华南复杂地形下GRAPES_Meso3km对流尺度模式前汛期精细化降水预报评估

为深入认识GRAPES_Meso(Global/Regional Assimilation and Prediction System)3 km对流尺度区域模式对华南前汛期精细化降水的预报性能,为模式改进及业务应用提供参考依据,利用广东省86个站点逐小时观测降水资料和国家气象信息中心多源融合降水资料,针对广东省复杂地形特点,结合距海岸线的远近及站点地形特点,将86个站划分为沿海东部、沿海西部和内陆地区三个子区域,采用二分类降水预报检验方法,定量评估了2020年5月18日—6月18日华南前汛期降水预报效果。结果显示,GRAPES_Meso 3 km模式精细化降水预报技巧受广东复杂地形影响较大,广东沿海东部和内陆地区24 h时累积降水的小雨、中雨、大雨量级预报成功指数(Threat Score,TS)、公平成功指数(Equitable Threat Score,ETS)评分高于沿海西部地区,尽管暴雨预报评分具有此相同特征,但三个子区域的暴雨预报评分总体较低;从3 h累积降水预报评分看,沿海东部、沿海西部及内陆地区等三个子区域存在明显的日变化特征,但是沿海东部及西部与内陆地区表现有所不同,沿海东部和西部降水预报评分夜间较低(预报偏差偏高),白天相对较高(预报偏差偏低),而内陆地区则是夜间较高(预报偏差偏低),白天相对较低(预报偏差偏高)。沿海西部预报评分相对较低的原因是由于检验时段内广东地区存在一个弱的风切变,而沿海西部大部分地区正好处于切变线南侧的温度高值区控制,但模式模拟该区域的日平均温度较实况偏低,导致沿海西部模式预报降水空报较多,降低其降水预报技巧。

Evaluation of fine-scale precipitation forecast of GRAPES_Meso 3 km convective-scale model in early summer rainy season in South China under complex topographical conditions

To learn more about the performance of GRAPES_Meso 3 km (Global/Regional Assimilation and Prediction System) model for precipitation prediction in pre-rainy season in South China under complex topographical conditions,and to provide a reference for model development and business application,we divide the 86 stations in Guangdong Province into three sub-regions:the eastern coastal area,the western coastal region and the inland region according to the complex terrain characteristics of Guangdong Province and in combination with the distance from the coastline and the terrain characteristics of the station by using the hourly precipitation data of 86 stations in Guangdong Province and the multi-source fusion precipitation data of the National Meteorological Information Center.The binary event forecast verification methods are used for quantitative evaluation of precipitation forecast effect during the pre-rainy season from May 18 to June 18,2020 in South China.The results show that the refined precipitation prediction skills of GRAPES_Meso 3 km model are greatly affected by the complex terrain of Guangdong.The scores of prediction success index (TS) and equitable thread score (ETS) of 24-hour cumulative precipitation of light rain,moderate rain and heavy rain magnitude in the eastern coastal and inland areas of Guangdong are higher than those in the western coastal areas;although the rainstorm prediction scores have the same characteristics,the rainstorm forecast scores of the three-sub regions are generally low;from the perspective of 3 h cumulative precipitation forecast score,there are obvious diurnal variation characteristics in the three sub regions of the eastern coastal region,the western coastal region and the inland region,but the performance of the eastern coastal region and the western coastal region is different from that of the inland region.The precipitation forecast score of the eastern coastal region and the western coastal region is low at night (the forecast deviation is high),relatively high during the day (the forecast deviation is low),while that of the inland region is high at night (low forecast deviation),relatively low in the daytime (high forecast deviation).The reason for the relatively low prediction score in the western coastal region is that there is a weak wind shear in Guangdong during the test period,and most of the western coastal region is just controlled by the high temperature area in the south of the shear line,but the daily average temperature in the region simulated by the model is lower than the actual situation,resulting in more empty precipitation forecasts in the western coastal region,which reduces its precipitation prediction skills.