosystems are under stress. In the following sections, we guide you through using publicly available ocean datasets to investigate the scientific questions behind the Gulf of Maine and its warming waters. By accessing current and archival datasets you will be able to visually show the seawater temperatures going up and connect that to other environmental stories or investigations about the Gulf. ### 3.2 Data acquisition In order to investigate warming in the Gulf of Maine, we will analyze surface temperatures from two different datasets: a local *in situ* temperature sensor and the global-average SST. With the global SST as our baseline, we’ll be able to determine how much faster the Gulf of Maine is warming compared to the rest of the world. This analysis involves database downloads, data post-processing/analysis, and data visualization. If you don’t have experience with coding and want to get started with the Python programming language, see the appendix for tips on getting setup. Otherwise, you can always consider contacting a scientist to support you with your investigation (see section 2.5.2 Scientific support). #### 3.2.1 Gulf of Maine buoy temperature dataset First, we’ll go to the [National Data Buoy Center (NDBC)](https://www.ndbc.noaa.gov/) website and look for a buoy in the Gulf of Maine with a long historical record of temperature measurements. Clicking on the “Historical Data & Climatic Summaries” link at the bottom of [Station 44007’s page](https://www.ndbc.noaa.gov/station_page.php?station=44007) reveals annual text files going back to 1982.  *Screenshot from the NDBC website showing potential buoys to use for Gulf of Maine case study.* The task now is to process all of this data into a more usable format. We’ll do this with a python script using the [pandas](https://pandas.pydata.org/) data analysis library. 1. Loop through the years 1982-2024 and create the dataset url for each year, using the NDBC website to deduce the url structure. 2. Load the text data directly from each url via `pandas.read_csv()` 3. Convert the year, month, day, hour columns into a single pandas datetime column. 4. Combine all of the data into a single dataframe. 5. Save our data to a new CSV file.  *An example of what the available buoy data looks like for the year 1985. The highlighted sections show the parts of the dataset that we’re interested in: the date/time and the water temperature.* #### 3.2.2 Global mean SST dataset Next, we want a corresponding dataset for the globally-averaged SST, in order to determine whether the Gulf of Maine is warming faster or slower than the average. The [Climate Reanalyzer](https://climatereanalyzer.org/clim/sst_daily/?dm_id=world2) displays globally-averaged SST from the [NOAA 1/4° Daily Optimum Interpolation Sea Surface Temperature (OISST)](https://www.ncei.noaa.gov/products/optimum-interpolation-sst), a long term Climate Data Record that incorporates observations from different platforms (satellites, ships, buoys and Argo floats) into a regular global grid. 1/4° refers to the grid resolution—about 25 km. There is an option to download the underlying data from the Climate Reanalyzer website, which will save us a lot of time vs. trying to access decades of data and doing the global averaging ourselves. The data is available as a JSON file, which is a different text file format that will require a more custom approach for converting into a pandas dataframe.  *Screenshot of the [Climate Reanalyzer](https://climatereanalyzer.org/clim/sst_daily/?dm_id=world2) website. In the dropdown menu, we want to download the JSON data.* One key concept to note is how this data handles dates. Each year includes a list of 366 temperatures, without any explicit list of the corresponding dates. This is using the format of “day of year” and we see that the last temperature is “null” for non-leap years. When processing this data, we need to take this into account and ignore the null final value. Similar to the buoy data, we’ll re-format this data in a pandas dataframe and save to a new CSV file.  *A look inside the globally-averaged SST JSON file. The data is arranged as a list of years where each year has a list of 366 temperatures.* ### 3.3 Climatological data analysis A standard method for analyzing climate change anomalies is to first remove the climatological “seasonal” signal from the data. This will allow us to show, for each data point, how much warmer or colder it was than the average temperature for that day of the year. The first step is choosing which time period we’ll use for our climatology “baseline”. Here we’ve chosen 1991 to 2020 since it is fully covered by our data and matches the climatology period used by the Climate Reanalyzer website. Next, we’ll use some built-in pandas methods to get the climatological average temperature for each day and then map that to each datapoint in our timeseries. The following code snippet shows the steps used for both the buoy data and the global SST: ```python # Select just the data in the range of the climatology period df_clim = df[(df.index.year >= 1991) & (df.index.year <= 2020)].copy() # Assign the day of year (1-366) to each data point in the timeseris df_clim["day_of_year"] = df_clim.index.dayofyear # Take the mean for each day_of_year df_clim = df_clim.groupby("day_of_year")["temp"].mean() # New variable in df: the climatological temperature for that day df["day_of_year"] = df.index.dayofyear df["climatology_value"] = df["day_of_year"].map(df_clim) # Temperature anomaly is observed temperature minus climatological temperature df["anomaly"] = df["temp"] - df["climatology_value"] ```  *Our resulting dataframe includes new columns for climatology and temperature anomaly.* ### 3.4 Analyzing and Visualizing the results First, we’ll simply plot the full temperature timeseries and see what we find.  The warming signal is instantly apparent in the global SST data because the seasonal signal is so small. The Gulf of Maine, however, varies by more than 15° C throughout the year so any long term changes are difficult to see in this format. Plotting the climatology signal illustrates this point (pay attention to the y-axis).  Next we’ll view our temperature anomaly data (observed temperature minus climatology). As expected, there is more noise in the buoy data since it’s taken from a single point and any given day can vary by as much as 4 °C from climatology. The globally-averaged temperature has much less variance.  For the final version of our plot, we’re incorporate 3 changes: 1. Fit a simple linear regression using [numpy’s polyfit](https://numpy.org/doc/stable/reference/generated/numpy.polyfit.html) in order to quantify the average rate of warming for the two datasets. 2. Plot the monthly averages instead of the daily values in order to simplify the visual clutter. 3. Use the same y-axis range for the two plots for direct visual comparison.  Comparing our warming rate calculations against the published literature finds good agreement: - Gulf of Maine SST: our rate of 0.496°C/decade is within 5% of the 0.47°C/decade reported by the [Gulf of Maine Research Institute](https://gmri.org/stories/2024-gulf-of-maine-warming-update/). This is likely due to differences in methods—we used a single buoy and they used the OISST data averaged across the entire Gulf. - For global SST, our rate of 0.188 °C/decade is within 5% of the 0.18 °C/decade (over the past 50 years) published by [Samset et al. (2023)](https://www.nature.com/articles/s43247-023-01061-4). These final plots provide simple visual evidence of the Gulf of Maine’s rapid warming over the past 40 years. We showed the data transform from text files, to noisy timeseries, and finally to expert-validated trend lines. By removing the strong seasonal signal and focusing on the anomalies, we can clearly see the long-term warming trend in both the Gulf of Maine buoy data and the global mean SST. Finally, note that the linear regression is useful for quantifying the recent warming in an easily understandable number but is not necessarily a predictor of future warming. The [Maine Climate Science Dashboard](https://climatecouncil.maine.gov/future/climate/science-dashboard) shows the potential for human emissions to either accelerate or slow down this rapid warming.  *The [Maine Climate Science Dashboard](https://climatecouncil.maine.gov/future/climate/science-dashboard) combines historical water temperature measurements with different climate scenario forecasts.* ## 4. Conclusion Our investigation into Gulf of Maine temperatures, using readily available public datasets, highlights one local manifestation of global climate change. This rapid warming isn't merely an abstract data point, it continues to have profound implications for the region’s biodiversity and the human communities who rely on the ocean. Marine species are highly sensitive to temperature changes, and the Gulf of Maine has been experiencing a noteworthy decline in native species and [increase in warmer-water species](https://online.ucpress.edu/elementa/article/9/1/00076/118284/Climate-impacts-on-the-Gulf-of-Maine-ecosystemA). The next steps in this story might look to other data sources to explore: Why is the Gulf of Maine warming so quickly? and What will the region look like in the future? or How exactly are local fisheries affected by warming waters? This case study is one example of how to find the connections between global environmental change, local ocean data, and tangible human impacts. This process offers a template for investigating similar stories in your own regions: 1. **Start with a local observation or community concern:** What are people witnessing or experiencing in your local environment? 2. **Explore the scientific context:** Consult with scientists, read relevant research, and understand the underlying environmental drivers. 3. **Seek out publicly available data:** As shown in section 2, there is a large assortment of high-quality public ocean datasets that can be used to investigate countless questions. 