台风路径多模式集成预报技术研究  

A study on multi-model ensemble forecast technique for Typhoon track

作　　者：郭蓉[1] 余晖[1] 漆梁波[2] 江漫[2] GUO Rong;YU Hui;QI Liangbo;JIANG Man(Shanghai Typhoon Institute of CMA,Shanghai 200030,China;Shanghai Meteorological Central Observation,Shanghai 200030,China)

机构地区：[1]中国气象局上海台风研究所,上海200030 [2]上海中心气象台,上海200030

出　　处：《气象科学》2019年第6期839-846,共8页Journal of the Meteorological Sciences

基　　金：国家自然科学基金资助项目(41775065;41405060;4185080);科技部重点专项政府间国际合作项目(2017YFE0107700);中国气象局预报员专项资助项目(CMAYBY2017-024)

摘　　要：利用NCEP、ECMWF、日本数值、英国数值、上海台风模式和广州模式包含全球模式和区域在内的6家数值模式资料,利用近似SEAV方法,设计台风路径多模式集成预报方法(SHME),并用2014—2016年的台风客观预报数据进行多模式集成预报的效果检验,且与ECMWF模式进行比较,通过比较发现,SHME方法较ECMWF在12~48 h预报上有明显改进,在72~120 h预报2014年尤其突出,2015—2016年均与ECMWF预报效果相当。Using global and regional models from six different forecast centers including NCEP,ECWMF,Britain model,Japan model,Shanghai typhoon model and Guangzhou model,and the approximate SEAV method,this paper designed a multi-model ensemble forecast technique for typhoon track(SHME).The typhoon objective forecast data of 2014—2016 was tested for the effect of SHME method and compared with the data from the ECMWF model.The result showed that the SHME method had a significant improvement over the ECMWF during 12—48 h,and SHME was much better than ECMWF during 72—120 h in 2014.In addition,SHME method was equivalent to the ECMWF in 2015—2016.

关 键 词：TC 多模式集成 路径预报 平均误差 

分 类 号：P457.8[天文地球—大气科学及气象学]