王汉东，黄瓅瑶，朱思蓉，罗 斌

(长江勘测规划设计研究有限责任公司，湖北 武汉 430010)

　　摘 要：空间插值是常用的面雨量计算方法，以三峡区间为研究对象，选取反距离加权、普通克里金、协同克里金3种空间插值算法，进行面雨量空间插值计算和对比分析，其中协同克里金插值将地形作为影响因素。三峡区间面雨量插值结果表明，克里金插值在精度上要优于反距离加权插值，插值效果也更平滑;协同克里金的插值精度优于普通克里金，平均绝对误差、误差均方根相对于普通克里金插值分别降低了19%，10%。因此，在山区或地形复杂区域的面雨量计算，将地形作为协同区域化属性的协同克里金插值更合适。

　　关键词：面雨量;空间插值;对比分析;反距离加权插值;普通克里金插值;协同克里金插值

　　Intercomparison of area rainfall spatial interpolation in SanXia Region

　　WANG Handong, HUANG Liyao, ZHU Sirong, LUO Bin

　　(Changjiang Institute of Survey, Planning, Design and Research, Wuhan 430010, China)

　　Abstract: Spatial interpolation methods are often used to calculate the area rainfall. The following interpolation methods are used to calculate the area rainfall of SanXia region: inverse distance weighting (IDW), ordinary Kriging and co-Kriging, and their results are compared and analyzed. The terrain of the region is applied to improve the interpolation accuracy in the co-Kriging method. The results of area rainfall spatial interpolation in SanXia Region show that: Compare with IDW method, the two Kriging methods’ interpolation precision are better. The spatial distribution of interpolation results calculated by the two Kriging methods are smoother also. The accuracy of co-Kriging is better than the ordinary Kriging. The mean absolute error (MAE) of the co-Kriging is less than the ordinary Kriging by a rate of 19%. And the root mean squared interpolation error (RMSIE) of the co-Kriging is less than the ordinary Kriging by a rate of 10%. Because the co-Kriging method allows for the topographic effects on rainfalls. The co-Kriging interpolation method is more suitable for those regions which rainfalls are significantly affected by topography.

　　Key words: area rainfall; spatial interpolation; comparative analysis; inverse distance weighting interpolation; ordinary Kriging interpolation; co-Kriging interpolation