基于集合预报优选台风路径订正技术的广西风雨预报研究  

Research on wind and rain forecasts in Guangxi based on the ensemble forecast optimization technique for typhoon track revision

作　　者：梁依玲 梁存桂 韩慎友 李佳颖 翟舒楠 LIANG Yiling;LIANG Cungui;HAN Shen you;LI Jiaying;ZHAI Shunan(Guangxi Meteorological Observatory,Nanning 530022,China)

机构地区：[1]广西壮族自治区气象台,南宁530022

出　　处：《气象研究与应用》2024年第3期21-28,共8页Journal of Meteorological Research and Application

基　　金：珠江流域(华南区域)气象科研开放基金项目(ZJLY202304);广西气象科研计划项目(桂气科2020QN03、桂气科2022M09)。

摘　　要：利用2017—2019年欧洲中心全球集合预报数据以及中国气象局热带气旋资料中心的最佳路径数据集,基于统计分析和集合预报释用技术,对影响广西的台风路径预报提出一种实时订正技术。通过优选N个误差较小的集合预报成员,生成台风路径订正产品、台风概率匹配平均降水产品和台风大风合成分析产品。结果表明,利用优选集合成员的算术平均模型预报台风路径比集合预报平均、确定性预报效果改善,且预报时效越长,改善效果越明显,24 h、48 h时效的台风路径预报误差比集合平均减小4 km和6 km,比确定性预报减小5 km和19 km;优选集合成员的概率匹配平均降水产品,在一定程度上能改善传统的集合平均降水产品,特别是对大暴雨落区预报,36 h、60 h时效大暴雨TS评分较集合平均提高10%和12%;优选集合成员的大风合成分析产品,能预报出台风影响时极大风区移动、增强、减弱的态势,36 h时效6、7、8级以上极大风的TS评分较集合平均提高18.3%、15.7%、13.4%,对9~11级以上的极大风有一定的预报能力。Based on the ECMWF global ensemble forecast data and the best track dataset of the Tropical Cyclone Data Center of the CMA from 2017—2019,this paper proposed a real-time revision technology for the forecasting of the tracks of typhoons affecting Guangxi by using statistical analysis and the ensemble forecast release technique.By selecting the optimal N ensemble forecast members with smaller errors,the revised typhoon track products,the typhoon probability-matching average precipitation products,and the the synthetic analysis products for typhoon gales are generated.The results showed that the use of the arithmetic average model of the preferred members to forecast the typhoon tracks shows improvement over the ensemble average forecast and the deterministic forecast.The longer the forecast time is,the more obvious the improvement effect is.The errors of typhoon track forecast for 24 h and 48 h are reduced by 4 km and 6 km compared with the ensemble average,and by 5 km and 19 km compared with the deterministic forecast.The optimal selection of probability-matching average precipitation products of ensemble members can improve the traditional ensemble mean precipitation products to a certain extent,especially for the prediction of heavy rainstorm,and the TS scores of 36 h and 60 h time of heavy rainstorm are improved by 10%and 12%compared with the ensemble average.The optimized gale synthesis analysis products of the ensemble members can forecast the movement,strengthening and weakening of the extreme wind area during the impact of typhoons,improving the TS scores of the strong wind exceeding level 6,7 and 8 at the 36 h timescale by 18.3%,15.7%and 13.4%compared with the ensemble average products,and having some forecasting ability for the strong wind above level 9 to 11 or higher.

关 键 词：台风路径 集合预报 概率匹配平均降水产品 

分 类 号：P457.8[天文地球—大气科学及气象学]