The new Meteorological Research Institute coupled GCM (MRI-CGCM2) - Model climate and variability

Abstract

A new version of a global coupled atmosphere-ocean general circulation model (MRI-CGCM2) has been developed at the Meteorological Research Institute (MRI). The model can be used to explore climate change associated with anthropogenic forcings. We aimed to reduce the drawbacks of the former version of the model (MRI-CGCM1, Tokioka et al., 1996) and achieve a more realistic climatic mean and variability to predict climate changes with greater accuracy. In a preliminary analysis of the control run, the model showed generally good performance in reproducing the mean climate (including seasonal variation) in representative aspects; surface air temperature, precipitation, snow and sea ice distribution, and ocean structure and circulation. The model is capable of making a stable integration longer than 200 years. The sea ice distribution is much improved and is close to the observed extent and thickness. The model simulates realistic strength of meridional overturning in the Atlantic Ocean that MRI-CGCM1 failed to simulate. The model realistically simulates variabilities such as Arctic Oscillation (AO) and ENSO. Temporal variation of the sea surface temperature (SST) anomaly in the NINO3 region (150°W to 90°W, 4°S to 4°N) shows a large positive value (max. +4°C) with several years interval. The SST anomaly pattern is similar to the observed El Nino with a strong positive anomaly in the central-eastern equatorial Pacific. The model still has some biases at present. The surface air temperature in winter at high latitude has a warm bias due to weaker stability in the boundary layer. The surface temperature over land in summer also shows a warm bias associated with a problem concerning the hydrological process.