Yang Qi-Yong, Jiang Zhong-Cheng, Ma Zu-Lu, Cao Jian-Hua, Luo Wei-Qun, Li Wen-Jun, Duan Xiao-Fang
Institute of Karst Geology, Chinese Academy of Geological Sciences, Guilin 541004, China.
Huan Jing Ke Xue. 2012 Apr;33(4):1404-8.
In order to master the spatial variability of the normal different vegetation index (NDVI) of the peak cluster karst area, taking into account the problem of the mountain shadow "missing" information of remote sensing images existing in the karst area, NDVI of the non-shaded area were extracted in Guohua Ecological Experimental Area, in Pingguo County, Guangxi applying image processing software, ENVI. The spatial variability of NDVI was analyzed applying geostatistical method, and the NDVI of the mountain shadow areas was predicted and validated. The results indicated that the NDVI of the study area showed strong spatial variability and spatial autocorrelation resulting from the impact of intrinsic factors, and the range was 300 m. The spatial distribution maps of the NDVI interpolated by Kriging interpolation method showed that the mean of NDVI was 0.196, apparently strip and block. The higher NDVI values distributed in the area where the slope was greater than 25 degrees of the peak cluster area, while the lower values distributed in the area such as foot of the peak cluster and depression, where slope was less than 25 degrees. Kriging method validation results show that interpolation has a very high prediction accuracy and could predict the NDVI of the shadow area, which provides a new idea and method for monitoring and evaluation of the karst rocky desertification.
为掌握峰丛岩溶地区正常植被指数（NDVI）的空间变异性，考虑到岩溶地区遥感影像存在山体阴影“信息缺失”问题，运用图像处理软件ENVI，在广西平果县果化生态实验区提取无阴影区域的NDVI。运用地统计方法分析NDVI的空间变异性，并对山体阴影区域的NDVI进行预测与验证。结果表明，受内在因素影响，研究区NDVI呈现出较强的空间变异性和空间自相关性，变程为300 m。通过克里金插值法得到的NDVI空间分布图显示，NDVI均值为0.196，具有明显的条带状和块状分布特征。较高的NDVI值分布在峰丛地区坡度大于25度的区域，而较低值分布在峰丛山脚和洼地等坡度小于25度的区域。克里金法验证结果表明，插值具有很高的预测精度，能够预测阴影区域的NDVI，为岩溶石漠化的监测与评价提供了新的思路和方法。