4. **Connect the data back to human issues:** How do the environmental changes revealed by the data affect local cultures, livelihoods, health, and economies? The key thing to remember is that there are multiple angles to uncover and expose often-invisible impacts to the ocean. Datasets provide one lens to report on climate and environmental changes, but these stories impact communities and are thus both political and social. Just as ocean science has changed and begun to decolonize, it's crucial to investigate and tell stories that reflect diverse experiences. Ocean data can help highlight intersecting issues—such as deep seabed mining, marine health, and colonial continuums—with evidence-based information and compelling visualizations. We hope this guide offers a practical starting point for navigating ocean science, accessing and interpreting data, and connecting your investigation to real-world consequences that are planetary in scale yet intimately local. ```cik-note ``` >**APPENDIX: Getting started with python** > >If you have not done any coding before, the initial task of setting up your coding environment can be a challenging hurdle. There are multiple options for code editors/IDEs (integrating development environment), ways of handling dependencies (the external packages you install to give you advanced functionality), and other decisions that are outside the scope of this article. Luckily, once you’ve chosen your tools, there are good resources online so here are a few recommendations and then you can seek out more detailed tutorials: > >1. Use Visual Studio Code as your code editor (the application where you will write and run code). This is the most popular option and there is an extensive ecosystem of 3rd party plug-ins and help resources. <https://code.visualstudio.com/> > >2. Use [conda](https://docs.conda.io/projects/conda/en/latest/user-guide/getting-started.html) for package management. Your computer’s operating system may come with a version of python pre-installed but it's not a good idea to install packages onto this global location. Instead, create separate conda "environments" for different projects. This will allow you to experiment in a safe and organized way. Here is a helpful article on the VS Code website: <https://code.visualstudio.com/docs/python/environments>. For example, to create a new environment that we'll name "ocean-study" and install the "matplotlib" plotting package would look like this: > > ```bash > conda create -n ocean-study > conda activate ocean-study > conda install matplotlib > ``` > Now, in VS Code, just make sure your Python Interpreter is using this environment (it will look something like `~/miniconda3/envs/ocean-study/bin/python` and you will be able to use the matplotlib package in your code. > >3. Finally, consider using Jupyter Notebooks for exploratory coding where you’re loading datasets and making plots. Notebooks have the file extension `.ipynb` and allow you to run chunks of code independently in code "cells" and view the output right below. You can also use Markdown text cells to write notes and explanations for yourself and collaborators. Instructions on using VS Code: <https://code.visualstudio.com/docs/datascience/jupyter-notebooks> > <hr class="thick"> ##### About the authors **Mae Lubetkin** is an ocean scientist, transmedia artist, and writer based in Paris and at sea. Their practice-led research remaps our relations to bodies of water and digital worlds by means of investigation, counter-narrative, and memory. With a background in marine geology and subsea imaging, their artistic practice is in dialogue with Science while situated in queer, intersectional, anti-extractivist, and decolonial frameworks. Guided by wet-techno-critical studies and thinking with other-than-human worlds, they compose environmental traces in installations and digital outputs. Their core practice is in solidarity with submerged, ancient, ephemeral and imaginary environments. **Dr. Kevin Rosa** is an oceanographer and the founder of Current Lab, a startup specializing in computational ocean forecasting. He holds a B.A. in Physics and a PhD in Physical Oceanography from the University of Rhode Island, with a focus on ocean physics and hydrodynamic modeling. <hr class="thick"> *Published in June 2025*
Ocean Datasets for Investigations
=============================== By Mae Lubetkin and Kevin Rosa  ```cik-in-short ``` **In short:** Learn how to identify and use Ocean datasets as a tool for revealing the unseen or underreported dynamics of the world's most significant bodies of water. --- ## 1. Introduction: Ocean Science, Data, Storytelling Given our current planetary condition, many of the world's most pressing stories are linked to the ocean. Covering 71% of Earth's surface and interacting constantly with the atmosphere, the ocean is our greatest carbon sink and an essential indicator of climate change. Despite its critical role in maintaining a habitable planet and supporting coastal livelihoods, the ocean is often invisible to the lived experience of most individuals, particularly those far from its shores. There are so many ways to tell stories about the ocean, and countless diverse perspectives from which to understand it. Now, more than ever, we need to incorporate cross-cultural and trans-disciplinary strategies for investigative projects, particularly those concerning the ocean. This guide presents ocean datasets as a tool for revealing the unseen or underreported dynamics of the world's most significant bodies of water. For informed investigations and impactful storytelling, oceanographic datasets can be an essential resource for journalists, activists, and anyone interested in data-driven methods to communicate the climate crisis, environmental change, natural disasters, extractivism, and associated ocean justice issues. From bathymetric maps, subsea imagery, and 3-D habitat models, to satellite-derived and *in situ* real-time monitoring data – a vast amount of oceanographic media and data is publicly available. In this Exposing the Invisible Guide, we begin with an introduction on the broader scientific history and context within which ocean data is collected, stored, and made accessible. Section two outlines the diversity of datasets, including some history, trade-offs, use cases, data collection methods, data sources, and resources to learn more about each data type that we present. Section three offers a specific application of ocean data in a case study, including: steps explaining why the data is useful for supporting this particular story; how to source and present the data; and, finally, how to bring it into a meaningful investigative report, journalistic piece, or other storytelling format. The guide concludes with a summarized approach for using ocean datasets in investigations and outlines strategies for identifying the right ocean scientists who could support you and your investigation.  *Boulder resting on the seafloor offshore the Revillagigedo Archipelago in the Pacific Ocean. Credit: Ocean Exploration Trust.* ### 1.1 Ocean Science: History and Context Ocean science generally refers to the observation and investigation of biological, geological, physical, and chemical processes that shape and constitute global marine environments. This broad disciplinary field includes numerous sub-disciplines that focus on intricate, detailed studies of specific scientific questions concerning the ocean. Ocean scientists monitor habitat change, measure biodiversity, investigate geological phenomena, and study human impacts on ocean systems (i.e., global warming, pollution, overfishing, and extractive projects). Interactions between marine ecosystems and human activities, as well as atmospheric and coastal processes, are all carefully investigated by ocean scientists today. Despite niche specializations, there are increasingly multidisciplinary projects that involve diverse experts in order to more comprehensively understand the interconnections between oceanic processes and phenomena. Collectively, this research improves our baseline knowledge of the ocean, which can then support preservation strategies while maintaining sustainable and regenerative relationships with diverse marine ecosystems. Although ‘contemporary’ ocean science has deep roots in European colonialism and imperial exploration, ocean knowledge systems long predate Western scientific inquiry. Indigenous and coastal communities across Oceania, as well as the Atlantic and Indian Oceans, carefully studied and navigated the seas for thousands of years. These forms of ocean science are less known or dominant on a global scale, but they nevertheless involve highly sophisticated techniques for understanding the stars, ocean swells, winds, and currents. Navigators across Oceania used interconnected and embodied forms of ocean science, on their own terms, to journey vast distances across the seas with tremendous precision. While other coastal Indigenous peoples around the world developed specific place-based systems of knowledge, including both land and marine management practices that viewed these spaces as highly linked. Most of these communities understood the ocean not as a monstrous or alien-filled void (as European explorers often depicted it), but as a vast world interconnected with our own.  *Rebbilib (navigational chart) by a Marshall Islands artist in the 19th to early 20th century. These stick charts were used by Marshall Islander navigators during long ocean voyages. Credit: Gift of the Estate of Kay Sage Tanguy, 1963. Source: <https://www.metmuseum.org/art/collection/search/311297>* When European seafarers began worldwide exploration voyages in the 15th-16th centuries, they dismissed or actively erased these Indigenous ocean knowledge systems. European or 'Western' scientific models were linked to a colonial mindset that often viewed the natural world as a space to examine in order to master and own its elements, rendering them as 'resources'. Notions of relationality were strongly opposed to the point that scientists considered their surroundings as objects to be studied rather than elements to relate to or work with. At the core, these opposing ocean scientific methods or knowledge systems reflected the specific values and worldviews of each culture, respectively. The Challenger Expedition (1872–1876) was the first European-led systematic scientific exploration of the global oceans. In some ways, it was groundbreaking. However, it also played a key role in the broader colonial project, which aimed to map, control, and extract 'resources' from around the world. Today, Western or 'contemporary' ocean science continues to use investigatory methods that stem from European knowledge systems. Oceanographic research often occurs in the context of economic or territorial expansion and military-supported science projects. Nevertheless, these methods are beginning to open up to other forms of knowledge creation that move beyond the geopolitical interests of wealthy nations.  *Map of ocean currents created by John Nelson using the "WGS 1984 Spilhaus Ocean Map in Square" projected coordinate system in ArcGIS. The Spilhaus Projection (developed by oceanographer Athelstan Spilhaus in 1942) reflects aspects of decolonial cartography by shifting focus from land-centered perspectives to an oceanic worldview. Source <https://storymaps.arcgis.com/stories/756bcae18d304a1eac140f19f4d5cb3d>* Thanks to the enduring activism and decolonizing work led by many Indigenous and coastal communities, there is increasing recognition of the need to reclaim ocean science by amplifying the knowledge and perspectives of those who have long understood the seas. Ocean science is just beginning this deep process of decolonization, which first seeks to acknowledge and reckon with the violent and ongoing impacts of colonization. Decolonizing ocean science requires a fundamental shift in who holds agency and sovereignty over their own ocean waters. This also relates to international waters and who is included, excluded, or undervalued throughout negotiations concerning the legal and regulatory structures governing global oceans. Today, many ocean scientific projects are co-designed and co-led by ocean knowledge holders from diverse backgrounds. Collecting ocean datasets requires a team of experts who follow cultural protocols, ensure environmental safety, and apply diverse scientific methods, all while striving for more relational practices. ### 1.2 Ocean Data Collection Today Today, ocean datasets are collected by ocean scientists in collaboration with ocean engineers. These datasets are gathered from several sources to understand the global ocean and its role in maintaining Earth's habitability and critical planetary cycles. Ocean engineers develop the tools, platforms, and instruments that are required for data collection, such as underwater vehicles, satellite-mounted sensors, and buoys. By designing technologies that can operate in diverse and sometimes extreme conditions, these engineers support and expand ocean scientific capabilities. Together, ocean scientists and engineers advance our understanding of the planet for both research and conservation. There is a considerable variety of ocean data types, tools for data collection, and associated databases to store these recorded entities. This diversity of datasets is outlined in section 2. Like most scientific fields, funding can be secured from public governmental bodies or private sources. The ocean datasets we focus on here are publicly accessible and typically funded by governments via taxpayer contributions. This means that ocean datasets are for the people and should be accessible. Unfortunately, many public ocean datasets are stocked in complex databases and require specialized software, programming experience, or extensive knowledge to access. That said, there are plenty of datasets that can be retrieved and visualized more easily, with little to no background knowledge. There are also some ocean datasets that can be accessed with helpful tips and instructions, which is what we will focus on here. The Exposing the Invisible Toolkit is designed as a self-learning resource, we hope that this guide will support future investigations and make ocean datasets more accessible to communities and investigators around the world. ### 1.3 Data Gaps, Capacity Initiatives, Ocean Defenders Conducting ocean science can be a costly endeavor. Depending on the environment, scientific goals, and technical requirements, some ocean scientific work can only be conducted by wealthy nations or private organizations. Typically, this kind of science takes place at sea or uses remote sensing techniques. For example, deep ocean exploration and research requires a ship, vehicles, or platforms to deploy down to the deep ocean, technical and computing systems aboard to process the data, and a diverse team of experts to manage these operations. In contrast, satellite remote sensing used for ocean research typically covers the entire Earth surface. Publicly funded satellite-derived ocean datasets can be useful across territorial waters, throughout international seas, and are accessible regardless of one's nationality. Near-shore ocean science and *in situ* coastal monitoring efforts are more financially affordable, especially as diverse knowledge systems merge with lower-cost technologies and capacity initiatives. In this context, capacity refers to the skills, resources, and knowledge needed to effectively conduct ocean science. As ocean science undergoes a process of decolonization, this emphasis on capacity building, development, and sharing is also strengthening. Additionally, several international ocean law and policy frameworks specifically aim to increase ocean science capacity. Ocean defenders are also central to these efforts. As groups, individuals, or organizations dedicated to protecting marine environments, defenders play a key role in advocating for capacity building within and outside scientific structures. Many defenders are fisherpeople, coastal communities, or groups directly affected by changes in climate and ocean health. Beyond advocating for sustainable and generative oceanic futures, they also fight to overcome political resistance and funding barriers. Ocean defenders, like land defenders, face challenging or dangerous obstacles while pushing for local and global ocean preservation. Ocean science and policy clearly needs collaborative approaches that bring multiple knowledge systems forward while prioritizing those most impacted by climate change, pollution, and other threats to both marine habitats and coastal livelihoods.  *Ocean-defending artisanal fishers and their supporters in South Africa celebrate upon receiving the news that Shell’s permit to conduct a seismic survey on the Wild Coast had been set aside by the Makhanda High Court, in September 2022. Photo credit: Taryn Pereira. Source: <https://oceandefendersproject.org/case-study/no-to-seismic-surveys/>* **More on capacity initiatives, knowledge gaps, and ocean defenders:** - Guilhon, M., M. Vierros, H. Harden-Davies, D. Amon, S. Cambronero-Solano, C. Gaebel, K. Hassanali, V. Lopes, A. McCarthy, A. Polejack, G. Sant, J.S. Veiga, A. Sekinairai, and S. Talma. (2025). [Measuring the success of ocean capacity initiatives.](https://doi.org/10.5670/oceanog.2025.122) *Oceanography* 38(1). - Saba, A.O., I.O. Elegbede, J.K. Ansong, V.O. Eyo, P.E. Akpan, T.O. Sogbanmu, M.F. Akinwunmi, N. Merolyne, A.H. Mohamed, O.A. Nubi, and A.O. Lawal-Are. 2025. [Building ocean science capacity in Africa: Impacts and challenges.](https://doi.org/10.5670/oceanog.2025.133) *Oceanography* 38(1). - Behl, M., Cooper, S., Garza, C., Kolesar, S. E., Legg, S., Lewis, J. C., White, L., & Jones, B. (2021). [Changing the culture of coastal, ocean, and marine sciences: strategies for individual and collective actions.](https://www.jstor.org/stable/27051390) *Oceanography*, *34*(3), 53–60. - The Ocean Defenders Project (2025). [Ocean Defenders: Protectors of our ocean environment and human rights.](https://oceandefendersproject.org/project-publications/ocean-defenders-protectors-of-our-ocean-environment-and-human-rights/) The Peopled Seas Initiative, Vancouver, Canada. - Belhabib, D. (2021) [Ocean science and advocacy work better when decolonized.](https://doi.org/10.1038/s41559-021-01477-1) *Nat Ecol Evol* 5, 709–710. - Kennedy, R. and Rotjan, R. (2023). [Mind the gap: comparing exploration effort with global biodiversity patterns and climate projects to determine ocean areas with greatest exploration needs.](https://doi.org/10.3389/fmars.2023.1219799) *Front. Mar. Sci.* (10). - Bell, K.L.C, Johannes, K.N., Kennedy, B.R.C., & Poulton, S.E. (2025) [How little we’ve seen: A visual coverage estimate of the deep seafloor.](https://www.science.org/doi/10.1126/sciadv.adp8602) *Science Advances*, Vol 11, Issue 19. ### 1.4 Ocean Datasets for Investigations and Storytelling Ocean datasets play a crucial role in both scientific investigations and storytelling by providing evidence-based insights into the health and dynamics of our global ocean. These datasets help scientists better understand the ocean, but beyond research, they serve an important role in ocean-related investigations. Ocean datasets can support communities, journalists, and activists in raising data-backed awareness about critical marine and environmental justice issues. By sharing this data in accessible ways, oceanographic narratives can amplify the voices of coastal communities, engage the public, and inspire action in support of more regenerative ocean futures. Numerous well-resourced journalistic and forensic organizations use ocean data to support their stories or reporting, such as Forensic Architecture, LIMINAL, Forensis, Earshot, Border Forensics, and others. In this guide, we will demonstrate how you can access datasets and conduct your own oceanic investigations. By the end, you will be able to illustrate a well-defined ocean or climate question using publicly available oceanographic datasets and media collections, which will enhance your evidence-based and visually engaging story. ## 2. Diversity of Data Types The following sub-sections serve as a collection of key ocean data types, organized by how they are collected and what they reveal. Each broad data source type (i.e., the overarching technique used to gather certain kinds of ocean data) begins with a bit of history and trade-offs, and then is further broken down into specific data products. For each data product, we share use cases, data collection methods, key considerations, and some databases (often from U.S. and European agencies), followed by resources to learn more. This structure is designed to help you understand how each dataset is produced, grasp the significance of the data, and know where to go for deeper investigations or analyses. There are many data types presented below which are organized in these four broad categories: *in situ* sensors; deep ocean observation, exploration, and research; mapping; and, satellite remote sensing. See Section 3 to review an example case study which demonstrates how some of these datasets may be used to support an investigative ocean and climate story.  *Illustration of a range of ocean sensor platforms—ships, profiling drifters, gliders, moored buoys, and satellites. Source: <https://www.ecmwf.int/en/about/media-centre/news/2021/world-meteorological-day-focuses-role-ocean-weather-and-climate>* ### 2.1 *In Situ* Sensors *In situ* sensing refers to the direct measurement of ocean properties using instruments that are physically located within the water. Sensors for measuring temperature, salinity, pressure, currents, and of biochemical concentrations are deployed on a range of platforms with various advantages and drawbacks. While satellites provide broad spatial coverage of the ocean’s surface, *in situ* platforms are essential for monitoring the ocean’s interior, tracking coastal change, and measuring water properties that cannot be detected remotely. - **History:** - Sailors have used thermometers to measure ocean temperature since at least as early as Captain James Cook’s [1772 voyage](https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/rog.20022) to the Antarctic Circle (another example of colonial science forming the foreground to Western ocean science). - The Nansen bottle (1896) and later the Niskin bottle (1966) enabled the capture of water samples at specific depths, which could then be pulled up and tested for temperature and salinity on the ship. - The first bathythermograph was developed in 1938 and featured a temperature sensor on a wire which recorded a temperature profile as it was lowered into the ocean. This was used by the US Navy in WWII to improve sonar accuracy, since temperature layers alter acoustic propagation. - Today, there are thousands of advanced sensors across the world’s oceans which transmit readings in real time via satellite. - **Trade-offs:** - *In situ* sensors can only measure the ocean properties at their exact location so great consideration is taken in their placement and timing. - Powering the instruments is a constant challenge and factors into decisions about sampling frequency. - The extreme pressure in the deep ocean limits the operating depth of some sensors. - Harsh operating conditions