by Yvonne Firing, NOC | June 27, 2025
Yvonne is a physical oceanographer at National Oceanography Centre in the United Kingdom. In the EuroGO-SHIP project, her work has focussed on expanding secondary quality control and serving as the salinity parameter expert to gather evidence on the effect of practical differences in procedures toward developing a framework for an updated salinity measurement best practice. In addition, Yvonne has been involved in pilot skills-sharing related to the calibration of salinity data through collaborations with other EuroGO-SHIP partners.
Best practices and quality control are key components of the EuroGO-SHIP project, so we asked Yvonne to help unpack these terms and clarify why they are incredibly important to ensure the delivery of much-needed climate data.
Why did you want to become a scientist and how did you end up focussing on best practices and quality control?
I was always interested in how things worked, that’s how I got started in physics and then in physical oceanography. My current focus is strongly on ocean observations, especially those made “in person” or in situ. And I think I navigated toward this partially because I found I really enjoyed going to sea, and partially by chance, because there were opportunities to work on projects where I would make these kinds of observations at sea.
Over time in my current job, I saw opportunities to improve some of the ways we did things, in particular some of our data processing and quality control tools (code). Then, I got more opportunities to apply the same approach or same skills to other projects. And rather than focusing on one science question or one region of the ocean, I’m now applying these tools and procedures to numerous questions and areas.
How do you go about understanding what a best practise is? Do you start with comparisons?
Yes, and this has been interesting. We wanted to compare different procedures based on the quality of the result, and compare the steps involved in a procedure to see how they contribute to or detract from the success or the quality. And we discovered that the first difficulty was people generally don’t write down exactly every step they do in a procedure. It is sort of intuitive, and one doesn’t think about it. Or they feel that it is an obvious step and doesn’t need to be recorded. And this leads to potential gaps between what has been recorded, or what the person trying to follow the procedure thinks needs to happen and what actually needs to happen when conducting the same type of measurement. Even from the same person, so if you think about trying to follow instructions that you wrote years earlier, this may sound familiar!
So how important is it that everyone is following the same procedure? It is about defining and adopting best practices or is it about quality control?
That is actually a very good question because in some cases, there might be multiple procedures that meet a certain standard for measurement quality. In oceanic hydrography, we don’t generally take a metrological approach. Instead, we may assess uncertainty by comparing different datasets. So, observers may not need to follow the exact same procedure if their methods produce the same results, based on these comparisons. And again, that acceptable level of uncertainty will be different for different uses of the data. On GO-SHIP cruises we take 10% replicate samples because the GO-SHIP community has worked out that is what is required given our target level of uncertainty. In theory we could take more replicates, or we could do things like run a certified reference to go with every single sample, but that would be beyond what we need.
But what we’re looking for in EuroGO-SHIP is to identify where you could easily get a better dataset, which can be by performing quality control on the measurements you have, or it could be by changing something about your procedure to get the measurements. So, we are not telling everyone to adopt all these standards and spend all this effort on making top quality measurements, we are asking, do you want to get this level of uncertainty? And if yes, then this is what you need to do.
And why is that overall important, that comparison?
One of the big questions that we try to answer with hydrography is what are the long-term changes in the ocean properties and ocean transports? Some of those changes are very small relative to other signals, relative to seasonal signals, or relative to mesoscale variability – oceanic weather. All the real ocean variability on different scales combines with the uncertainties that we know are introduced by our measurement systems, like the precision and accuracy of a sensor, as well as things we can’t always control, like something that goes wrong that you may or may not have noticed in your lab. But to be confident that we are detecting long-term trends, shifts, and cycles, we need to be sure that a measurement taken in 2008 can be compared to a measurement taken in 2020, even if they’re on different ships and by different people. So we approach the problem from two angles, first do we know what procedures were followed in each case and should they produce the same results – this is the best practices aspect — and second, how can we detect if we have an unexplained error in one of them, so we don’t mistake this for the signal we are looking for – this is the secondary quality control aspect.
What do you think the legacy of EuroGO-SHIP will be?
I hope the legacy of EuroGO-SHIP will be a robust structure, in some form, that will enable researchers to make better observations and more observations of ocean properties. Ocean physical and biogeochemical properties really tell us a lot about how the ocean works, and they set the environment for ocean ecosystems, so they are relevant not only for questions of climate and forecasting global change, but also for understanding regional and local ecosystems and what is happening there.
We have already seen improved observations as a result of the pilot best practices and skills-sharing work, which highlighted a difference in procedure we hadn’t realised was there, and which turned out to be pretty easy to fix. So, I hope awareness of some of these improvements will spread beyond the consortium to our wider community including different subfields, experts and non-experts.
Another example is, NOC has one of the UK’s GO-SHIP lines that we typically do every 10 years across the subpolar North Atlantic from Scotland to Greenland. But we don’t have the capability to do all the types of measurements we need to monitor this line. It was thanks to EuroGO-SHIP that, first, we spun up a collaboration with project partner NORCE to measure transient tracers, joining from Norway, and then we applied to another research infrastructure project, AQUARIUS, and now the cruise that will support this work, programmed for next summer, will be on the Celtic Explorer, one of the Irish vessels, and we’ve been talking with other EuroGO-SHIP partners about their observations in the area and potentially coordinating. So, I think this collaborative expedition is part of the legacy of EuroGO-SHIP already.
How has the project benefitted the wider community and what would happen if it just stopped?
If we didn’t have a follow-on of any sort, I don’t think it would come to a grinding halt because within our 14-partner consortium we have made connections, and we’ve worked with the wider community to understand their needs and determine how to address them. We have new knowledge that has come from our pilot work, figuring out things about improved practices, quality control, or requirements to support these transnational collaborations. Some of this has fed into other community efforts already, and as a result of this progress, I think that the collaboration would keep going outside of the focus of this project, but it would be much slower.
About the author
Name: Yvonne Firing
Work Package: WPs 2,3 and 5
Organisation: National Oceanography Centre
