Climate model simulation of the South Indian Ocean Convergence Zone: mean state and variability

Evaluation of climate model performance at regional scales is essential in determining confidence in simulations of present and future climate. Here we developed a process-based approach focussing on the South Indian Ocean Convergence Zone (SIOCZ), a large-scale, austral summer rainfall feature extending across southern Africa into the southwest Indian Ocean. Simulation of the SIOCZ was evaluated for the Coupled Model Intercomparison Project (CMIP5). Comparison was made between CMIP5 and Atmospheric Model Intercomparison Project (AMIP) models to diagnose sources of biases associated with coupled ocean-atmosphere processes. Models were assessed in terms of mean SIOCZ characteristics and processes of interannual variability. Most models simulated a SIOCZ feature, but were typically too zonally oriented. A systematic bias of excessive precipitation was found over southern Africa and the Indian Ocean, but not particularly along the SIOCZ. Excessive precipitation over the continent may be associated with excessively high low-level moisture flux around the Angola Low found in most models, which is almost entirely due to circulation biases in models. AMIP models represented precipitation more realistically over the Indian Ocean, implying a potential coupling error. Interannual variability in the SIOCZ was evaluated through empirical orthogonal function analysis, where results showed a clear dipole pattern, indicative of a northeast-southwest movement of the SIOCZ. The drivers of this shift were significantly related to the El Niño Southern Oscillation and the subtropical Indian Ocean dipole in observations. However, the models did not capture these teleconnections well, limiting our confidence in model representation of variability.