权重对空间插值方法的影响分析

GIS软件中包含了一些通用的空间插值方法模板,用户在使用空间插值方法时由于参数选取不当可能带来误差,例如:IDW法的幂指数默认值为2,但实验证明幂指数为2并不一定是最佳的.以反距离加权平均法和样条函数为例进行研究,结果表明:IDW法权重的选择对插值结果有很大的影响,幂指数的选择要根据研究区域的具体情况确定;在插值点稀少或数据缺乏的地区,适当的提高IDW的幂指数值可以提高插值精度;相比IDW法,样条函数法更适合应用于表面逐渐变化的情况,如:温度、高程、地下水位温度、污染浓度等.

Analysis on Impact of Weight to Spatial Interpolation Methods

GIS software consists of some generic interpolation methods model. When the users use the interpolation methods, because of the inappropriate parameters, it leads to unnecessary error. For example, the IDW ,its default value of exponential is 2, but experiments prove that 2 is not the best. The problems were discussed by Anti-distance weighted average method and examples of spine function ,the conclusion shows that,( 1)The choice of weights of IDW has great impact on interpolation results, we must choose the exponent according to the specific circumstances of the regional that we surveyed ; ( 2) In some areas ,the interpolation points are scarce or lack of data, to increase the exponential value of IDW appropriately can improve its precision.; ( 3) Compare to IDW, Spine is more suitable for these situations that surface changes are gradually,such as temperature, elevation, groundwater temperature, concentration of pollution and so on.