基于均一化资料的中国极端地面气温变化分析

Analysis of Extreme Temperature Changes in China based on the Homogeneity-Adjusted Data

利用经过质量控制和均一化处理的中国693个气象站点1961-2014年逐日最高气温、最低气温和平均气温资料,对极端气温的空间变化进行了分区划分,并分析了11个极端气温指数的变化特征。结果表明:极端高温指数(极端最高气温、夏季日数、暖夜日数、暖昼日数、持续暖期和生长日)在各区呈增加趋势,而极端低温指数(极端最低气温、霜冻日数、冷夜日数、冷昼日数、持续冷期)在各区呈减小趋势(极端最低气温除外,呈增加趋势),其中夜间极端气温指数变化程度最大(暖夜日数增加和冷夜日数减小最显著);每个指数在大部分区域均有明显的趋势转折,极端高温指数的转折时间基本在1995-1998年之间,而极端低温指数的转折时间在1985-1986年和1995年前后。极端高温指数转折前/后有减小/增加趋势,而极端低温指数正好相反(但极端最低气温除外,转折前/后有减小/增加趋势)。几乎所有区域的极端气温指数均与平均气温呈显著的相关,其中阈值指数与平均气温的相关性最好。

Based on homogeneity-adjusted and quality-control daily maximum, minimum and mean temperature data of 693 meteorological stations in China from 1961 to 2014, then divided the area of China into several re- gions due to the spatial variability of extreme temperature, and analyzed the change characteristic of 11 extreme temperature indices （were defined by World Meteorological Organization） by linear trend and cumulative depar- ture methods. The result shows that an increasing trend occurs on extreme high temperature indices （including ex- tra-maximum air temperature, summer days, warm nights, warm days, a continuous warm stage and the grow- ing season length） in all regions, while extreme low temperature indices （including extra-minimum air tempera- ture, frost days, cold nights, cold days and continuous cold stage） present a decreasing trend, except for extra- minimum air temperature, which shows a trend of increase. Among them, the degree of change of extreme tem- perature indices in night is the greatest （increase of warm nights and decrease of cold nights are the most signifi- cant）. Furthermore, in most of regions of each index have an obvious trend turning. The turning time of extreme high temperature indices is nearly in 1995 to 1998, while extreme low temperature indices is nearly in 1985 to 1986 and 1995. Extreme high temperature indices show a decreasing/increasing trend before/after turning, but ex- treme low temperature indices are adverse, except for extra-maximum air temperature that is increased/decreased before/after turning. There is a significant correlation between all extreme indices with average temperature near- ly in all regions. The correlation between the threshold indices with average temperature is the most significant.