The OpenCTD: Open-source Oceanography for Everyone

Below is a transcript and slides from the above talk, delivered at the October 19, 2023 GOSH Community Call.

Good afternoon, good evening, and good morning, and thank you for inviting me.

Access to the tools of science is rarely equitable, and nowhere is this inequality of access more pronounced than in the ocean sciences, where all but a few entities have the capital to mount major oceanographic research campaigns. I come from the world of deep-sea ecology, where our budgets can quickly climb into the tens of millions of dollars. But even small-scale coastal research can be stymied by the need for vessels, equipment, and instruments, access to which is often controlled by research institutions. For ocean knowledge seekers who lack academic credentials or significant financial resources, accessing the fundamental tools of marine science can represent an insurmountable barrier.

This is a huge problem. As the need to understand the dramatic changes happening both at the surface and beneath the waves accelerates, barriers to access that precludes the participation of the full breadth of ocean stakeholders erodes our potential to understand, anticipate, and mitigate those changes.

I believe that the ocean belongs to everyone and that the tools to study the ocean should be available to anyone with the curiosity and motivation to pursue that inquiry.

Chief among those tools is the workhorse of oceanography, the CTD.

A CTD is a device that measures salinity, temperature, and depth. These are the fundamental parameters that you need to conduct almost any ocean science study. It can be used to profile a water column, affixed to other instruments to correlate observations to water conditions, or deployed as a fixed mooring for long term monitoring. Almost all marine scientific research includes a CTD cast. Ecologists need to understand the environment that their study species live in. Aquaculturists need to continuously monitor the water quality around their fish or shellfish pens. Conservation workers need to document changes in water health and quality. Climatologists, SONAR technicians, lifeguards, documentarians, fisherman, river keepers, and even the kid taking a swim in the Chesapeake Bay who just wants to know if the sea nettle season is over, all benefit from CTD data.

But commercial CTDs are expensive. Handheld commercial units cost upwards of $6000. Larger systems can easily climb into the six-figures. This creates a barrier to access for many of the people most imminently affected by our changing oceans.

The OpenCTD is a low-cost, open-source alternative to commercial CTDs designed for budget-restricted scientists, educators, and practitioners working in nearshore coastal ecosystems, where entire research projects can be conducted for less than the cost of a commercial CTD. The sensor quality is acceptable for the majority of ecology and conservation studies – you probably won’t want to use one for chemical or physical oceanography, but it is well within the accuracy and precision range for ecology, conservation, and environmental monitoring – and it can operate to a depth of 140 meters.

An OpenCTD, in contrast to commercial units, can be built for about $350 in parts and another $200 in tools and consumables.

Getting into the guts of it, the brain of the OpenCTD is an Arduino microcontroller that talks to a real-time clock, which timestamps all the data, as well as an SD card reader where all the data is logged. All these components, when combined with a commercial chip to interpret salinity and a 3.7-volt lithium polymer battery form the control unit. The control unit then interfaces with a sensor package that includes a battery of three temperature sensors, an absolute pressure sensor which provides depth, and a graphite probe which measures conductivity, which is then translated to salinity. Everything is housed within a very boring PVC pipe and the sensors are fully potted in epoxy. The open end is capped with an off-the-shelf pressure cap that tradespeople use to test pipework but that we’ve found will hold pressure down to 140 meters. This means that a full watertight enclosure can be constructed for about $8. In comparison, a housing of similar quality from Blue Robotics, which is a fantastic company that makes excellent housings at fair prices,  costs about  $117.

But that’s not the entire OpenCTD story.

From the beginning, the vision of the OpenCTD was that it would be built by the end user. We a guide that walks users through the process of sourcing materials, downloading software, 3D printing components, assembling electronics, and calibrating the finished instrument. This helps alleviate the other major hurdle to operating an oceanographic program – commercial equipment requires an ongoing relationship with the manufacturer and service contracts over a 5- or 10-year period can dramatically exceed the original cost of the device. So not only does the OpenCTD provide a device at relatively low cost, but it also provides the expertise necessary for the creation of local capacity. OpenCTD users can service and maintain their own equipment, repair damaged units, and even build new ones as needed, entirely independent from the OpenCTD team.

