CMIP6 model performance in the Subpolar North Atlantic in relation to the AMOC

Abstract

In a warming climate, it is necessary to understand how the ocean, a crucial component of the climate system, will react. The Atlantic Meridional Overturning Circulation (AMOC) transports heat, nutrients, and carbon, but is often not well represented in climate mod- els. Deep convection in the Subpolar North Atlantic (SPNA) influences the strength of this circulation and is an important feature to represent accurately in climate models for reli- able future projections. This thesis explores how the representation of circulation and hydrography impacts stratification and regions of deep convection in the SPNA for a range of CMIP6 models in comparison to CMEMS reanalysis. All of the models show temperature and salinity biases throughout the Labrador, Irminger, and Nordic Seas, influencing the stratification and location or depths of deep convection. These biases can be caused by an overestimation of sea ice concentration, especially in the Labrador Sea, or too high temperatures in the North Atlantic Current with a stronger AMOC compared to observations. Increasing ocean resolution is not found to improve these biases, and results in an overestimation of the mixed layer depth in the Labrador Sea. If these processes are not well represented in comparison to observations or reanalysis products, it may influence their validity for future AMOC and climate projections.