GIS技术支持下岩溶区植被指数的Kriging估计

研究如何修复扫描行校正器SLC故障造成的数据坏行区域植被指数信息缺失问题，对全面掌握区域植被指数空间变异有重要的意义。以湖南省慈利县石漠化严重的区域为例，利用遥感图像处理软件ENVI（Environment for Visualizing Images）提取研究区域植被指数。在GIS技术支持下，利用地统计学的方法对坏行区域的植被指数进行了Kriging估计。结果表明，植被指数主要受内在因子的作用，具有强烈的空间自相关性，自相关距离达到501 m；Kriging估计结果表明，植被指数在空间分布表现为条带状和斑块状分布，植被指数较高的区域主要分布在研究区的中南部。Kriging插值验证表明利用地统计学的方法很好地估计了遥感影像坏行区域植被指数的“缺失”信息，提高了ETM+SLC off数据的利用率，为岩溶石漠化评估提供参考。 

Estimates of Vegetation Index in Karst Area Based on GIS and Kriging

How to solve the problem of the SLC off data has a great significance for mastering the spatial variability of karst rock desertification. Vegetation index were extracted in a representative district，which was in Cili county of Hunan province by applying image processing software of ENVI.The spatial variability of the vegetation index was analyzed with geostatistical method.The vegetation index of the study area were impacted intrinsic factors and performed strong spatia1 autocorrelation，and the range was 501 m.The spatial distribution maps of the vegetation index interpolated by Kriging interpolation method showed that the higher indices emerged in the middle and south of the study area，and the spatial distribution of the vegetation index was apparently strips and blocks.Using geostatistical methods to predict the SLC off vegetation index has a very high prediction accuracy，which can improve the utilization ratio of ETM+ SLC off data to some extent and provide the reference to the assessment rocky desertification in the karst area.