Spatial correlation length of summer extreme heat stress over eastern China

This study attempts to measure the spatial correlation length (SCL) of summer extreme heat stress in any location by using a characterized scale identification method. Daily datasets of multiple meteorological variables from 2,134 observation stations over eastern China during 1961-2010 were used. Three types of heat indexes (a total of seven indexes) were applied to characterize the heat stress. The first type used a single variable, that is, daily maximum temperature (T-max) or daily minimum temperature (T-min), while the second used mean temperature (T) and relative humidity, and the third used T, vapour pressure, and 10-m wind speed. A 90th percentile of the climatology of local heat stress was applied to identify hot days. The SCLs of heat stress were analysed in three regions: North China (NC), the Yangtze River Valley (YRV), and South China (SC). Results showed that the trend changes in heat stress had obvious temporal and geographical characteristics, especially in NC and YRV. Generally, the SCLs of heat stress in NC were the largest, reaching more than 440 km for T-max, with YRV second, about 350 km, and SC the smallest, only about 185 km. This phenomenon could be found for almost all indexes. Moreover, the SCLs of the first two types of heat indexes (except T-min) for the three regions were greater than that of the third one, particularly in NC and YRV, which was related to inconsistent changes in the variables used and the primary role of which one characterized the heat stress. The spatial distributions of high SCLs for all indexes were in line with that of the major urban agglomeration in eastern China. The SCL of heat stress for a location was related not only to the changes themselves, but also to the surroundings, involving the number and spatial distribution of hot days.