limit the lifespan of sensors and necessitates regular maintenance/replacement, often in remote locations at high costs. This leads to less accessible areas being undersampled. #### 2.1.1 Moorings and Fixed Platforms  *Example of a coastal data mooring. Source: Baily et al., 2019, <https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2019.00180/full>* - **About:** - A range of platforms anchored in place, collecting time-series data at a fixed location. - Used in the deep ocean (e.g., the TAO array across the Pacific Ocean), the continental shelf (e.g., NOAA’s National Data Buoy Network), and at the coastline (e.g., tide gauges). - **Use cases:** - Long-term climate monitoring of ocean heat content and circulation. - Data inputs for forecast models, improves accuracy of ocean and weather predictions. - Early warning systems for tsunamis and hurricane storm surge. Tracking tidal heights and local sea level rise. - Water quality monitoring for pollutants, algal blooms, and hypoxia. - **Data collection:** - Sensor packages measure temperature, salinity, pressure, biochemistry, and more. - Some moorings have Acoustic Doppler Current Profilers (ADCPs) to measure water current velocities throughout the water column. - **Key considerations:** - Data today is mostly broadcasted in near-real time, but there are some platforms that require physical retrieval before the data is downloaded. - Spatial coverage is extremely limited and focused around a small set of nations. - The diversity of databases and data types can pose a challenge for accessing and working with the data. - **Data sources:** - [US NOAA National Data Buoy Center (NDBC)](https://www.ndbc.noaa.gov/) - [EU Copernicus Marine Service In Situ dashboard](https://marineinsitu.eu/dashboard/) - [Global Tropical Moored Buoy Array](https://www.pmel.noaa.gov/gtmba/) - **Resources to learn more:** - [WHOI - Moorings & Buoys](https://www.whoi.edu/what-we-do/explore/instruments/instruments-moorings-buoys/) - [Ocean Observatories Initiative](https://oceanobservatories.org/) #### 2.1.2 Drifters and Floats  *Map of Argo float locations. Source: <https://argo.ucsd.edu/about/status/>* - **About:** - Unanchored and unpropelled instruments that drift with the currents and take ocean measurements. - Drifters stay at the surface and provide information about surface conditions, and surface currents are calculated from their GPS trajectory. - Floats profile the water column by adjusting their buoyancy to move up and down. The Argo program is the largest and most significant, with over 4,000 Argo floats profiling the world’s oceans. - Capable of providing global coverage at lower cost than moored sensors, especially in remote open-ocean regions. - **Use cases:** - Drifters: Mapping near-surface currents and tracking pollutants and marine debris transport. - Floats: Measuring subsurface temperature and salinity for climate studies. Some Argo floats also have biochemical sensors. - **Data collection:** - Drifters: GPS-tracked, measure SST, pressure, sometimes salinity, sometimes waves. - Argo floats: Profile down to 2,000 m every 10 days, transmitting data via satellite. - **Key considerations:** - Drifters and floats are always moving, so you can’t get a clean timeseries for a single location like you can with moorings. Additionally, Argo floats only take one profile every 10 days in order to preserve battery life. - Argo floats don’t generally operate near the coast on the continental shelf. - Some drifters/floats lack real-time telemetry (data transmission). - **Data sources:** - [Global Drifter Program](https://www.aoml.noaa.gov/phod/gdp/) - [Argo Program](https://argo.ucsd.edu/) - [Copernicus Marine Service drifters](https://data.marine.copernicus.eu/products?facets=featureTypes%7ETrajectory) - [SOCCOM Biogeochemical Floats](https://soccom.org/) - **Resources to learn more:** - [About Argo](https://argo.ucsd.edu/about/) #### 2.1.4 Autonomous Vehicles - ASVs and Gliders  *Illustration of a glider’s sawtooth propulsion pattern. Source: <https://earthzine.org/going-deep-to-go-far-how-dive-depth-impacts-seaglider-range/>* - **About:** - Autonomous Surface Vehicles (ASVs) and gliders are robotic platforms that enable long-duration, energy-efficient monitoring over vast areas. - Bridging the gap between targeted, expensive, ship-based measurements and low-cost, but uncontrolled, drifters/floats. - **Use cases:** - Significant overlap with the use cases for moored sensors and drifters/floats. - Targeted measurements in dangerous conditions like hurricanes. - Mapping surveys, autonomous of ship or in tandem with other vehicles. - **Data collection:** - Autonomous Surface Vehicles (ASVs) use solar panels, wind, or waves as a power source to supplement and recharge their batteries. - Gliders are underwater vehicles that create propulsion by adjusting their buoyancy and gliding horizontally while sinking/rising (similar to an airplane). This enables longer battery range than propellers or thrusters. - **Key considerations:** - Gliders and ASVs are often used in targeted studies rather than continuous global monitoring and thus have lower data availability. - Shorter mission durations than moorings or drifters/floats. - **Data sources:** - [NOAA Glider Data Assembly Center](https://gliders.ioos.us/) - [OceanGliders](https://www.oceangliders.org/) - **Resources to learn more:** - [National Oceanography Centre UK - Gliders](https://noc.ac.uk/facilities/marine-autonomous-robotic-systems/gliders) ### 2.2 Deep Ocean Observation, Exploration, and Research Deep ocean science is typically conducted to observe long-term changes at specific seafloor sites, to explore marine habitats that are unknown to science, and to conduct applied or experimental research on focused environmental questions. A range of deep ocean data collection tools include platforms or landers, cabled observatories, and deep submergence systems—such as human-occupied vehicles (HOVs), remotely-occupied vehicles (ROVs), and autonomous underwater vehicles (AUVs).  *Human-occupied vehicle (HOV) 'Alvin' being recovered in 2024 during an expedition to the East Pacific Rise hydrothermal vent fields. Photo credit: Mae Lubetkin* - **History** - 1872–1876: The *HMS Challenger* expedition, a milestone for deep ocean science but deeply tied to imperialism - Mid-20th century: Cold War military priorities further developed submersible vehicle capabilities, leading to *Trieste*’s 1960 Mariana Trench dive - 1964 onward: HOVs expanded access to the deep ocean, e.g., *Alvin* (US), *Nautile* (France), *Shinkai* (Japan), and *Mir* (Russia) - 1980s–2000s: ROVs and AUVs developed by industry (oil, mining, and defense) and scientific institutions, in parallel - 2000s–present: Cabled observatories (e.g., Ocean Networks Canada, DONET in Japan), public research campaigns (e.g., NOAA, IFREMER), and oceanographic instruments expanded reach and scope. - Today: Many regions still face barriers to participation and funding in deep ocean science (as outlined in the introduction). Meanwhile, deep submergence science in wealthy nations increasingly utilizes AI, autonomous systems, 4K and 3-D imaging techniques. - **Trade-offs:** - Provides direct access to deep ocean environments which are inaccessible by surface vessels or remote sensing - High spatial and contextual resolution: can capture detailed imagery, samples, and detailed *in situ* measurements - Resource-intensive: operations usually require ships, launch/recovery teams, and specialized personnel - Limited temporal and spatial coverage: data collection is episodic, site-specific, and dependent on expedition funding, schedules, and weather conditions at-sea - High costs and technical barriers mean deep ocean science is dominated by a few well-funded institutions or nations, with limited global access - Colonial legacies persist in relation to who sets research agendas, who makes funding decisions, and who benefits from collected data #### 2.2.1 Deep Submergence Systems (HOVs, ROVs, AUVs) - **About:** - Vehicles that operate in the deep ocean water column or along the seafloor, including: - Human-occupied vehicles (HOVs): carry scientists directly, typically 1-3 observers and a pilot - Remotely operated vehicles (ROVs): tethered and piloted from a surface vessel like a ship - Autonomous underwater vehicles (AUVs): untethered and pre-programmed - These systems can operate from hours to days and are depth-rated around 4000-6000 m, but some may reach full ocean depths (11 km) and others may work well in shallower waters. - **Use cases:** - High-resolution visual surveys - Precise targeted sampling with environmental context and imagery at diverse environments including hydrothermal vents, methane seeps, cold-water coral habitats, and others - Biogeographic habitat mapping - Wreck exploration and infrastructure inspection - Imagery of deep ocean environments can support visual storytelling, public engagement, and education - **Data collection:** - Data is streamed directly to the support vessel for tethered operations - While for untethered submersibles (HOVs and AUVs) most data is retrieved when the vehicle is recovered - All physical samples are retrieved and processed upon vehicle recovery - **Key considerations:** - Requires experienced pilots and operational support (expensive) - AUVs need detailed mission planning (mission failure could lead to vehicle loss) - Navigation and environmental risks must be managed carefully - **Data sources:** - SeaDataNet - [EU research vessel data](https://csr.seadatanet.org/), including cruise summary reports and more - EuroFleets - European initiative to compile [EU research cruise data](https://www.eurofleets.eu/data/) - Rolling Deck to Repository - [US research vessel data](https://www.rvdata.us/data), including: expedition summary, shiptrack navigation, scientific sampling event log, post-processed data - JAMSTEC Databases - [Japan research vessel data](https://www.jamstec.go.jp/e/database/), including: HOV *Shinkai 6500* and ROV *Kaiko* mission data, cruise reports, and dive logs - **Resources to learn more:** - [Woods Hole Oceanographic Institution - National Deep Submergence Facility](https://ndsf.whoi.edu/) - [Ocean Exploration Trust - Science and Technology](https://nautiluslive.org/science-tech)  *An imaging elevator equipped with two camera systems, lights, battery packs, and crates to store additional sampling tools to be used by an HOV during a dive in the same region. Source: Woods Hole Oceanographic Institution* #### 2.2.2 Landers and Elevators - **About:** - Landers are relatively simple systems that descend to the seafloor and remain stationary for the duration of their deployment. - They are sometimes referred to as 'elevators' since they descend to the seafloor then ascend back to the surface - There are no people on landers, but they typically carry sensors, cameras, samplers, and other instruments - Depending on power supply and scientific goals, they can spend hours to sometimes months on the seafloor - **Use cases:** - Collecting environmental data (e.g., conductivity, temperature, pH, oxygen) - Capturing imagery