A PRELIMINARY VALIDATION STUDY OF THE SEASONAL FORECAST OF CCCMA MODEL OVER CHINA

In this study we validate the raw ensemble mean forecasts of the CCCma’s GCM2 modelagainst surface temperature and precipitation data obtained from 160 Chinese stations.It is foundthat despite the lagre biases,the model was able to produce seasonal anomalies that haveproperties that are reasonably close to those that are observed.This anomaly is the quantity ofinterest when forecasting seasonal climatic conditions.The root mean squared difference(RMSD)between the forecast and observed anomaly leads us to be modestly optimistic about the prospectsfor using dynamical models to forecast the interannual variability of some meteorological elements.The correlation analysis of the forecast and observation also supports the result given by theRMSD analysis and provides a tool for identify the forecast confidence level in various regions,