North Sea carbon burial and transport off the northwest European shelf in global Earth system models (ESMs) – validation and variability

The North Sea is a very productive and heavily exploited continental shelf sea that acts as a sink for atmospheric CO₂. The balance between carbon buried in North Sea sediments and exported off the northwest European shelf into the North Atlantic is highly uncertain, rendering carbon budgets difficult to make and future changes of the system hard to predict. As part of the NoSE (North Sea-Atlantic Exchange) project, this study evaluates the uncertainty and variability in the exchange (with the North Atlantic) and burial of carbon in the North Sea as simulated by global Earth System Models (ESMs).  

From six state-of-the-art ESMs (ACCESS-ESM1-5, CanESM5, CMCC-ESM2, CESM2-WACCM, IPSL-CM6A-LR, NorESM2-LM), the years matching the observational period were selected from the CMIP6 historical experiment (1850 to 2014) and the ScenarioMIP of CMIP6 ssp245 experiment (2015 to 2100).

Here, we compare simulated values for sea surface temperature, salinity, phosphate, alkalinity and dissolved inorganic carbon (DIC) to an internally consistent data product for the northwest European shelf (NWESDAP; 347 datapoints). The models show good agreement for temperature (r² > 0.8), but weak simulation of salinity (r² < 0.4). The weak fit is mainly caused by a bad representation of low salinity, whereas in the higher salinity range, the comparison between the models and the observations is better. The model simulation of phosphate shows a weak fit with the observations (r² < 0.5), mainly caused by a bad model representation of higher phosphate concentrations in the observational dataset. Although the model representation of total alkalinity and surface DIC is weak (r² < 0.3), the mean observed DIC values are reasonably well represented by the models. As the horizontal flux of DIC is a product of horizontal water transport and DIC concentration, the representation of the variability in DIC might not be as important for calculating lateral fluxes as an adequate simulation of its spatial mean value.

The model ensemble representation of the main in- and outflow areas, as well as representation of the spatial variability of the variables is considered good enough to compare the present-day and the future North Sea carbon cycle.

Using the model ensemble, we will present the spatial and temporal variability in carbon burial, carbon concentrations, horizontal carbon transport and the balance between these components, both for present-day and future climate conditions.