Interannual climate variability in the Tropical Indian Ocean: A study with a hierarchy of coupled general circulation models

The Indian Ocean shows significant climate variability at interannual timescales. This variability is affected by the El Niño--Southern Oscillation (ENSO) phenomenon. A hier-archy of coupled model simulations is used as well as observational datasets and reanaly-sis products to investigate the role ENSO plays in determining the interannual climate variability in the Indian Ocean and to investigate the spatial and temporal characteristics of the ENSO--independent climate variability. The hierarchy of coupled model simula-tions comprises three experiments. Two experiments are conducted with the ECHO-G model, which is a state--of--the--art ocean--atmosphere general circulation model (CGCM). The first experiment with ECHO-G is a standard integration, which serves as a control integration. The second experiment with ECHO-G is similar to the control ex-periment but the interannual ocean--atmosphere interactions have been inhibited in the tropical Pacific Ocean, so that in this experiment the ENSO signal is excluded by defini-tion. The third experiment is conducted with a fixed--depth mixed--layer model coupled to the same atmosphere model that has been used in the two experiments with ECHO-G. This integration allows the investigation of the role of ocean dynamics in the generation of interannual sea surface temperature (SST) variability. It is shown that the variability of several atmospheric and oceanic fields in the Indian Ocean region, including SST, is consistent with a cycle only if ENSO is included in the simulation. The ENSO signature in the Indian Ocean includes both a dynamical and a thermodynamical response. ENSO influences the Indian Ocean at two different interannual timescales and provides conditions favorable for coupled ocean-atmosphere interac-tions in the Indian Ocean. In contrast, the climate variability that is independent of ENSO shows a response of the ocean to the atmospheric forcing that is not significantly more energetic than expected from a stochastic forcing approach, thereby reflecting either a passive oceanic response or weak coupling in the ENSO independent Indian Ocean variability