Assessment Analysis on Winter and Spring Temperature and Rainfall Forecasts over China with Regional Climate Model

Regional climate model of National Climate Center (RegCM_NCC) is developed from the second generation regional climate model (RegCM2) of NCAR.Horizontal resolution of RegCM_NCC is 60 km, andthe center position is 35°N, 110°E, with 151 by 79 g rids, covering most areas of the East Asian.The vertical dimension is divided into 16 layers, and the top-level atmospheric pressure is 100 hPa.15 laps are adopted for buffer zone.The initial and lateral boundary conditions come from the corresponding nested National Climate Center global coupled model (CGCM_NCC).In order to study the prediction ability of RegCM_NCC, operational assessment (five parameters) ofNational Climate Center are used to assess the hindcast of average temperature and precipitation of RegCM_NCC from 1983 to 2002 in winter and 1984 to 2003 in spring, as well as the real-time forecasting from2003 to 2007.The hindcast of ave rage temperature in winter by RegCM_NCC for 20 years is similar to the observation.The forecast of the average temperature is low er than the observation in general.The hindcast ofprecipitation distribution in winter is obviously different from the observation.Forecasts of precipitationare in southwest China are especially inaccurate.The correlation coefficient distributions of the averagetemperature and precipitation between the forecast and the observation in most areas are positive.The P score of w inter average temperature in most years are greater than 60, and the average is 66.4.The ACC of winter average temperature is -0.7—0.4 and the average is -0.07.The P score of the winter precipitation in most years are 60—80 and the average is 69.9.The ACC of precipitation are -0.2—0.5, with the average of 0.05.The hindcast of average temperature in spring by RegCM_NCC for 20 years is similar to the observation.The forecast of average temperature in eastern China is higher, and that in western China is low erthan observation.There are great differences between hindcast of spring precipitation distribution and theobservation.In southwest China the precipitation forecasts accuracy is rather poor.The correlation coefficient distributions of the average temperature between the forecast and the observation in most area arepositive.Negative correlation distribution of precipitation is to the north of the Yangtze River.The P score of the spring average temperature in most years are greater than 60, and the average is67.8.The ACC of spring average temperature is -0.5—0.5 and the average is 0.05.The P score of thespring precipitation in most years are 60—75 and the average is 65.6.The ACC of precipitation are-0.4—0.3, with the average of -0.01.The assessment of the forecasts with regional climate model of National Climate Center approve s themodel' s ability in forecasting average temperature and precipitation in winter and spring.The forecastinglevel needs further improving for the low mean score and ACC.