中国近40年日平均气温的概率分布特征及年代际差异

Probability Distribution of Mean Daily Temperature and Inter-decadal Variation in China in Recent 40 Years

利用全国129站日平均气温资料,从偏态系数、峰度系数入手分析了日平均气温的概率分布特征及年代际差异,结果表明：近40年中国各季日平均气温的均值分布大致呈南高北低,夏季日平均气温的日际变化最小,四季日平均气温不服从正态分布的地理差异明显。1961-1975年时段至1976-2000年时段,夏季、春季日平均气温不服从正态分布的范围明显加大,冬季不服从正态分布范围,北部加大,南部减小,秋季与冬季大致相反;在夏、春、秋季日平均气温方差的变化对日平均气温概率分布的影响是主要的,均值变化的影响次之,冬季在黄河中下游和长江下游流域均值的变化对概率分布影响是主要的,方差变化影响次之,其他地区相反。

Using 129 stations mean daily temperature data, probability distribution of mean daily temperature and inter-decadal variation are analyzed by deflection coefficient and kurtosis coefficient. The result is that： in recent 40 years mean of mean daily temperature of each season approximately assume high in south and low in north, daytime change is the least in summer, geographical diversity of unconformity normal distribution is evident to four seasons. From period 1961--1975 to period of 1976--2000, unconformity areas to normal distribution is increase in summer and spring, unconformity areas to normal distribution in winter increase in north of China and minish in south of China, change of unconformity areas to normal distribution in full approximately reverse to winter. Influence of change of variance to probability distribution of mean daily temperature is primary, influence of change of mean is secondary in summer, spring and autumn. Influence of change of mean to probability distribution of mean daily temperature is primary, influence of change of variance is secondary in middle and underside reaches of Yellow River and underside reaches of Changjiang River in winter, the other area reverse.