The carbon budget, an estimate of how much CO2 humanity may still produce before the atmosphere heats beyond a certain threshold, is the figure at the heart of all climate targets, nationally decided contributions, and net-zero pledges. Models are the foundation of that figure. These models make assumptions about how the ocean functions as a carbon sink, including how much of the carbon emitted by humans is absorbed by seawater, how this absorption rate varies as the ocean warms, and how various biological and chemical processes in the water column influence how long the carbon remains sequestered once it enters the ocean. According to a recent analysis authored by 72 scientists from 23 different nations and organized by IOC-UNESCO, the models differ from one another on this issue by 10 to 20 percent worldwide, and by even greater percentages in some areas. That disparity in the basis of global climate accounting is not insignificant.
The most thorough attempt to date to identify the uncertainties impacting ocean carbon science is the Integrated Ocean Carbon Research Report, but the image it paints is unsettling. The gaps are fundamental to the processes that decide whether the ocean continues to absorb carbon at its current pace or begins to absorb less as conditions change. They are not esoteric technical issues. how circulation patterns that raise carbon-rich deep water to the top are altered by ocean warming.
How changes in microbial and phytoplankton communities affect the amount of carbon that is finally stored in the deep ocean after being absorbed at the surface. How carbon is exchanged with the atmosphere on seasonal and decadal timescales in coastal and polar regions, two of the ocean monitoring network’s least data-rich environments. These details cannot be reliably parameterized using current data. These are real sources of uncertainty, and the degree of uncertainty is significant enough to have an impact on the emission reduction estimates now being made by governments.
Discussions on climate policy frequently ignore the practical implications of this uncertainty, but it merits further consideration. More CO2 remains in the atmosphere if global climate models, which currently assume that the ocean will continue to absorb between 25 and 30 percent of human carbon emissions annually, show a decline in the actual absorption rate, whether due to warming reducing the ocean’s capacity, circulation changes altering transport patterns, or biological changes reducing the efficiency of the biological pump. Higher atmospheric concentrations than predicted by the models are produced by the same emission route. The carbon removal requirements increase and steeper emission reductions are needed to meet the temperature targets that were established based on certain carbon budgets.
One notable and significant observation made in the paper is that coastal and polar regions are particularly data-poor. In addition to being some of the ocean’s most important carbon exchange zones, the high latitudes—the Arctic and Southern Oceans—are also the areas experiencing the greatest rates of biological system change and climate change. It is technically challenging and costly to monitor these areas at the density required to lower model uncertainty since the surroundings are harsh, remote, and equipment-intensive. Only a small portion of what would be required for accurate regional carbon flux estimations is covered by the current in-situ measuring networks. Carbon exchange at the ocean’s surface is more difficult to measure remotely than other oceanographic variables like temperature and sea surface height, however satellite observations are helpful.
A global ocean carbon monitoring system that incorporates satellites, self-governing float and glider platforms, and continuous readings from the surface to the deep ocean is the report’s suggested solution. Global subsurface ocean monitoring is currently supported via the Argo float network, although conventional Argo floats lack the biogeochemical sensors required to detect carbon flux. These skills are being expanded, albeit slowly and at a high expense, by the Biogeochemical-Argo program. For the same reason that IPCC modeling centers treat data quality as a prerequisite for projection reliability, the UNESCO report presents this expansion as an essential investment: you cannot reduce model uncertainty without reducing data uncertainty, and you cannot reduce data uncertainty without more and better measurements.
Although it receives less attention than the technological topic, the report’s capacity development angle—the focus on developing monitoring infrastructure in underrepresented regions—may be just as significant. The areas with the poorest characterization of ocean carbon exchange are frequently those with smaller scientific groups addressing these issues, fewer research vessels, and less infrastructure for oceanographic research. For the modeling issues the paper outlines, a global monitoring system that only provides adequate coverage of the North Atlantic and Pacific quadrant isn’t really a global monitoring system in any meaningful sense.
Observing how this problem has evolved over time, it’s hard not to believe that the scope of the modeling uncertainty revealed here merits greater attention than it usually gets in discussions on climate policy. The disparity between the high and low estimates of ocean carbon absorption produced by current models is not a rounding error; rather, it is a range that, when stretched over decades, indicates a substantial difference in the amount of space the atmosphere has for additional emissions before reaching specific warming thresholds. It is feasible to build climate commitments on models with that degree of uncertainty in one of their primary inputs if the uncertainty is recognized and the monitoring expenditure required to lower it is given top priority. Among other things, the UNESCO study carefully argues that it ought to be.
