区域水土流失研究中降雨量空间插值方法选取  

Selection of the Spatial Interpolation Methods for Precipitation in the Regional Soil Erosion

作　　者：汪翠英[1,2] 杨勤科[1,2] 郭伟玲[1,2] 姚志宏[1,2] 李文凤[1,2] 

机构地区：[1]中国科学院水利部水土保持研究所,陕西杨陵712100 [2]西北农林科技大学资源环境学院,陕西杨陵712100

出　　处：《水土保持研究》2008年第2期88-91,共4页Research of Soil and Water Conservation

基　　金：国家重点基础研究发展计划“973计划”项目(2007CB407204);“十一五”国家科技支撑计划项目(2006BAD09805);黄委会治黄专项(2004SZ01-04)

摘　　要：在区域水土流失模型研究中,空间插值可提供每个计算栅格的气象要素资料。考虑到研究区域降雨与高程相关性很弱,不宜采用梯度距离反比法(GIDS),故采用距离反比法(IDW)和普通克里格法(Kriging),对延安示范区及其周围共50个站点2000—2003年的5—10月逐月降雨量进行插值。交叉验证结果表明:对2种插值方法,二者经过对数变换后平均相对误差(MRE)为8.30%和7.67%,分别比原始数据插值后的MRE下降了23.17%和23.50%,说明插值精度得到了提升,对研究区域某一年逐月降水的插值Kriging方法比IDW方法更加精确。Every cell＇s climatic information can be used by interpolation method when researching district soil erosion model,it is important to compare the results obtained using alternative methods applied to each data set. In this study, two methods for spatial interpolation of precipitation data were compared using 2003 year six months precipitation from the Yan,an demonstration area was interpolated using inverse distance weighted （IDW） and Universal Kriging methods. Corresponding cross-validation test show that a log-transformation improves the accuracy,by a log-transformation the value of MRE is 8.30% and 7.67% in the methods of IDW and Kriging respectively that is lower 23.17% and 23.50% than the result of original data. In addition,the method of GIDS can not be used in the study. Universal Kriging method is more accurate in predictions than IDW in this area.

关 键 词：普通克里格法 距离反比法 降雨 

分 类 号：S157[农业科学—土壤学]