长江中下游地区雪和雨夹雪垂直热力特征的聚类分析

Clustering Analysis of Vertical Temperature Characteristics of Snow and Sleet Phase in Eastern Yangtze River Region

基于ERA5逐小时再分析资料和地面观测资料,对2019—2022年长江中下游地区雪和雨夹雪的垂直热力特征进行分析,并应用K-means聚类方法分型探讨其物理过程特征。结果表明:1)雪和雨夹雪的聚类均可分为“冷型”和“逆温型”两簇,“冷型”降雪、雨夹雪多数出现在28°N以北地区,江南地区以“逆温型”为主。2)通过湿球温度曲线的聚类分析可以较好地区分雪、雨夹雪相态,优于样本合成平均。3)大部分雪、雨夹雪的云顶温度都低于-10℃,且量级越大,云顶温度越低,“冷型”雪、雨夹雪云顶温度低于“逆温型”。4)“逆温型”雪和雨夹雪的降雪率高于对应的“冷型”样本,且“逆温型”南风、北风风速均强于“冷型”。5)“逆温型”雪和雨夹雪比“冷型”样本出现暖层比例高,且“冷型”样本暖层绝大部分都位于近地面层。

Based on hourly ERA5 reanalysis data and surface observations,the vertical temperature characteristics between snow and sleet over eastern Yangtze River region from 2019 to 2022 winters were analyzed,and the different physical processes were discussed using K-means clustering method.The results showed that:1)Two clusters were found through clustering analysis of snow and sleet:“cold”and“inversion”clusters.The cold cluster of snow and sleet often happened to the north of 28°N,while the warm cluster often happened to the south.2)The cluster method of wet-bulb temperature profiles better discriminated between snow and sleet phase types than sample averaging.3)The cloud-top temperature of most snow and sleet samples was lower than-10℃,and the larger the magni-tude,the lower the cloud top temperature.The cloud top temperature of“cold”snow and sleet was lower than that of“inversion type”.4)The snowfall rate of“inversion type”snow and sleet was higher than that of the corresponding“cold type”samples,and the south and north wind speeds of“inversion type”were stronger than those of“cold type”.5)Warm layers of“inversion type”snow and sleet presented higher occurrence ratio than that of“cold type”samples,and most of the warm layers from the cold cluster samples hap-pened near the surface.