An Assessment of the Oceanic Physical and Biogeochemical Components of CMIP5 and CMIP6 Models for the Ross Sea Region

The physical and biogeochemical performance of 16 CMIP5 and 16 CMIP6 Earth System models (ESM) are examined relative to present day (1976–2005) observational data sets for a Ross Sea Region (RSR) containing the Ross Gyre (RG) and the Ross Sea Continental Shelf. A relative ranking scheme using error metrics and published ESM properties (including climate sensitivity parameters and anomalous deep ocean convection statistics) enables identification of a “best” ensemble of models for the RSR. Over the RSR the CMIP6 models are generally found to have improved physical representations compared to the CMIP5 set (based on our metrics for sea ice concentration, sea surface temperature, salinity, and height, and mixed layer depth), but the CMIP5 and CMIP6 biogeochemical representations remain similar. Examination of mean properties for the period 2081 to 2100 for RCP8.5 and SSP585 for CMIP5 and CMIP6, respectively, shows significant surface temperature increases, with significant decreases in sea surface salinity, sea ice concentration, and mixed layer depth across the RSR. Biogeochemically, there are generally small increases in surface values for chlorophyll, integrated primary production, and zooplankton carbon concentrations. The projections also have robust reductions in surface nitrate, phosphate, and silicate across the RSR. For that part of the RG circulation to the east of 180°E—which we refer to as the “inner RG”—significant barotropic transport increases are found by the end of century. Plain Language Summary Global climate simulation models have been tested to see how well they represent observations of the ocean in a region around Antarctica known as the Ross Sea. Here we are interested in both the physical (temperature, flows etc) and biogeochemical (chlorophyll, nutrients etc) properties of the models compared to the observations. This process allows us to identify the “best” models, or at least to rank them all from “best” to “worst.” This then provides us with more confidence about their future predictions, the argument being that models that do well for our present ocean will do better into the future. By the end of the century we find that the predictions show the surface ocean is warmer and fresher than today, with much reduced sea ice, and surface nutrients are depleted but the surface chlorophyll not so. The large scale circulation is also increased. This shows that the Ross Sea is likely to experience significant physical and biological change