桂林海洋岩溶区植被覆盖度的多时相遥感预估模型

摘要

Evolution of vegetation coverage is one of the characteristic indicators to quantify the development of rocky desertification in karst region,and macro-distribution and trends are important for environmental change. Landsat remote sensing images of five phases in Guilin Haiyang karst experimental area are from 2000,2001, 2006,2008 and 2009 in dynamic analysis for local topography.Renewed steep drop,two-division model is pro-posed and amendment NDVI vegetation index to estimate the corresponding year vegetation coverage are 0.223 39,0.278 77,0.353 664,0.368 54 and 0.373 96.Vegetation coverage is based on the value of the estab-lishment of a corresponding decision tree model and divided into five grades.Vegetation coverage study area cov-ers moderate -low coverage level in 2000,and moderate -high coverage level in 2009.Fast remote sensing veg-etation coverage data from quantitative experiments establishes predictive model for estimation results in the study area.According to the prediction model for future environmental analysis,vegetation coverage is increas-ing year by year to a stable saturation value.%植被覆盖度的演变规律是喀斯特地区石漠化程度量化的特征指标之一，其宏观分布与变化趋势是认识环境变化的重要参数。以桂林海洋岩溶试验区2000、2001、2006、2008及2009年5个时相的 Landsat遥感影像为动态分析源，针对陡升陡降的地形地貌特征，提出了修正 NDVI 的二分模型，估算出对应年份的植被覆盖度分别为0．22339、0．27877、0．353664、0．36854、0．37396；根据植被覆盖度值建立了相应的决策树分类模型并划分出5个等级，研究区的植被2000年为中—低覆盖级，2009年为中—高覆盖级，以提取的实验数据建立的遥感快速植被覆盖度定量估算预测模型与研究区的实测结果符合。根据预测模型对未来环境的演变进行了相应的预估分析，植被覆盖度在逐年提升并直至达到稳定的饱和值。