小流域时段降雨量空间插值方法研

在作小流域分布式水文模型研究时，首先需要对时段降雨进行空间插值。本文基于时段降雨与年（月）降雨量的不同特点，以桑园流域为例，结合空间信息技术，通过分析小流域时段降雨与经度、纬度、高度等因子的内在关系，提出了一种适用于小流域的时段降雨空间插值方法——回归距离平方反比法（RIDS法），并与距离平方反比法（IDS法）和梯度距离平方反比法(GIDS)作了对比。通过对桑园流域2003年512个时段降雨进行空间插值对比试验，计算结果表明RIDS法的平均相对误差为4.0％，IDS法的平均相对误差为9％，GIDS法的平均相对误差为30.4％，本文提出的RIDS方法的计算结果明显优于IDS法和GIDS法。

Spatial Interpolation Methods for Hourly Precipitation in Small Basin

When searching for the distributed watershed hydrological model in a small basin,first of all we need to estimate area precipitation from point hourly precipitation through spatial interpolation.In this paper,the authors analyze the inherent relations between hourly precipitation and the longitude,latitude and altitude,and raise a method,Inverse Regression Distance Square(IRDS),which is suited to interpolate the hourly precipitation in the small basin.Taking Sangyuan Basin as a study area,this paper interpolates 512 hourly rainfall data in 2003 using the method developed and compares the results with those of Inverse-Distance-Squared(IDS) method and Gradient-plus-inverse-distance-squared(GIDS) method.The mean relative error of IDS method is about 9%,and the mean relative error of GIDS method is about 30.3%,while the mean relative error of IRDS method is only about 4%.Results show that the IRDS method proposed in this paper has evident advantages over the other two methods.