与数值模式预报结合的冬季延伸期强降温过程预测方法及业务化应用  

A Predicting Method of the Strong Cooling Process during Winter with Numerical Model Prediction and its Operational Application

作　　者：陈伯民[1] 周坤[1,2] 信飞 马俊[3] 靳利梅 CHEN Bomin;ZHOU Kun;XIN Fei;MA jun;JIN Limei(Shanghai Climate Center,Shanghai 200030,China;Qingpu Weather Office,Shanghai,Shanghai 200030,China;Shanghai Sucslife Science and technology Co.Ltd.,Shanghai 200090,China)

机构地区：[1]上海市气候中心,上海200030 [2]上海市青浦区气象局,上海201700 [3]上海成生科技有限公司,上海200090

出　　处：《高原气象》2023年第2期495-505,共11页Plateau Meteorology

基　　金：国家重点研发计划课题(2018YFC1505601);国家自然科学基金重点项目(41730959);中国气象局气象预报关键技术发展专项[YBGJXM(2018)04-04]。

摘　　要：首先对NCEP-CFSv2模式预报的2017年1-3月和10-12月1~30天700 hPa 8个关键区低频波效果做了定量评估,然后,在此基础上将模式1~30天低频波预报结果和低频波预测模型相结合,在2018年1-4月和2018年11月至2019年1月对上海地区做了15次延伸期强降温过程业务预测,结果显示:(1)模式预报的关键区低频波位相和演变趋势与实况高度吻合,预报的延伸期(11~30天)关键区低频波与实况相关系数达0.839;预报的延伸期(3~6候)关键区低频波趋势平均准确率达83.3%,准确率为100%的比例高达45.8%。(2)15次延伸期强降温过程业务预测的平均准确率、Cs评分和Zs评分分别为61.2%,0.149和0.158,并准确给出2018年年初和年末两次最强降温过程发生时段,预报时效分别为18天和16天,显示出一定的预测技巧;评分结果显著高于未应用CFSVv2模式结果的2015年和2016-2017年1月同期业务预测。At first this paper quantitatively evaluated the low-frequency wave performance of NCEP-CF Sv2 model over the eight key areas on 700 hPa from January to March and from October to December of 2017,and then made operationally the fifteen extended-range operational predictions of strong cooling process for January to April of 2018 and for November to January 2019 with the 1~30 days prediction given by CFSv2 model as well as the low-frequency wave conceptual predicting model.The results show that the phase and evolution trend of the low-frequency wave in the key area predicted by CFSv2 model are highly consistent with the reality,with the correlation coefficients of 0.839 of the predicted low-frequency waves with the observed for the extended-range(11~30 days),the accuracy of low frequency wave trend by the model over 3~6 pentad up to percent of 83.3 on average,and the percentage of 100-percent accuracy of the trend even up to 45.8.The average accuracy,Cs and Zs scores of 15 strong cooling process operational predicting are 61.2%,0.149 and 0.158 respectively,and at the same time the occurrence of the two strongest cooling processes at the beginning and the end of 2018 were accurately given with the lead-time of 18 and 16 days in turn,which are significantly higher than those of the operation predicting for the same period of 2015 to 2017 without CFSv2 results.

关 键 词：延伸期强降温过程预测 低频波 数值模式 

分 类 号：P456[天文地球—大气科学及气象学]