of habitats or operations - Deploying baited cameras or traps to study biodiversity - Using the platform to carry additional gear or instruments to the seafloor that a deep submergence system could not transport on its own due to space limitations - **Data collection:** - Typically data is retrieved when the lander is recovered back on deck - Some landers will transmit data acoustically or remotely from the seafloor - The frequency that imagery or other data are collected is pre-programmed before deployment - **Key considerations:** - Requires careful site selection, recovery planning, and often ship time - Currents can impact the intended landing location on the seafloor, sometimes drifting the platform or lander far off-site - Limited in capabilities, not as advanced as deep submergence vehicles, but also much cheaper and easier to custom build - **Data sources:** - Deep ocean lander data (e.g., imagery and environmental sensor data) would be found in the same databases and repositories listed in section 2.2.1 Deep Submergence Systems (HOVs, ROVs, AUVs). - **Resources to learn more:** - [Schmidt Ocean Institute - Elevators and Landers](https://schmidtocean.org/technology/elevators-landers/) - [The Deep Autonomous Profiler (DAP), a Platform for Hadal Profiling and Water Sample Collection (Muir et al., 2021)](https://journals.ametsoc.org/view/journals/atot/38/10/JTECH-D-20-0139.1.xml) - [Lander Lab: Technologies, Strategies and Use of Ocean Landers (Hardy, 2022)](https://magazines.marinelink.com/Magazines/MarineTechnology/202201/content/technologies-strategies-landers-594271)  *Map of Ocean Networks Canada NEPTUNE and VENUS Observatories near Vancouver Island, Canada. Each orange square represents a node or station along the cabled observatory where instruments or sensors are mounted. Source: <https://www.oceannetworks.ca/>* #### 2.2.3 Cabled Observatories - **About:** - Mostly permanent, wired infrastructure on the seafloor that transmit real-time power and data via fiber optic cables connected to shore stations - Similar in some ways to the data collection tools described in section 2.1 *In Situ* Sensors, but these networks are fixed in location and networked - People do not visit these observatories, instead they support a wide range of sensors (e.g., temperature, pressure, seismometers, hydrophones, cameras, samplers) - Can integrate with ROVs or AUV docking stations, and are also typically maintained and serviced by ROVs - Designed for continuous, high-frequency monitoring of deep ocean processes across years or decades - They connect highly diverse environments from hydrothermal vent regions to abyssal plains and continental shelves - **Use cases:** - Long-term and consistent monitoring of geophysical activity (e.g., earthquakes, hydrothermal vents) - Real-time data for early warning systems (e.g., tsunamis, gas releases) - To study oceanographic processes (e.g., currents, biogeochemical fluxes, and ecosystem change) - Supports public engagement and education through livestreams - **Data collection:** - Real-time data is livestreamed to shore stations and then available via online portals - Most are operated by national or international research infrastructures - **Key considerations:** - Extremely costly, high maintenance needs (ROVs are often used for annual servicing) - Site selection is key since they are fixed installations - **Data sources:** - Ocean Networks Canada - [Oceans 3.0 Data Portal](https://data.oceannetworks.ca/), including all datasets, dashboards, and visualizers (more info on [ONC data](https://www.oceannetworks.ca/data/)) - US Ocean Observatories Initiative - [OOI Data Portal](https://oceanobservatories.org/data-portal/), includes cable-linked arrays on East and West Coasts and deep Pacific - EU [EMSO ERIC Data Portal](https://data.emso.eu/home), real-time and archived data, tools and research environment to investigate seafloor observatories across European margins - **Resources to learn more:** - [Ocean Observatories Initiative](https://oceanobservatories.org/observatories/) - arrays, infrastructure, instruments - [Ocean Networks Canada](https://www.oceannetworks.ca/observatories/) - observatories - [Interactive map of ONC locations](https://www.arcgis.com/home/webmap/viewer.html?webmap=fcea4e5f087f41c58bcc5e51b13fffa1&extent=-158.3094,39.6681,-29.8133,75.4182) - [Regional Cabled Observatories](https://www.whoi.edu/what-we-do/explore/ocean-observatories/about-ocean-observatories/types-of-observatories/regional-cabled-observatories/) - summary by Woods Hole Oceanographic Institution ### 2.3 Mapping Marine hydrography or ocean scientific mapping involves the creation of high-resolution representations of the seafloor, water column, and other associated features or phenomena (e.g., fish migrations, vents or seeps bubbling up) using vessel-based sonar, autonomous vehicles, acoustic or optical tools. It is a type of remote sensing since the mapping instrument is not on the seafloor. Unlike satellite remote sensing, which observes only the ocean surface from space, hydrographic mapping is conducted from platforms within or on the ocean surface. These mapping systems can resolve fine-scale topography (seafloor bathymetry), subsurface geologic layers, water column imaging, and habitat mapping that integrates both physical and biological data. Laser and 3-D reconstruction are other forms of high-resolution mapping. - **History:** - 1870s–1900s: Early bathymetric charts created using lead lines, linked to colonial navigation and maritime claims as well as scientific knowledge creation - 1920s–1940s: Echo sounding developed for military and commercial navigation, later repurposed for seafloor mapping - 1950s–1970s: Multibeam sonar developed, enabling wider swath coverage (i.e. can map wider seafloor area, not just single points) and broader seafloor topography or bathymetry mapping - 2010s–present: Autonomous vehicles, numerous specialized sonar systems, and 3-D photogrammetry advance deep mapping capabilities - Today: Mapping remains uneven globally—nations with limited funding or access to ships and processing capacity are underrepresented and do not have detailed seafloor maps of the their waters - **Trade-offs:** - High-resolution, fine-scale maps of seafloor and water column features - Enables geologic, biologic, and habitat-based spatial analysis - Requires significant ship time, technical expertise, and post-processing - Data coverage is patchy, most of the seafloor remains unmapped - High cost and national interests impact where mapping occurs and who benefits from the data  *Bathymetric mapping using a hull-mounted multibeam sonar system. Black lines indicate the ship’s track, while the coloration represents depth differences (red is shallow, purple is deep) used for visualizing the bathymetric or topographic features of the seafloor. Source: https://www.worldofitech.com/mapping-the-ocean-floor-water-bathymetry-data/>* #### 2.3.1 Bathymetric Mapping - **About:** - Measurement and charting of the depth and shape of the seafloor - Typically uses sonar-based systems (e.g., single-beam or multibeam echosounders), mounted on ships, AUVs, or towed platforms - Short-range systems (e.g., ROV-mounted sonar) provide highly detailed data over small areas (centimeters in resolution), while medium-range systems (e.g., hull-mounted multibeam on ships or AUVs) cover much larger swaths with lower resolution - **Use cases:** - Mapping underwater topography and geological features - Planning submersible dives and identifying hazards - Supporting infrastructure projects like cables or offshore wind farms - Creating base maps for habitat mapping or biogeographic studies (i.e. understanding what marine life lives where and how their habitats are linked to geologic features as well as currents and physical oceanographic phenomena) - **Data collection:** - By research vessels or autonomous vehicles using sonar systems - Key manufacturers include Kongsberg, Teledyne, R2Sonic, and Edgetech - Data is processed using specialized hydrographic software (e.g., QPS Qimera, CARIS, MB-System) - **Key considerations:** - Requires calibration (e.g., sound speed profiles) and correction for vessel motion - Deep ocean mapping can be slow and resource-intensive - Interpretation of raw bathymetry data requires trained analysts and geospatial tools, it is not yet fully automated - **Data sources:** - European Marine Observation and Data Network - [EMODnet Bathymetry](https://emodnet.ec.europa.eu/en/bathymetry), multibeam datasets and other maps with a built in visualizer - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access), archive of US and global bathymetric surveys with visualizer - General Bathymetric Chart of the Oceans - [GEBCO Gridded Bathymetry Data](https://www.gebco.net/data-products/gridded-bathymetry-data), global map interface of compiled bathymetry - Global Multi-Resolution Topography - [GMRT data](https://www.gmrt.org/), global compilation of multibeam data (includes graphical map tool) - **Resources to learn more:** - [NOAA - What is bathymetry?](https://oceanservice.noaa.gov/facts/bathymetry.html) - [Seabed 2030](https://seabed2030.org/) – global initiative to map the entire seafloor by 2030 - [Using open-access mapping interfaces to advance deep ocean understanding](https://link.springer.com/article/10.1007/s40012-025-00410-2) (Johannes, 2025) - [GeoMapApp](https://www.geomapapp.org/) - free map-based application for browsing, visualizing and analyzing a diverse suite of curated global and regional geoscience data sets  *Plumes of bubbles emanating from the seafloor, indicating that there were methane gas seep ecosystems in this region. Sound waves reflect strongly off the gas bubbles and are visible in the water column data. Source: <https://nautiluslive.org/blog/2018/08/08/more-just-bathymetry-seafloor-mapping-tool-exploration>* #### 2.3.2 Water Column Mapping - **About:** - Acoustic systems—usually multibeam echosounders or special water column sonars—to detect and visualize features suspended in the ocean between the surface and the seafloor - Key for detecting gas plumes, biological layers (schools of fish or migrations in the twilight zone), suspended sediments, etc. - **Use cases:** - Observing midwater scattering layers (biological migrations) - Detecting hydrothermal plumes amd tracking gas plumes from methane seeps - **Data collection:** - Most multibeam systems include water column data modes, so it is often collected in tandem with bathymetry - From ships, AUVs, and ROVs - Data must be interpreted alongside oceanographic profiles (e.g., CTD casts) and often requires manual cleaning to reduce noise - **Key considerations:** - Processing and interpreting water column data is a bit tedious not yet standardized - Detection is sensitive to sonar frequency, range, and sea conditions - Validation with ground-truth sampling (e.g., bottle casts, nets, sensors) is helpful - **Data sources:** - European Marine Observation and Data Network - [EMODnet Physics](https://emodnet.ec.europa.eu/en/physics), includes some water column data layers - NOAA National Centers for