Longitudinal Biases in the Seychelles Dome Simulated by 34 Ocean-Atmosphere Coupled General Circulation Models

The upwelling dome of the southern tropical Indian Ocean is examined by using simulated results from 34 ocean-atmosphere coupled general circulation models (CGCMs) including those from the phase five of the Coupled Model Intercomparison Project (CMIP5). Among the current set of the 34 CGCMs, 12 models erroneously produce the upwelling dome in the eastern half of the basin while the observed Seychelles Dome is located in the southwestern tropical Indian Ocean (Figure 1). The annual mean Ekman pumping velocity is almost zero in the southern off-equatorial region in these models. This is in contrast with the observations that show Ekman upwelling as the cause of the Seychelles Dome. In the models that produce the dome in the eastern basin, the easterly biases are prominent along the equator in boreal summer and fall that cause shallow thermocline biases along the Java and Sumatra coasts via Kelvin wave dynamics and result in a spurious upwelling dome there. In addition, these models tend to overestimate (underestimate) the magnitude of annual (semiannual) cycle of thermocline depth variability in the dome region, which is another consequence of the easterly wind biases in boreal summer-fall. Compared to the CMIP3 models (Yokoi et al. 2009), the CMIP5 models are even worse in simulating the dome longitudes and magnitudes of annual and semiannual cycles of thermocline depth variability in the dome region. Considering the increasing need to understand regional impacts of climate modes, these results may give serious caveats to interpretation of model results and help in further model developments.; Figure 1: The longitudes of the shallowest annual-mean D20 in 5°S-12°S. The open and filled circles are for the observations and the CGCMs, respectively.