滇池流域降水时空变异特征分析

降水是区域水资源形成的主要影响因素,其时空变化趋势也直接影响着各种生态系统的结构、服务功能及空间分布与演变。降水时空变异分析是认识区域水资源形成与时空演变的主要手段和方法。本文利用滇池流域及周边雨量站逐月数据,采用回归分析、距平、空间相关性分析、Mann-Kendall检验、Co-kriging插值及交叉验证等方法,对1953-1987年和2007-2012年2个时序系列的时空变异特征分析结果表明：（1）1953-1987年春、秋和冬季降水量有升高趋势,夏季呈减少趋势,但各季节的增减趋势不显著,2007-2012年春、夏、冬季呈减少趋势,秋季为增加趋势,近期降水量明显有减少趋势;（2）1953-1987年流域降水量呈现增加趋势（11.12 mm/10a）,大致经历下降-上升-下降过程,2007-2012年流域降水量呈显著的锯齿状减少趋势,处于枯水期;（3）1953-1987年各时段的雨量主要呈现负相关性（不显著）,2007-2012年间呈现正相关性,通过LISA统计分析认为,空间异质性随地理位置和时间而变化;（4）年均降水量与雨季降水量的空间分布特征基本相似,出现2个降水高值区和2个低值区对顶分布态势。但在2007-2012年,降水量的高值范围有所减少,低值区范围相应有所扩大。

An Analysis on the Spatio-temporal Variability of Precipitation in Dianchi Basin

Precipitation is one of the most important factors of climate, and the analysis of spatio-temporal variability of precipitation is a primary way to understand the formation and development of regional water resources. Furthermore, the trend of the spatio-temporal variability of precipitation directly affects the structure, service function and spatial distribution of various ecosystems. Based on the monthly precipitation data from 1953 to 1987 and from 2007 to 2012, using regression analysis, anomaly, spatial autocorrelation, Mann-Kendall test, Co-kriging interpolation considering the influence of elevation (DEM), and cross-validation, we conducted an analysis on the spatial and temporal variability of precipitation in Dianchi basin. The results are stated in the following sections: (1) from 1953 to 1987, the precipitation tends to increase in spring, autumn and winter, but decrease in summer; from 2007 to 2012, however, the precipitation tends to decrease in all seasons except in autumn. (2) An increasing trend was discovered in the annual precipitation (11.12mm/10a) during the 35 years, and it generally revealed a fluctuation pattern of “down-up-down”; however, a downward trend of precipitation is significant from 2007 to 2012. (3) The indices of Moran’s I are negative during the 35 years, which reveals a main negative correlation that is different from the period of 2007 to 2012. The analysis of LISA shows that the spatial heterogeneity tends to change with respect to the geographical location and time. (4) The spatial distribution of annual rainfall is similar to the rainfall during the rainy season, which is a cross-distribution indicated by two high-value and two low-value areas. In addition, the spatial extent of heavy rain events is decreasing, and the spatial extent of light rain events is increasing.