Improved air–sea flux algorithms in an ocean–atmosphere coupled model for simulation of global ocean SST and its tropical Pacific variability

A revised algorithm for air–sea exchange parameterization of momentum, sensible and latent heat flux improves the climate simulation of the global distribution of sea surface temperature (SST) and tropical Pacific variability of SST. Based upon an analysis of studies from field programs, we apply the revised algorithm with new expressions for surface momentum and scalar roughness length dependent on 10-m winds in neutral condition, and evaluate them in the ocean–atmosphere coupled model of the Australian Community Climate and Earth-System Simulator. The revised algorithm improves simulations for mean global SST distribution, demonstrated with Pearson’s correlation indices showing corrections to a net fraction of 28 % over the global oceans. Being focused on the tropical Pacific, the algorithm eases the tropical SST cold tongue bias, and improves predictability of ENSO variability with better representations of the standard deviation of the Nino-3.4 index, especially the skewness of the index for nonlinearity of ENSO variability. Bjerknes and thermodynamical feedbacks are applied to understand the effects of the revised algorithm on the predictability of the Nino indices.