摘要
以美国爱达荷州为例,根据105个气象站点及其30年年平均降水量数据,采用反距离加权平均法、普通克立金法、规则样条函数法及趋势面法等方法进行插值,分析了不同插值方法中站点数量变化、像元尺度变化对降雨数据空间插值结果的影响。通过对数据和精度指标的对比分析,得出每一种插值方法在某一标度范围插值的精度及适用范围。研究表明:①插值精度并不与样本点数量成正比例关系;②像元尺度的变化对空间插值精度的影响很小;③不同插值方法对插值结果有很大的影响,规则样条函数具有优秀的插值精度。要得到最理想的插值结果,须对不同研究区,进行实测样本数据进行分析,反复试验比较,以选择最佳插值方法。
Taking Idaho State in the USA as an example,according to the precipitation data of the 105 meteorological stations in Idaho as well as their 30-years annual average rainfall data,adopting inverse distance weighted,ordinary Kriging,regularized-spline function and trend method to interpolation,it has analyzed different effects on the results of spacial interpolation of rainfall data caused by the change of the sites' number,pixel sizes of different interpolation methods.The interpolation precision in different ranges of every interpolation method with its applicable range has been concluded through a comparison between data and precision index.The research has shown:(1)there is no proportionality between interpolation precision and samples' number;(2) the change of pixel size has little effect on the interpolation precision;(3) different interpolation methods have great effects on the interpolation results with regularizedspline function method having an excellent interpolation precision.Ideal interpolation results need analysis on mearsured samples' data of different studying regions and repeated-experiments comparison for choosing the best interpolation method.
出处
《矿产与地质》
2007年第6期684-687,共4页
Mineral Resources and Geology
基金
国家自然科学基金项目(40671153)资助项目
