摘要
文章对GIS软件中几种常见插值方法的原理进行了详述,剖析了插值过程中参数设置对插值结果产生的影响。以内蒙古多年平均气温插值为例,对比了各种插值方法的优劣,结果表明:普通克里金法的插值结果最能反映内蒙古多年平均气温的分布特征,均方根误差为1.138℃;反距离权重法可反映基本特征但等值线不够平滑、局部有"牛眼"现象,均方根误差为1.260℃;趋势面法不能反映基本特征,均方根误差为1.425℃;使用协同克里金法,将高程作为协同因子对普通克里金算法进行改善,未取得明显效果。
This article introduced several common interpolation methods in GIS software, and analyzed the influence of the parameters setting on interpolation result. To average temperature for example, various interpolation method were contrasted. The results showed that ordinary Kriging method can best reflect the distribution character of average temperature in Inner Mongolia, its RMS error is 1.138℃. The inverse distance weight method can reflect the distribution character of average temperature basically, but the isoline wasn＇t smooth, there was ＂ buphthalmos＂ phenomenon, its RMS error is 1.260℃. Trend surface method cann＇t reflect the basic characteristics, RMS error is 1.425 ℃, using synergy Kriging method, elevation was used cofactor, rdinary Kriging method was improved, the effect wasn＇t obvious.
出处
《内蒙古气象》
2013年第3期11-14,共4页
Meteorology Journal of Inner Mongolia