Environmental Information - [NCEI water column sonar data](https://www.ncei.noaa.gov/maps/water-column-sonar/) - **Resources to learn more:** - [OET - More than just Bathymetry](https://nautiluslive.org/blog/2018/08/08/more-just-bathymetry-seafloor-mapping-tool-exploration) - Seafloor Mapping as a Tool for Exploration - [Migration in the Ocean Twilight Zone](https://twilightzone.whoi.edu/explore-the-otz/migration/) - often monitored with water column data #### 2.3.3 Seafloor Backscatter - **About:** - Analyzing the intensity of sound that is reflected or ‘scattered back’ from the seafloor when using sonar systems - Provides information about seafloor texture, hardness, and composition (e.g., sand, rock, mud) - Often conducted simultaneously with bathymetric mapping during ship-based or AUV surveys - **Use cases:** - Seafloor habitat classification or substrate mapping - Detecting anthropogenic objects or features (e.g., cables, wrecks) - Complements bathymetry for geologic or habitat models - **Data collection:** - Similar to bathymetry and water column mapping, backscatter data is collected using the same sonar systems and processed using similar software - **Key considerations:** - Requires calibration and post-processing to produce usable mosaics - Interpretation of sediment type from backscatter typically should be verified by ground-truth sampling (e.g., grabs, cores with ROVs or HOVs) - **Data sources:** - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access) - **Resources to learn more:** - NOAA - [How does backscatter help us understand the sea floor?](https://oceanservice.noaa.gov/facts/backscatter.html)  *Sediment layers seen in the sub-bottom profiler data collected in 2021 at the New England and Corner Rise Seamounts expedition on the NOAA Ship 'Okeanos Explorer'. Source: <https://oceanexplorer.noaa.gov/technology/sub-bottom-profiler/sub-bottom-profiler.html>* #### 2.3.4 Sub-bottom Profiling - **About:** - Uses low-frequency acoustic pulses to penetrate below the seabed and image sediment layers or other buried geologic features - Reveals vertical structures beneath the seafloor - Typically deployed from research vessels or towed systems - **Use cases:** - Studying sedimentation and geological processes - Locating subseafloor gas pockets or archaeological sites - For infrastructure planning or hazard assessment (e.g., submarine landslides) - **Data collection:** - Chirp profilers (high resolution, shallow penetration) and boomer/sparker systems (deeper penetration) are used - Operated from vessels with sonar equipment often collected simultaneously while collecting bathymetry and other mapping data, even if the sonar systems are different - **Key considerations:** - Resolution and penetration are inversely related (deeper = less detail) - Can be noisy and hard to interpret without ground-truthing (e.g., sediment cores) - **Data sources:** - NOAA National Centers for Environmental Information - [NCEI data access](https://www.ncei.noaa.gov/access) - European Marine Observation and Data Network - [EMODnet Geology](https://emodnet.ec.europa.eu/en/geology), includes sub-bottom and other forms of seafloor geological data - **Resources to learn more:** - NOAA - [Sub-Bottom Profiler](https://oceanexplorer.noaa.gov/technology/sub-bottom-profiler/sub-bottom-profiler.html)  *3-D reconstructed seafloor lava flows and hydrothermal vent field from the East Pacific Rise. This 3-D model was produced using downward facing video imagery and photogrammetry techniques. Credit: Mae Lubetkin* #### 2.3.5 Photogrammetry and 3-D Reconstruction - **About:** - Stitching together overlapping images or video frames from subsea camera systems often mounted on ROVs or AUVs - To create detailed mosaics or 3-D models of seafloor features - Uses optical data, offering true-color, high-resolution imagery (unlike acoustic mapping techniques described above) - **Use cases:** - Mapping hydrothermal vent fields, coral reefs, archaeological sites, etc. - Change detection in dynamic environments (e.g., volcanic or vent habitats, biological growth or loss) - Public engagement and educational tools - **Data collection:** - Collected by vehicle-mounted cameras with precise navigation and positioning - Software like Agisoft Metashape or custom photogrammetry pipelines are used for processing (which is easier now than ever before, becoming much more common in ocean sciences) - **Key considerations:** - Requires good lighting and water clarity - Processing is computationally intensive, and vehicle navigation data helps with plotting 3-D reconstructions onto broader bathymetric maps - Can be limited to small survey areas due to time constraints and battery limitations - **Data sources:** - Monterey Bay Aquarium Research Institute - [MBARI Sketchfab](https://sketchfab.com/mbari) - 3-D models of seafloor sites can be found in academic papers or at individual institutions or government agencies data repositories - **Resources to learn more:** - [Seafloor Futures](https://garden.ocean-archive.org/seafloor-futures/) (Lubetkin, 2024) - [Realtime Underwater Modeling and Immersion](https://nautiluslive.org/tech/realtime-underwater-modeling-and-immersion) - Ocean Exploration Trust - [Underwater 3-D Reconstruction from Video or Still Imagery: Matisse and 3-D Metrics Processing and Exploitation Software](https://www.mdpi.com/2077-1312/11/5/985) (Arnaubec et al., 2023) - [Seeing the Sea in 3-D](https://schmidtocean.org/cruise-log-post/seeing-the-sea-in-3d/) - Schmidt Ocean Institute ### 2.4 Satellite Remote Sensing Satellite data provides the most familiar, spatially complete picture of the ocean. This bird’s-eye perspective is invaluable for understanding large-scale phenomena like currents, sea surface temperature patterns, and phytoplankton blooms, providing visual evidence that can enhance storytelling. However, there are unique considerations since, unlike the use of satellite imagery on land, most of our understanding of the ocean does not come from the visual spectrum. In this section, we’ll introduce three of the most important types of satellite ocean data and discuss the use cases for each. - **History:** - 1978: NASA launched Seasat, the first satellite designed for ocean research. - Significant expansion in the 1990s with missions including TOPEX/Poseidon (ocean altimetry), AVHRR (high-resolution sea surface temperature), and SeaWiFS (ocean biology). - Modern constellations are operated by NASA, NOAA, ESA, EUMETSAT, CNES, ISRO, and others. - **Trade-offs:** - Excellent spatial coverage that’s impossible to achieve with ships or buoys. - Very high costs, these platforms are operated by government agencies. - Only seeing the very surface of the ocean, no subsurface data. - Limited horizontal resolution (spatial detail) and temporal resolution (orbital repeat time).  *Gulf of Mexico SST on a cloud-free data. Source: <https://marine.rutgers.edu/cool/data/satellites/imagery/>* #### 2.4.1 Radiometry - Sea Surface Temperature (SST) - **About:** - Sea surface temperature (SST) is the oldest and most extensive application of satellite oceanography. - **Use cases:** - Tracking climate change, El Niño, and marine heat waves. - Key input for weather models (e.g., very important for hurricane forecasting). - Mapping ocean eddies, currents, and upwelling, which are critical to fisheries. - **Data collection:** - Two separate types of sensors measure SST: Infrared and microwave. - IR sensors have higher spatial resolution ([1-4 km](https://coastwatch.noaa.gov/cwn/product-families/sea-surface-temperature.html)) and finer temporal coverage but cannot “see” through clouds, which block over 70% of the ocean at any given time. - Microwave sensors can see through most non-precipitating clouds but have a lower spatial resolution (about 25 km) and don’t work near the coastline. - Measures temperature of the top ~1 mm of the ocean - Blended products: Combine multiple sensors for better coverage (e.g., GHRSST L4)  *Example SST data at different processing levels. (Merchant et al. 2019): <https://www.nature.com/articles/s41597-019-0236-x>* - **Key considerations:** - Make yourself aware of the different processing levels when accessing data. Level 4 (L4) will be the easiest to work with but may not be fully accurate. - L2: Data along the original orbital track. - L3: Gridded data, sometimes averaged over time. - L4: Cloud-free, gaps are filled by various methods depending on the source. - Temporal resolution depends on the satellite orbit. There are good options that blend multiple satellites. - **Data sources:** - [EU Copernicus Marine Service](https://marine.copernicus.eu/) - [US NASA Physical Oceanography DAAC](https://podaac.jpl.nasa.gov/) - [NOAA CoastWatch](https://coastwatch.noaa.gov/cw_html/cwViewer.html) - Graphical Interface - **Resources to learn more:** - [Group for High Resolution Sea Surface Temperature (GHRSST)](https://www.ghrsst.org/ghrsst-data-services/for-sst-data-users/) #### 2.4.2 Radar Altimetry - Sea Surface Height (SSH) - **About:** - Measures ocean surface height by sending radio pulses and measuring return time. - SSH can tell us the strength of large scale currents like the Gulf Stream, as the slope of the sea surface is used to calculate the “geostrophic current”. - **Use cases:** - Key to understanding ocean circulation and long-term sea level rise. - **Data collection:** - Radar altimeters on satellites measure SSH directly (e.g., Jason-3, Sentinel-6) and then the geostrophic currents are calculated in a post-processing step. - Spatial resolution is significantly worse than SST (25+ km) - The recent SWOT satellite is a new type of altimeter with much higher resolution but has very limited coverage since there is only one currently in orbit. - **Key considerations:** - SSH is useful for large-scale ocean currents but not coastal tidal currents. - Similar to SST, be careful about processing level and look for re-gridded datasets. - Can generally see through clouds, so gaps are not a significant issue. - **Data sources:** - [EU Copernicus Marine Service](https://data.marine.copernicus.eu/products?facets=mainVariables%7ESea+surface+height--sources%7ESatellite+observations) and [Aviso](https://www.aviso.altimetry.fr/en/data/products/sea-surface-height-products/global.html) - [US NASA PODAAC](https://podaac.jpl.nasa.gov/NASA-SSH) - [Copernicus MyOcean Pro](https://data.marine.copernicus.eu/viewer/expert) and [Aviso](https://seewater.aviso.altimetry.fr/) - Graphical Interfaces - **Resources to learn more:** - [NASA JPL - What is Ocean Surface Topography?](https://podaac.jpl.nasa.gov/OceanSurfaceTopography)  *Global map of marine Chlorophyll concentration. Source: <https://sos.noaa.gov/catalog/datasets/biosphere-marine-chlorophyll-concentration/>* #### 2.4.3 Optical - “Ocean Color” - **About:** - Ocean color sensors measure the reflectance of sunlight from the ocean surface to infer biological and chemical properties, such as algal concentration, suspended sediments, and water clarity. - **Use cases:** - Tracking phytoplankton blooms and changes in marine ecosystems. - Useful for monitoring water quality, including coastal