Seasonal Biases in the Tropical Atlantic Sector in Climate Models: Causes and Impact on Interannual Variability

The PIs propose to complete a diagnostic examination of the relationship between bias in the representation of the seasonal cycle and CGCM simulation of climate variability, and secondly a climate modeling study using the bias-corrected seasonal cycle. They focus on the Atlantic sector partly because the bias is more severe there than in the Pacific. This proposal will extend their previous study to a multi-model analysis in order to look at the impact of this seasonal bias on errors in representation of climate variability. The PIs will also apply these results to improve representation of climate variability in CGCMs, focusing on the NOAA/GFDL CM2.1 model in cooperation with members of the GFDL climate group. 

The current plan is to continue a diagnostic examination of bias in climate variability in the tropical Atlantic sector of NCEP, NCAR, and GFDL CGCMs. The analysis includes the relative roles of local dynamic and thermodynamic air-sea interactions and the remote influences of ENSO and extratropics on the tropical Atlantic sector. For CM2.1, they already have access to an interesting suite of experiments including experiments in which SST in the tropical Pacific and Indian sectors are replaced with climatological monthly SST and a flux-corrected experiment with an “improved” climatological seasonal cycle. They will design sensitivity experiments to examine the response of simplified and full atmospheric models to a bias correcting forcing. The ultimate goal of these sensitivity studies is to formulate suggestions for the improvement of coupled models.