基于GIS的沙澧河流域面雨量算法对比分析

为了获取更客观有效的沙澧河流域面雨量计算方法,提高暴雨强度和落区监测与预报的准确率,利用ArcGIS9.3中的空间分析功能,以沙澧河流域1∶25万地理信息数据、16个气象站和217个加密自动雨量观测站的实况雨量资料为背景数据库,选取反距离加权算法、克里金算法、样条函数算法、泰森多边形算法,对沙澧河流域6个分片区的面雨量进行计算,并对计算结果进行了对比分析。结果表明,面雨量计算精度与雨量观测站密度与分布、降雨强度有关。其中,克里金算法计算的流域面雨量精度较高,更适合于整个沙澧河流域面雨量的计算;其次是泰森多边形算法和反距离加权算法,样条函数法算法应用效果较差。当雨量站点分布密集且比较均匀、降水空间分布均匀时,4种算法均适用于面雨量的计算,其中以克里金算法计算的面雨量使用效果最好;当雨量站点分布密集且比较均匀,降雨强度大,降水空间分布不均匀时,反距离加权算法、克里金算法、泰森多边形算法均适用于流域面雨量的计算,其中以泰森多边形算法计算的面雨量效果最好,样条函数法算法不适用此情况下面雨量的计算。当雨量站点分布密集且比较均匀、降水空间分布均匀时,计算的面雨量较站点或降水空间分布不均匀时计算的面雨量更接近实际情况。

Comparative Analysis of Areal Rainfall Algorithms Based on GIS in the Sha River and Li River Basin

In order to obtain more objective and effective areal rainfall calculating method of the Sha river and Li river basin, and improve the accuracy of monitoring and predicting the intensity and drop zone of the heavy rainfall, based on the 1：250,000 geographical information, and the rainfall data col- lected from 16 meteorological stations and 217 encrypted automatic rainfall observation stations, by using the ArcGIS9.3 spatial analysis function, the areal rainfall in 6 subregions in the Sha river and Li river basin are calculated and comparatively analyzed by using 4 different algorithms, i. e. Inverse Distance weighted（IDW） , Kriging interpolation algorithm （ Kriging）, Spline function algorithm （Spline） , and Thiessen polygon algorithm（Thiessen）. The results show that： the estimated accuracy of areal rainfall is influenced by the precipitation intensity, the density and distribution of the rainfall observation stations. Kriging is the most accurate algorithm for estimating the rainfall in the river basin and suit for estimating the rainfall of the Sha river and Li river basin. Thiessen and IDW are the second best algorithm while Spline is least accurate. When the distribution of rainfall observation stations is dense and uniform, the spatial distribution of rainfall is even, the 4 algorithms mentioned are suit for the calculation of areal rain- fall, and Kriging is the best. When the precipitation observation stations distribution is dense and uniform, rainfall is heavy, rainfall spatial distribution is uneven, IDW, Kriging and Thiessen can be used in the calculation of areal rainfall, and Thiessen is the best algorithm, Spline is not suit for the calculation. When the precipitation observation stations distribution is dense and uniform, spatial distribution of rainfall is even, the calculated areal rainfall is more close to the actual situation.