A new technique for automatically locating the center of tropical cyclones with multi-band cloud imagery

作者：Xiaoqin LU,Hui YU,Xiaoming YANG,Xiaofeng LI,Jie TANG

摘要：A spiral cloud belt matching(SCBeM)technique is proposed for automatically locating the tropical cyclone(TC)center position on the basis of multi-band geo-satellite images.The technique comprises four steps:fusion of multi-band geo-satellite images,extraction of TC cloud systems,construction of a spiral cloud belt template(CBT),and template matching to locate the TC center.In testing of the proposed SCBeM technique on 97 TCs over the western North Pacific during 2012-2015,the median error(ME)was 50 km.An independent test of another 29 TCs in 2016 resulted in a ME of 54 km.The SCBeM performs better for TCs with intensity above“typhoon”level than it does for weaker systems,and is not suitable for use on high-latitude or landfall TCs if their cloud band formations have been destroyed by westerlies or by terrain.The proposed SCBeM technique provides an additional solution for automatically and objectively locating the TC center and has the potential to be applied conveniently in an operational setting.Intercomparisons between the Automated Rotational Center Hurricane Eye Retrieval(ARCHER)and SCBeM methods using events from 2014 to 2016 reveal that ARCHER has better location accuracy.However,when IR imagery alone is used,the ME of SCBeM is 54 km,and in the case of low latitudes and low vertical wind shear the ME is 45-47 km,which approaches that of ARCHER(49 km).Thus,the SCBeM method is simple,has good time resolution,performs well and is a better choice for those TC operational agencies in the case that the microwave images,ASCAT,or other observations are unavailable.

发文机构：Shanghai Typhoon Institute Ocean Department CAS Key Laboratory of Ocean Circulation and Waves Center for Ocean Mega-Science

关键词：tropicalCYCLONECENTERlocationgeostationarysatellitematchingtechnique

分类号： P44[天文地球—大气科学及气象学]