Analysis of the Zonal Mean Atmospheric Climate State in IAP/ LASG GOALS Model Simulations

The results of four versions of IAP/LASG Global Ocean-Atmosphere-Land System Model (GOALS) are analyzed separately over the oceans and over continents, and compared with observed data.Some fundamental atmospheric variables including surface air temperature (SAT), sea level pressure (SLP) and precipitation are examined to evaluate the ability of the GOALS model to simulate the contemporary climate and climate variability. In general, all four versions of the GOALS model are capable of reproducing the main features of the mean state and seasonal variation of the observed climate with reasonable accuracy.The evaluation also reveals some weakness of the model. According to this study, we can clearly see that the essential discrepancy of global averaged SLP lies over the continents in boreal summer. The simulated higher SAT over land versus the observed is mainly due to the effect of the land surface process. It is worth noting the underestimation by simulated precipitation rates mostly appears over the oceans, yet over-land precipitation is higher in high and middle latitudes than the observed for the boreal winter.Through intercomparisons among different versions of the model, it can be clearly seen that the incorporation of the diurnal cycle of solar radiation apparently improves the simulation of SAT, especially in the low and middle latitudes over land. Also, the introduction of the diurnal cycle shows a great improvement in precipitation in tropical continents and wintertime precipitation in high and middle latitudes. Furthermore,based on the daily flux anomaly exchange scheme (DFA), the latest version of GOALS model simulated over-ocean temperature variability is improved in the low and middle latitudes.Having compared the standard deviation of the annual mean surface air temperature (SAT) simulated by the GOALS model to observation, it is found that all four versions of the GOALS model underestimate surface air temperature variability over both oceans and land relative to observations. Several factors that may contribute to these differences between simulated and observed temperature variability are identified.