sediment and oil spills. - **Data collection:** - Sensors measure light reflected from the ocean at different wavelengths (e.g., MODIS, VIIRS, Sentinel-3 OLCI) and then apply algorithms in order to calculate variables such as Chlorophyll-a concentration. - **Key considerations:** - Ocean color data is significantly affected by cloud cover, aerosols, and atmospheric correction errors. - **Data sources:** - [EU Copernicus Marine Service](https://data.marine.copernicus.eu/products?facets=mainVariables%7EPlankton--sources%7ESatellite+observations) - [US NASA Ocean Color Web](https://oceancolor.gsfc.nasa.gov/data/find-data/) - [NASA Worldview](https://worldview.earthdata.nasa.gov/) - Graphical Interface - **Resources to learn more:** - [IOCCG (International Ocean-Color Coordinating Group)](https://ioccg.org/) ### 2.5 Additional databases and scientific support The four sub-sections above (*In Situ* Sensors; Deep Ocean Observation, Exploration, and Research Systems; Mapping; Satellite Remote Sensing) cover the main areas of ocean scientific data types and collection methods. There are some datasets that are not discussed in this guide since they are likely less useful for investigative storytelling or require technical skills to access and interpret the data. Below are some additional databases and information on contacting scientists to support your investigation. While in section 3, we outline a case study using real data to tell an ocean story. #### 2.5.1 Additional databases, collections, and visualizers Sites that either did not fit into one of the sub-sections above, or that contain information which is generated after scientific studies occur: - [PANGAEA](https://pangaea.de/) - data publisher for earth and environmental science (across disciplines) - [International Seabed Authority DeepData](https://www.isa.org.jm/deepdata-database/) - database hosting all data related to international deep-seabed activities, particularly those collected by contractors (i.e. nations or entities) during their exploration activities and other relevant environmental and resources-related data. Includes a dashboard and map to search for basic stats and information about what contractors have done during deep seabed mining exploration cruises. - [Marine Geoscience Data System](https://www.notion.so/Ocean-Datasets-for-Investigations-1caf92221af780c68873c2aecf9b3479?pvs=21) - geology and geophysical research data across collections - [USGS Earthquake Hazards Program](https://www.usgs.gov/programs/earthquake-hazards/earthquakes) - interactive map with magnitudes and additional information (earthquakes can occur on land and in the ocean) - [WoRMS – World Register of Marine Species](https://www.marinespecies.org/) - comprehensive taxonomic list of marine organism names - [OBIS – Ocean Biodiversity Information System](https://obis.org/) - global open-access data and information on marine biodiversity - [Windy](https://www.windy.com/) - animated weather maps, radar, waves and spot forecasts #### 2.5.2 Scientific support All of the datasets and databases we outlined above are free and open to the public. We hope that we outlined enough context and links to user-friendly platforms to access the data so that you feel empowered to conduct your own investigations with ocean datasets. That said, some data might be more challenging to work with depending on prior experience and computing skills, among other factors. When in doubt, you can always contact an ocean scientist to ask questions or seek support. Depending on your investigation or story, you will need to contact a specific type of ocean scientist since each has their own specialty. You can start by searching for and contacting scientists at nearby universities or research institutes. **Ocean scientists and their specializations:** - **Physical Oceanographers -** Study ocean currents, tides, waves, and ocean-atmosphere interactions. They can help explain phenomena like sea level rise or how ocean circulation affects weather and climate. - **Chemical Oceanographers -** Focus on the chemical composition of seawater and how it changes over time. Useful for stories involving ocean acidification, pollution, nutrient cycling, or chemical runoff impacts. - **Biological Oceanographers or Marine Biologists -** Study marine organisms and their interactions with the ocean environment. They are ideal sources for stories on biodiversity, fisheries, invasive species, and ecosystem health. - **Geological Oceanographers or Marine Geologists -** Study the structure and composition of the ocean floor. They can provide insights into underwater earthquakes, tsunamis, deep-sea mining, or the formation of underwater features. - **Climate Scientists with Ocean Expertise -** Examine how oceans influence and respond to climate change. They are helpful for broader climate stories that involve ocean heat content, carbon storage, or long-term trends in ocean conditions. - **Marine Ecologists -** Study relationships among marine organisms and their environment. They can clarify ecosystem-level impacts, like those from overfishing, coral bleaching, or marine protected areas. - **Fisheries Scientists -** Specialize in fish populations, fishing practices, and resource management. Helpful for reporting on commercial fishing, stock assessments, or policy/regulation issues. - **Ocean Data Scientists -** Work with large marine datasets and modeling, can assist with interpreting satellite data, ocean models, or big datasets. - **Marine Policy Experts and Ocean Economists -** Focus on the intersection of ocean science, law, and economics. Helpful for coverage of marine regulations, governance issues, or the ‘blue economy.’ - **Marine Technologists or Ocean Engineers -** Design and use tools like underwater drones, sensors, and buoys. They can help explain how ocean data is collected and what the limitations of certain technologies might be. ## 3. Case Study: Gulf of Maine Ocean Warming  As ocean scientists from the Northeastern United States, we have each witnessed how rapid ocean changes are affecting the ecosystems and communities around us. For this example case study, we focus on the Gulf of Maine—a region close to home. When telling ocean stories, it is helpful to have either first-hand or personal connections to the coastal or oceanic region you are investigating. ### 3.1 Motivation The Gulf of Maine is warming [faster than 99% of the global ocean](https://eos.org/features/why-is-the-gulf-of-maine-warming-faster-than-99-of-the-ocean), making it a key site to investigate local impacts of climate change on marine environments and coastal livelihoods. Stories of changing fish stocks and stressed fisheries are already discussed in communities within the Northeastern region. Before getting into the data, we will think through the historical and ecological context of the Gulf of Maine and its fisheries. For centuries, the regional identity has been deeply linked to the ocean. Indigenous [Wabanaki peoples](https://www.wabanakialliance.com/wabanaki-history/)—including the Abenaki, Mi'kmaq, Maliseet, Passamaquoddy, and Penobscot nations—relied on these coastal waters for food as well as cultural practices and trade. They managed their coastal and marine environments with ocean knowledge developed across generations. When European colonization began, [intensive cod fishing fueled transatlantic trade and early settlements](https://www.markkurlansky.com/books/cod-a-biography-of-the-fish-that-changed-the-world/). Europeans considered the cod fisheries to be so abundant that they were endless. The overfishing by settlers caused a massive collapse in cod stocks by the 1950s. Now, other important local fisheries like the American lobster are being impacted by the combination of ocean warming and historic overfishing. Harmful algal blooms have also increased in frequency which indicate that the broader Gulf ecosystems are under stress. In the following sections, we guide you through using publicly available ocean datasets to investigate the scientific questions behind the Gulf of Maine and its warming waters. By accessing current and archival datasets you will be able to visually show the seawater temperatures going up and connect that to other environmental stories or investigations about the Gulf. ### 3.2 Data acquisition In order to investigate warming in the Gulf of Maine, we will analyze surface temperatures from two different datasets: a local *in situ* temperature sensor and the global-average SST. With the global SST as our baseline, we’ll be able to determine how much faster the Gulf of Maine is warming compared to the rest of the world. This analysis involves database downloads, data post-processing/analysis, and data visualization. If you don’t have experience with coding and want to get started with the Python programming language, see the appendix for tips on getting setup. Otherwise, you can always consider contacting a scientist to support you with your investigation (see section 2.5.2 Scientific support). #### 3.2.1 Gulf of Maine buoy temperature dataset First, we’ll go to the [National Data Buoy Center (NDBC)](https://www.ndbc.noaa.gov/) website and look for a buoy in the Gulf of Maine with a long historical record of temperature measurements. Clicking on the “Historical Data & Climatic Summaries” link at the bottom of [Station 44007’s page](https://www.ndbc.noaa.gov/station_page.php?station=44007) reveals annual text files going back to 1982.  *Screenshot from the NDBC website showing potential buoys to use for Gulf of Maine case study.* The task now is to process all of this data into a more usable format. We’ll do this with a python script using the [pandas](https://pandas.pydata.org/) data analysis library. 1. Loop through the years 1982-2024 and create the dataset url for each year, using the NDBC website to deduce the url structure. 2. Load the text data directly from each url via `pandas.read_csv()` 3. Convert the year, month, day, hour columns into a single pandas datetime column. 4. Combine all of the data into a single dataframe. 5. Save our data to a new CSV file.  *An example of what the available buoy data looks like for the year 1985. The highlighted sections show the parts of the dataset that we’re interested in: the date/time and the water temperature.* #### 3.2.2 Global mean SST dataset Next, we want a corresponding dataset for the globally-averaged SST, in order to determine whether the Gulf of Maine is warming faster or slower than the average. The [Climate Reanalyzer](https://climatereanalyzer.org/clim/sst_daily/?dm_id=world2) displays globally-averaged SST from the [NOAA 1/4° Daily Optimum Interpolation Sea Surface Temperature (OISST)](https://www.ncei.noaa.gov/products/optimum-interpolation-sst), a long term Climate Data Record that incorporates observations from different platforms (satellites, ships, buoys and Argo floats) into a regular global grid. 1/4° refers to the grid resolution—about 25 km. There is an option to download the underlying data from the Climate Reanalyzer website, which will save us a lot of time vs. trying to access decades of data and doing the global averaging ourselves. The data is available as a JSON file, which is a different text file format that will require a more custom approach for converting into a pandas dataframe.  *Screenshot of the [Climate Reanalyzer](https://climatereanalyzer.org/clim/sst_daily/?dm_id=world2) website. In the dropdown menu, we want to download the JSON data.* One key concept to note is how this data handles dates. Each year includes a list of 366 temperatures, without any explicit list of the corresponding dates. This is using the format of “day of year” and we see that the last temperature is “null” for non-leap years. When processing this data, we need to take this into account and ignore the null final value. Similar to the buoy data, we’ll re-format this data in a pandas dataframe and save to a new CSV file.  *A look inside the globally-averaged SST JSON file. The data is arranged as a list of years where each year has a list of 366 temperatures.* ### 3.3 Climatological data analysis A standard method for analyzing climate change anomalies is to first remove the climatological “seasonal” signal from the data. This will allow us to show, for each data point, how much warmer or colder it was than the average temperature for that day of the year. The first step is choosing which time period we’ll use for our climatology “baseline”. Here we’ve chosen 1991 to 2020 since it is fully covered by our data and matches the climatology period used by the Climate Reanalyzer website. Next, we’ll use some built-in pandas methods to get the climatological average temperature for each day and then map that to each datapoint in our timeseries. The following code snippet shows the steps used for both the buoy data and the global SST: ```python # Select just the data in the range of the climatology period df_clim = df[(df.index.year >= 1991) & (df.index.year <= 2020)].copy() # Assign the day of year (1-366) to each data point in the timeseris df_clim["day_of_year"] = df_clim.index.dayofyear # Take the mean for each day_of_year df_clim = df_clim.groupby("day_of_year")["temp"].mean() # New variable in df: the climatological temperature for that day df["day_of_year"] = df.index.dayofyear df["climatology_value"] = df["day_of_year"].map(df_clim) # Temperature anomaly is observed temperature minus climatological temperature df["anomaly"] = df["temp"] - df["climatology_value"] ```  *Our resulting dataframe includes new columns for climatology and temperature anomaly.* ### 3.4 Analyzing and Visualizing the results First, we’ll simply plot the full temperature timeseries and see what we find.  The warming signal is instantly apparent in the global SST data because the seasonal signal is so small. The Gulf of Maine, however, varies by more than 15° C throughout the year so any long term changes are difficult to see in this format. Plotting the climatology signal illustrates this point (pay attention to the y-axis).  Next we’ll view our temperature anomaly data (observed temperature minus climatology). As expected, there is more noise in the buoy data since it’s taken from a single point and any given day can vary by as much as 4 °C from climatology. The globally-averaged temperature has much less variance.  For the final version of our plot, we’re incorporate 3 changes: 1. Fit a simple linear regression using [numpy’s polyfit](https://numpy.org/doc/stable/reference/generated/numpy.polyfit.html) in order to quantify the average rate of warming for the two datasets. 2. Plot the monthly averages instead of the daily values in order to simplify the visual clutter. 3. Use the same y-axis range for the two plots for direct visual comparison.  Comparing our warming rate calculations against the published literature finds good agreement: - Gulf of Maine SST: our rate of 0.496°C/decade is within 5% of the 0.47°C/decade reported by the [Gulf of Maine Research Institute](https://gmri.org/stories/2024-gulf-of-maine-warming-update/). This is likely due to differences in methods—we used a single buoy and they used the OISST data averaged across the entire Gulf. - For global SST, our rate of 0.188 °C/decade is within 5% of the 0.18 °C/decade (over the past 50 years) published by [Samset et al. (2023)](https://www.nature.com/articles/s43247-023-01061-4). These final plots provide simple visual evidence of the Gulf of Maine’s rapid warming over the past 40 years. We showed the data transform from text files, to noisy timeseries, and finally to expert-validated trend lines. By removing the strong seasonal signal and focusing on the anomalies, we can clearly see the long-term warming trend in both the Gulf of Maine buoy data and the global mean SST. Finally, note that the linear regression is useful for quantifying the recent warming in an easily understandable number but is not necessarily a predictor of future warming. The [Maine Climate Science Dashboard](https://climatecouncil.maine.gov/future/climate/science-dashboard) shows the potential for human emissions to either accelerate or slow down this rapid warming.  *The [Maine Climate Science Dashboard](https://climatecouncil.maine.gov/future/climate/science-dashboard) combines historical water temperature measurements with different climate scenario forecasts.* ## 4. Conclusion Our investigation into Gulf of Maine temperatures, using readily available public datasets, highlights one local manifestation of global climate change. This rapid warming isn't merely an abstract data point, it continues to have profound implications for the region’s biodiversity and the human communities who rely on the ocean. Marine species are highly sensitive to temperature changes, and the Gulf of Maine has been experiencing a noteworthy decline in native species and [increase in warmer-water species](https://online.ucpress.edu/elementa/article/9/1/00076/118284/Climate-impacts-on-the-Gulf-of-Maine-ecosystemA). The next steps in this story might look to other data sources to explore: Why is the Gulf of Maine warming so quickly? and What will the region look like in the future? or How exactly are local fisheries affected by warming waters? This case study is one example of how to find the connections between global environmental change, local ocean data, and tangible human impacts. This process offers a template for investigating similar stories in your own regions: 1. **Start with a local observation or community concern:** What are people witnessing or experiencing in your local environment? 2. **Explore the scientific context:** Consult with scientists, read relevant research, and understand the underlying environmental drivers. 3. **Seek out publicly available data:** As shown in section 2, there is a large assortment of high-quality public ocean datasets that can be used to investigate countless questions. 4. **Connect the data back to human issues:** How do the environmental changes revealed by the data affect local cultures, livelihoods, health, and economies? The key thing to remember is that there are multiple angles to uncover and expose often-invisible impacts to the ocean. Datasets provide one lens to report on climate and environmental changes, but these stories impact communities and are thus both political and social. Just as ocean science has changed and begun to decolonize, it's crucial to investigate and tell stories that reflect diverse experiences. Ocean data can help highlight intersecting issues—such as deep seabed mining, marine health, and colonial continuums—with evidence-based information and compelling visualizations. We hope this guide offers a practical starting point for navigating ocean science, accessing and interpreting data, and connecting your investigation to real-world consequences that are planetary in scale yet intimately local. ```cik-note ``` >**APPENDIX: Getting started with python** > >If you have not done any coding before, the initial task of setting up your coding environment can be a challenging hurdle. There are multiple options for code editors/IDEs (integrating development environment), ways of handling dependencies (the external packages you install to give you advanced functionality), and other decisions that are outside the scope of this article. Luckily, once you’ve chosen your tools, there are good resources online so here are a few recommendations and then you can seek out more detailed tutorials: > >1. Use Visual Studio Code as your code editor (the application where you will write and run code). This is the most popular option and there is an extensive ecosystem of 3rd party plug-ins and help resources. <https://code.visualstudio.com/> > >2. Use [conda](https://docs.conda.io/projects/conda/en/latest/user-guide/getting-started.html) for package management. Your computer’s operating system may come with a version of python pre-installed but it's not a good idea to install packages onto this global location. Instead, create separate conda "environments" for different projects. This will allow you to experiment in a safe and organized way. Here is a helpful article on the VS Code website: <https://code.visualstudio.com/docs/python/environments>. For example, to create a new environment that we'll name "ocean-study" and install the "matplotlib" plotting package would look like this: > > ```bash > conda create -n ocean-study > conda activate ocean-study > conda install matplotlib > ``` > Now, in VS Code, just make sure your Python Interpreter is using this environment (it will look something like `~/miniconda3/envs/ocean-study/bin/python` and you will be able to use the matplotlib package in your code. > >3. Finally, consider using Jupyter Notebooks for exploratory coding where you’re loading datasets and making plots. Notebooks have the file extension `.ipynb` and allow you to run chunks of code independently in code "cells" and view the output right below. You can also use Markdown text cells to write notes and explanations for yourself and collaborators. Instructions on using VS Code: <https://code.visualstudio.com/docs/datascience/jupyter-notebooks> > <hr class="thick"> ##### About the authors **Mae Lubetkin** is an ocean scientist, transmedia artist, and writer based in Paris and at sea. Their practice-led research remaps our relations to bodies of water and digital worlds by means of investigation, counter-narrative, and memory. With a background in marine geology and subsea imaging, their artistic practice is in dialogue with Science while situated in queer, intersectional, anti-extractivist, and decolonial frameworks. Guided by wet-techno-critical studies and thinking with other-than-human worlds, they compose environmental traces in installations and digital outputs. Their core practice is in solidarity with submerged, ancient, ephemeral and imaginary environments. **Dr. Kevin Rosa** is an oceanographer and the founder of Current Lab, a startup specializing in computational ocean forecasting. He holds a B.A. in Physics and a PhD in Physical Oceanography from the University of Rhode Island, with a focus on ocean physics and hydrodynamic modeling. <hr class="thick"> *Published in June 2025* |
source/en/the-kit-en.json” file was changed. 4 months ago