基于新疆51个站点1960-2005年的日降水资料,从最大连续降水事件出发,以年、夏、冬为研究时期,定义描绘最大连续降水事件的日数、降水量和降水强度的9个极端降水指标,研究最大连续降水事件的时空概率特征。本文应用改进的Mann-Kendall法对各指标变化趋势进行检验,采用基于F检验的线形分析计算其变化率。研究结果表明:(1)年和夏季最容易发生2天的最大连续降水天数,最大连续降水事件日数越长,降水强度越低;冬季易发生1天的最大连续降水天数,随最大连续降水事件日数的增加,降水强度增加;(2)近年来,日数短的连续降水天数事件频率减少,而随降水日数的增加;降水量有增加趋势;因此,新疆降水有极端化的趋势;(3)新疆有湿润化趋势,而南疆在夏季的湿润趋势比北疆明显,北疆在冬季比南疆显著。
Based on the daily precipitation data from 51 rain stations in Xinjiang during 1960-2005,the spatial and temporal possibility variations of maximum consecutive wet days in the whole year,summer and winter seasons are studied.Nine precipitation extreme indices which reflect the attributes of consecutive maximum wet days are defined,modified Mann-Kendall test is applied to detect the tendencies,and changing rates of indices are evaluated through linear regression with F test.Results show that:(1) two days consecutive wet days occurred most frequently during the year and summer,and the precipitation intensities decreased as the number of days of maximum consecutive wet days increased;in winter,one day consecutive wet days had the maximum possibility,and the intensities increased as the number of days of the maximum consecutive wet days increased;(2) the possibilities of consecutive wet days with short durations reduced,while those of the large number of days increased;the precipitation of consecutive wet days had increasing trends;the intensities of all numbers of days increased;(3) a wet tendency was identified in Xinjiang;wet trend in Southern Xinjiang was more significant than in Northern Xinjiang in summer,while in winter the wet tendency in Northern Xinjiang was more remarkable.