中文摘要：

DEM及数字地形分析的精度无疑受到包括高程数据空间内插方法在内的多种因素的影响，本研究利用模拟高斯曲面为基准数据，通过四种不同插值方法结果的对比，分析说明每种插值方法在某一标度范围内插值精度的优劣，从而在应用中针对不同条件采用相应优选的插值方法。在此基础上，对于四种插值方法生成相应的坡度、坡向值，通过与真值的比较探讨插值方法对用DEM提取坡度、坡向的影响。实验结果确定了不同内插方法对坡度和坡向因子提取精度的大小，发现在坡度方面，SPLINE方法内插出来的坡度最精确，其他依次为KRIGING，IDW，TIN；在坡向方面，SPLINE内插出来的坡向最接近真值，是最精确的插值方法，其他依次为KRIGING，TIN，IDW。以上结果为空间插值方法的选择提供了参考。

英文摘要：

DEMs and the terrain features derivation accuracy are usually affected by many factors such as sample points density and distribution, DEM form structure and interpolation methods. Since only the elevation of the sample point can be obtained, interpolation will be used to estimate the simulative value of the other points. In this study, four interpolation methods, IDW interpolation, Spline interpolation, Kriging interpolation and TIN interpolation, are evaluated on a Gauss synthesized surface with peak and valley just similar as a real surface, by comparison with the true values of gradient and aspect. Interpolation result is more sensitive to the errors when gradient is less. The uncertainty of gradient, in the four methods, is the least for SPLINE interpolation, and increases in turn of KRIGING, IDW, and TIN interpolations. And the uncertainty of aspect in the four methods, as that of gradient, is the least for SPLINE interpolation, and increases in turn of KRIGING, TIN, and IDW interpolations. The results provide information for the selection of interpolation algorithms.