So, anyone can build their own CTD, but the major way that we’ve kept this program funded is through hosting training workshops to teach students how to build the instrument. Over the last few years, I’ve hosted CTD building workshops local to me in Maryland, with marine educators at Stellwagen Bank National Marine Sanctuary and, remotely, with middle and high school students in Homer, Alaska. For students, the process of building an OpenCTD offers an introduction to coding, 3D-printing, hardware prototyping, electronics and can provide a practical foundation for courses in earth science and marine or environmental science. Students come out of these workshops feeling a sense of accomplishment and their newfound understanding for how data is produced promotes a greater appreciation for the process of science.

A student with no prior experience in electronics, soldering, coding, or fabricating can build, calibrate, and deploy an OpenCTD over a long weekend.

This approach highlights the core principles of the OpenCTD program. We want our device to be as inexpensive as possible while still producing high quality data. We want everything we possibly can to be released under an open-source license, so that anyone can take the base OpenCTD and expand, iterate, and adapt it to their needs. And we want the materials we use to be as accessible as possible, so that people who want to built a CTD can find the parts they need in their local hardware and electronics stores, and from major online retailers.

And we have another value: There’s this phenomenon called Parachute Science, which is when researchers from rich, usually western, countries drop into poorer countries, collect data to enrich their careers, and then disappear, providing no material support for the people and communities most directly connected to that data. A similar concept exists in the conservation technology world, where wealthy developers devise one-size fits all technological solutions that are then deployed to much fanfare in places that may neither need nor want those tools and lack the capacity or desire to maintain them. This phenomenon is called Imported Magic.

We want the OpenCTD to reach beyond Imported Magic, to promote ownership over not just data, but over the tools and skills to acquire that data. We focus on conservation technology that promotes ownership and data independence within communities so that research priorities can be set by knowledge seekers, not funders, and reflect and respect their specific needs and values.

But absolutely none of this matters if the data is no good. Salinity, for example, is an incredibly tricky thing to measure, not the least because, as researchers, we don’t really know what we’re measuring when we measure salinity. Which is why we no longer report salinity in parts per thousand when we don’t know what those parts are, but in the unitless Practical Salinity Unit. Over the course of 2023, we’ve conducted a series of head-to-head tests between OpenCTDs and both precision commercial benchtop probes and NOAA data buoys to assess not just the quality of OpenCTD data but the post-calibration sensor drift. This chart shows one test, where OpenCTDs calibrated using a variety of different calibration protocols designed to work under different environmental conditions, all fell within a 5% margin of error from the commercial benchtop instrument, and the CTD calibrated using our recommended calibration protocol fell within 1% of the commercial unit. 

And we see similar results in the field. Last year we sent OpenCTDs to naturalists working aboard commercial whale watching boats throughout New England to log and upload opportunistic data while bringing tourists out to view whales. This is a great example of a natural partnership because whale watching boats, at least in the US, are required to stop moving when whales are present, giving them a window of time for naturalists to talk about whales and whale conservation, and conduct a CTD cast to collect primary data during this pause. Over 2021 and 2022, we collected more than 50 CTD casts from along the North Atlantic seaboard and were able to validate their data against fixed NOAA monitoring buoys. These are areas of the coast that are highly trafficked, but rarely do researchers collect oceanographic data at this level from these areas. This kind of data can be invaluable as a baseline or for detecting local changes in oceanographic patterns.

And when we looked at sensor drift after these heavily used and abused CTDs came back to the lab, we found that even after 18 months and dozens of casts, the sensors held their original calibration with minimal drift.

I’m going to leave the thank you slide up while I talk about what’s next. We finally have a paper in review validating the OpenCTD data, which will help us establish greater trust with the formal scientific community. We’re imminently about the issue the fifth edition of the OpenCTD construction and operation manual after spending the summer reviewing every step of the workflow. We’re working on developing workshops in New England and Ghana for summer 2024. And we’re working on integrating a 3D printable Niskin bottle into the OpenCTD, so that researchers can take water samples at discrete depths. We’re also slowly developing a higher resolution OpenCTD with more precise sensors and a greater depth rating.

Thank you.

Southern Fried Science is free and ad-free. Southern Fried Science and the OpenCTD project are supported by funding from our Patreon Subscribers. If you value these resources, please consider contributing a few dollars to help keep the servers running and the coffee flowing. We have stickers.

Please review the comment policy before you reply.

This site uses Akismet to reduce spam. Learn how your comment data is processed.