Currently, one of the pressing questions in climate science is how much of the carbon dioxide (CO2), that humans produce through the burning of fossil fuels, changes in land use and other activities, ends up in the ocean. The ocean’s role in taking up atmospheric CO2 is vital, and without it, the planet would be warming even faster than at present. In fact, scientists estimate that around a quarter of all CO2 released as a result of human activities has been taken up by the ocean. Of that, around 40% goes into the Southern Ocean, the region south of latitude 35°S, making it the largest single contributor to the overall uptake of CO2 by the ocean - what’s known as the ocean carbon ‘sink’.
Unfortunately, due to its remoteness and harsh weather conditions, observations of the Southern Ocean by oceanographers have historically been sparse, both in space and in time. This makes it difficult to be confident in our estimates of the Southern Ocean carbon sink, and crucially it complicates the task of comparing observations with models. Where the models and the observations appear to disagree, as is the case for the carbon sink, do we conclude that the models are getting it wrong, and therefore need improving, or does the lack of available observations mean that our attempts to estimate the sink in the real ocean cannot be trusted? Ultimately, accurate models are our only way of confidently predicting what the future holds for the changing climate, and thereby allowing us to plan for it.
To estimate the amount of CO2 that moves between the atmosphere and the ocean, known as the flux, requires three ingredients. The first two are the concentrations of CO2 in the atmosphere and the surface ocean; it is the difference between these two measurements that drives a flux in one direction or the other. Since gases naturally diffuse from regions where they are more concentrated to regions where they are less so, if the CO2 concentration is higher in the atmosphere than the ocean, this drives a flux of CO2 into the ocean. Conversely, if the concentration is higher in the ocean than the atmosphere, this drives a flux out of the ocean. The third ingredient is a model of just how diffusion across the air-sea boundary works, which depends on various factors, most importantly the strength of the wind. The atmospheric CO2 concentrations are well-known because the atmosphere mixes quickly, so we can rely on continuous observations in only a few places to work it out. Winds can be measured using satellites, giving almost total coverage over the global ocean. This leaves the ocean surface CO2 concentrations, which are more variable than the atmosphere, and which satellites cannot observe; they are the focus of our study.
Existing observations of ocean surface CO2 concentrations have been collected mainly from ships over a period spanning the last 4 decades. In the northern hemisphere, many of these observations were acquired by so-called ‘ships of opportunity’, where the instrument needed to take the measurements is installed on cargo vessels that are constantly on the move, and contribute substantially to data coverage. In the Southern Ocean, there are no such ships, and historically we have relied on scientific research expeditions, that are generally confined to the summer months, to collect the CO2 data. The result is that our estimates of the Southern Ocean carbon sink and its variability over the past 40 years or so have been based on observations that are sparse and strongly biased towards the summer months.
So how do we go about filling in some of the gaps? We want to know what the surface concentrations of CO2 were in the Southern Ocean in winters where they were not observed. Clearly, we can’t travel back in time and take more measurements. But there are observations taken in summer that we can make use of: those taken below the surface. We take advantage of a special property of the waters in the top few hundred metres in the more southerly parts of the Southern Ocean: in summer there exists a layer of water that is colder than the waters above and below it. We call this layer a ‘temperature minimum’, and it is useful because it contains water that would have been at the surface at some point during winter the previous year, and whose properties have been preserved since that time. In effect, it remembers what the conditions were like at the surface when we were not around to observe them. We take observations of carbon concentrations from this temperature minimum layer in summer, and use them to extrapolate new values of wintertime surface CO2 concentration. The number of additional surface concentrations is quite small, just a few hundred over a 15-year period, but they are spread all around the Southern Ocean and so provide wintertime data coverage for past years over a large area where we previously had none.
What difference does the new data coverage make to our estimate of the carbon sink? Well, it turns out that not very much is changed compared with previous estimates. This might seem like a rather mundane result, and it is indeed not terribly exciting. However, it is useful because it helps answer the question of whether our models or our observational estimates are at fault where the two disagree: it points in the direction that the problem is with the models. Of course, as with much scientific enquiry, there is more to learn before we can be definitive in pointing the finger of blame, but we are edging closer. Figuring out why the models are wrong, if they are wrong, is another task altogether, and one for another day!
Please sign in or register for FREE
If you are a registered user on Research Communities by Springer Nature, please sign in