THE EVALUATION OF THE ENSO-LIKE INTERANNUAL VARIATION IN THE GOALS/LASG MODEL

The features of the ENSO-like interannual variation of the sea surface and subsurface temperatures have been explored in the equatorial Pacific, their zonal propagation, and teleconnect ion based on 30-year integrat ion results of the new version of the global ocean-atmosphere-land system (GOALS) model at State Key Laboratory of Atmospheric Sciences and Geophysical Fluid Dynamics, Institute of At mospheric Physics (LASG/IAP).In this model, the R42L9 at mospheric general circulation model (AGCM) is full coupled with the T63L30 ocean general circulation model between 40 S and 40 N. The results show that the GOALS model captures some obse-rvational phenomena. For example, the irregular interannual variation of the sea surface temperature (SST) in the equatorial Pacific, the anomalous SST in the equatorial eastern Pacific propagates west ward with time, the opposit edirection propagation process of the anomalous sea subsurface emperature in the equatorial Pacific from the west to the east and from the deep upward to the shallow layer, and the variation of the SST in the east of the equatorial Pacific are negatively correlated with that in the west and the south west of the Pacific and positively correlated with that in the South Indian Ocean and the subtropics of the At lantic. However, there still existevident biases in the GOALS model with reference to observations. For instance, the mag nitude of the interannual variability of the SST in the equat orial eastern and central Pacific is too small. The model cannot simulate the feature of larger mag nitude of the SST variation in the eastern Pacific than that in the central Pacific. The propagation velocity of the SST anomalies from the east to the west of the equatorial Pacific is much slower than that calculat ed from the observation dat a but the propagation of the subsurface temperature anomalies from the west to the east of the equatorial Pacific is faster than the observation. The teleconnections such that the SST anomalies in equatorial eastern Pacific is negatively correlated with that in the northwestern Pacific and the positive correlation with that in the northern Indian Ocean are not simulated by the GOALS model. Those modelerrors, there fore, influence the precipitation simulation in the South and Sout heast Asia.