2016年梅汛期降雨环流特征及ECMWF中期预报偏差分析

利用NCAR/NCEP逐日再分析资料和台站观测日平均降雨资料，分析2016年梅汛期的大气环流演变特点和期间3次强降雨过程的环流特征，对比了欧洲中期数值模式(EC模式)的预报能力，并对其中期预报降雨的落区偏北、强度偏弱的偏差原因进行分析。结果表明，2016年梅汛期中高纬度环流多变化，多冷空气活动但势力总体不强，夏季风在6月下旬和7月上旬逐步增强，西太平洋副热带高压稳定维持，为强降雨的发生提供了有利动力和水汽条件。在梅汛期前期EC在中期时效对于夏季风的预报强度偏强、副高位置偏北，直接造成模式预报的雨带位置偏北。EC对于乌拉尔山一带的环流系统预报能力较好，但对于日本海-鄂霍茨克海一带的环流系统预报能力较差，从而使得影响我国的冷空气路径和强度预报均出现偏差，这对于7月初的强降水的强度和落区预报也有明显影响。 

ANALYSIS OF MID-RANGE FORECAST BIAS BY ECMWF ON PRECIPITATION FEATURES AND RELATED LARGE-SCALE CIRCULATION DURING 2016 MEIYU SEASON

In 2016, the precipitation was significantly above normal over the middle and lower reaches of the Yangtze River basin during the Meiyu season. It is rare that, in recent years, ECMWF showed obvious forecasting bias on both the location and the intensity of severe rainfall processes, particularly on July 1 to 4. In ECMWF forecast, the location of the rainfall belt is obviously northward and its intensity is much weaker than observations. In this paper, using daily NCEP/NCAR reanalysis data and observational daily precipitation data, characteristics of rainfall and related atmospheric circulation during the 2016 Meiyu season are analyzed, and then forecasting bias of ECMWF is carefully investigated. The results show that, during the Meiyu season, a strong and stable western Pacific subtropical high (WPSH), a gradually strengthened East Asian summer monsoon (EASM) from late June to early July, and together with frequent but weak cold air activities, provide favorable dynamic and moisture conditions for strong rainfall. From the point of the mid-range forecast, the ECMWF forecasting bias is mainly attributed to the overestimation of the intensity of EASM and more northward location of WPSH, which leads to a northward forecasting bias of the location of the strong rainfall belt. Additionally, although the model gives a good forecast for mid-latitude circulation near Ural Mountains, the forecast in the Sea of Japan and near Okhotsk is relatively bad, which result in an obvious bias on the track and the intensity of the cold air activity, especially in early July. It is perhaps helpful for forecasters to understand these forecasting bias and carry out a reasonable correction based on the numerical forecast.