河南省雨季降水时空变异特征分析

降水的时空变异分析是认识区域水资源形成与演变的重要方法。时空变异特征分析不仅可以系统地对降水的时间序列进行分析,而且能从空间上把握降水的分布格局。本文将河南省近51年雨季降水资料,结合数字高程模型(DEM),利用回归分析、空间自相关分析、空间插值模拟及交叉验证等,对河南省降水时空变异特征进行分析。结果表明:(1)河南省雨季降水整体来看呈增加趋势,近年来尤为明显;但9月份表现异常,呈下降趋势。(2)月降水量差异明显,最大降水量在7月份,平均达到178.3 mm;(3)在空间上降水呈现出明显的南多北少,东多西少的格局;有明显的集聚特点,在南部以罗山、潢川为中心形成降水丰沛聚集区,北部以辉县为中心形成降水稀少聚集区;林县、栾川和西峡表现为空间例外,明显高于相邻区域的降水量。

Analysis on the Spatio-temporal Variability of Rainy Season Precipitation in Henan Province

With the extensive application of geographic information systems and the deeply development of geography disciplines, the spatial and temporal structure and process analysis are receiving more attention. The analysis of spatio-temporal variability of precipitation is the basis for the understanding of formation and development of regional water resources. It not only reveals the change of time-series, but finds the spatial structure and changing pattern. Thus, it provides the basis for predicting the drought and waterlogging. Based on the rainy season precipitation data of the past 51 years in Henan Province, combined with a digital elevation model (DEM), using regression analysis, spatial autocorrelation, simulation of spatial interpolation, and cross-validation, we conducted an analysis of spatial and temporal variability of precipitation in Henan Province. The result reveals that: (1) the trend is clear, and the rainy season precipitation in Henan Province overall has shown an increasing trend and in recent years it is particularly evident; (2) The differences of monthly precipitation are obvious, the maximum value is in July, and the average reaches 178.3 mm. (3) The spatial variance exists. There is a clear pattern that the precipitation in the south and east are more than that of north and west in spatial. There is a strong clustering characteristic that in the south, Luoshan and Huangchuan counties as the center formed the rainfall abundant areas, while in the north, Hui County as the center formed the rainfall scarce areas. Lin, Luanchuan and Xixia counties as spatial outliers, are significantly higher than the adjacent regional precipitation. After the spatial autocorrelation analysis, the spatial and temporal anisotropy can be acquired. Therefore, according to the spatio-temporal analysis, we get the interpolation map with Cokriging method, and it tallies with the prior conclusion.