摘要: 利用黄河流域143个气象站的逐日降水和气温资料，应用经验正交分解法(EOF)、趋势及小波分析法对STARDEX极端气候核心指标进行时空分析，结果表明，时空间分布特征明显的是极端气温即暖日阈值、冷日阈值、霜冻日数和最长热浪天数，整个流域呈现了南北或东西向的渐变和中间高两头低或相反的空间变化特征，综合表现为流域中部极端气温事件呈现增多的趋势，即暖日阈值、冷夜阈值和最长异常热浪天数增加，霜冻日数减少，且除霜冻日数外，其他三个指数在20世纪90年代和20世纪末均有突变发生，20世纪末后加强。极端降水的分析表明极端降水事件相对较弱，前3个模态形式表现为上、中、下游不同步的渐变和中高两头低或中低两头高的空间分布形态，时间系数的年际和年代际变化表明较强降水阈值、较强降水日数、最大5日降水量和日降水强度呈现不显著的增加和减少趋势。
Abstract: The daily rainfall and temperature data of 143 weather stations in the Yellow river are used for verifying the space-time distribution of extreme climate by using empirical orthogonal decomposition (EOF) and wavelet analysis method with STARDEX core index. The results show that the spatial distribution characteristics txq90, tnq10, tnfd and txhw90 are obvious, the entire basin shows distribution shape of north-south, east-west gradient and middle high both sides low or middle low both sides high. Increase tendency is shown for extreme temperature, that is to say, txq90, tnd10 and txhw90 show increase and tnfd shows decrease. Except for tnfd, the others three indexes happened abrupt change in 1990s. Analysis of precipitation suggests that the extreme precipitation events are relatively weak. The former three modal forms show gradient spatial change distribution pattern of the upper, middle and lower reaches and middle high both sides low or middle low both sides low. The time coefficient change of interannual and interdecadal shows that pq90, pnl90, px5d and pint have no-significant increase and decrease change trend.