岩溶塌陷地层的样本数据中常含有大量的客观存在的特异值,普通克里格法对此类数据的处理存在自身难以克服的缺陷。本文首先介绍了指示克里格的基本原理,运用空间数据分析技术对样本数据的空间分布和空间结构等性质进行分析,确定了最佳模型和相关参数,最后以徐州市塌陷区为例,在GIS平台上构建了基于指示克里格法的复杂岩溶地面塌陷区特征提取模型,通过交叉验证法证实了该模型在此研究范围的适用性,同时这一方法的估计方差较小并且标准均方根误差接近于1,表明指示克里格法在塌陷区高程插值中具有较高的插值精度和较好的稳定性。 Sample data of karst collapse strata often contain a large number of objective values ​​that exist objectively. Ordinary Kriging methods have some inherent flaws in dealing with such data. In this paper, the basic principle of Krieger Indication is introduced firstly. The spatial data analysis technology is used to analyze the spatial distribution and spatial structure of the sample data, and the best model and related parameters are determined. Finally, taking Xuzhou Subsidence Area as an example, The model of complex karst ground subsidence based on the indicator kriging is constructed on the GIS platform. The cross validation method is used to verify the applicability of the model in this research area. The variance of this method is small and the standard mean square The root error is close to 1, indicating that the kriging method has higher interpolation precision and better stability in elevation interpolation of subsidence area.