Evaluation of the interior ocean ventilation of biogeochemical tracers in a global ocean model using observation-based metrics

The ocean has absorbed approximately 25% of anthropogenic CO2 emissions since the industrial era, playing a critical role in the global carbon cycle. However, the current ocean carbon sink as simulated by the ocean biogeochemistry models of the Global Carbon Budget shows a spread larger than the European Union’s fossil carbon emissions and mismatches with current observation-based estimates. The prime suspect for this deviation is the poorly constrained transfer of carbon between the surface and the interior ocean. This process is called ventilation and is based on the interior ocean carbon gradients that depend on mixing, advective and biological processes.

To address this, we developed a set of metrics based on the new dataset from biogeochemical Argo floats (BGC-Argo) that offer unprecedented observations from the surface to 2000 m, and the GLODAP bottle data. These metrics are a tool to evaluate and optimize ocean ventilation processes and carbon transport between the surface and the interior in ocean models. They target the stratification and mixing (physical variables) as well as the gradients of tracers such as apparent oxygen utilization, dissolved inorganic carbon or dissolved inorganic nitrate. We compute metrics quantifying these depth gradients averaged across large-scale biomes.

With this methodology, we evaluate the ventilation in the model FESOM-REcoM. Our results identify model-observation differences in terms of absolute values and magnitude of gradient in salinity and in the biogeochemical variables in many biomes. Biases in the gradients of biogeochemical properties can partially be explained by biases in the physical stratification of the water column, especially in biomes with high mixing at higher latitudes.  In other biomes, biases are attributed to an imperfect representation of biogeochemical processes in the model.  We characterize the distribution of biases in FESOM-REcoM, and discuss how to reduce them.