甘肃省降水的空间内插方法比较

气象要素的空间内插对于观测台站十分稀少,而台站分布又非常不合理的地区具有十分重要的实际意义。然而在众多的气象要素空间插值方法中,并没有适合每一种气象要素的普适的最佳插值方法[1]。本文以甘肃省52个气象站及相邻省区18个气象站1961———1990年整编资料中的逐月降水量为数据源,选用了反距离加权方法(IDW)、样条函数方法(Spline)、趋势面方法(Treng)、Kriging等4种内插方法,进行了最优方法选取的探讨。交叉验证结果表明: 4种方法中,相对均方差(RMSE)的排列顺序为Kriging

Comparison of Four Precipatation Spatial Interpolation Methods in Gansu

Spatial interpolation of climate data is very important to those areas where the meteorological stations are rare and its distribution is unreasonable. Four methods Kriging、IDW、Spline and Trend were used to interpolate precipatation in Gansu province .The results presented that Kriging method was the best one for spatial interpolation .From the spatial distribution of precipation map obtained by Kriging in Gansu province ,we can see that the rainfull increased from southeast to northwest and that in the summer and autumn is much more than that in other seasons.