淮河流域夏季极端降水频次空间分布的客观分类及其形成机理

本文基于1961～2016年淮河流域四省（江苏、安徽、河南、山东）逐日降水观测资料及全球大气再分析资料，利用K均值聚类、旋转经验正交函数分解对淮河流域夏季极端降水频次分布进行了客观分类，利用统计诊断和数值模拟的手段讨论了其相关环流异常和形成机理。结果表明:（1）淮河流域夏季极端降水频次的空间分布可客观分为以极端降水主要发生在淮河流域33°N以南地区的南部型，主要发生在32°～36°N之间的中部型，和主要发生在34°N以北的北部型这三种分布类型；（2）南部型极端降水频次分布与西北太平洋副热带高压异常偏西偏南有关，西北太平洋异常反气旋北侧的异常气旋性环流使得水汽输送停留在淮河流域南部，导致南部极端降水频次偏多。中部型对应淮河流域受鞍型场环流结构控制，导致中部极端降水频次偏多。北部型极端降水频次分布时，淮河流域处于反气旋性环流异常西南侧，偏南风将水汽输送至淮河流域北部，导致北部极端降水频次偏多；（3）南部型和北部型的极端降水频次分布相关环流异常分别受厄尔尼诺和拉尼娜相关海表温度异常所影响，而中部型极端降水频次分布的相关环流异常是巴伦支海/喀拉海海冰异常在欧亚大陆上空激发的自西北向东南传播的准定常罗斯贝波所导致的。

Objective Clustering of Spatial Patterns of Summer Extreme Precipitation Frequency over the Huaihe River Basin and Their Formation Mechanisms

Based on the daily precipitation over the Huaihe River basin and reanalysis datasets from 1961 to 2016, the spatial patterns of the summer extreme precipitation frequency over the Huaihe River basin are objectively classified through the K-means cluster and rotated empirical orthogonal function analysis. Their formation mechanisms are investigated via observational diagnosis and numerical simulations. Results show that: (1) The spatial distribution of the summer extreme precipitation frequency over the Huaihe River basin can be objectively divided into three types: S-Type, C-Type, and N-Type, in which the extreme precipitation mainly falls south of 33°N, 32°–36°N, and north of 34°N, respectively. (2) The S-Type is related to the anomalous southwestward shift of the northwestern Pacific subtropical high (NWPSH). Cyclonic anomalies north of NWPSH hinder the water vapor further to the north, resulting in more extreme precipitation in the south of the Huaihe River basin. For the C-Type, the Huaihe River basin is controlled by saddle circulation anomalies pattern. Therefore, extreme precipitation is mainly confined to the central Huaihe River basin. The N-Type is associated with the northwestward shift of the NWPSH. The southerly wind could transport water vapor to the north of the Huaihe River basin, resulting in more extreme precipitation in the region. (3) The formation and the related circulation anomalies of the S-Type and N-Type are closely associated with El Ni?o-like and La Ni?a-like sea surface temperature anomalies, while the C-Type-related saddle circulation pattern is linked to two southward-propagated stationary Rossby waves over the Eurasian continent induced by the Arctic Sea ice anomalies over the Barents Sea/Kara Sea.