兴隆山国家级自然保护区的植被指数及其变化特征

Changes in Vegetation Indices of the Natural Forest Ecosystem in the Xinglong Mountain National Nature Reserve

运用ENVI中感兴趣区、Arcgis中AML编程、基于境内独立水体和裸露路面反射率进行大气影响校准等方法，计算并自动统计出兴隆山国家级自然保护区各小班2000-2002年7～8月的3种植被指数（差值指数、归一化指数和绿度指数）及生态因子的平均值，建立多种要素和监测指标的空间数据库，旨在探索复杂地形条件下植被指数的年际变化，为该区域的森林生态系统及退耕还林监测奠定数据对比与技术基础。研究发现：2001年和2002年各林种的差值指数和绿度指数均比2000年有所减少，只有2001年的NDVI比2000年增加1．11％。研究认为，就年际变化幅度来说，阔叶林3种指数的最大说明其对生态环境变化的敏感性；针叶或针叶混交林处于第二位，NDVI变化值只有前者的1／3和1／7，适应性和稳定性较前者高；各种未成林和宜林地变化较小，其适应性最高。如此大的变化幅度可能与生态因子剧烈变化有关。建议适时更新植被调查，进一步完善遥感监测手段，建立植被指数并与多种生态因子时间序列上的映照关系。

This study is aiming at exploring the annual changes in vegetation Indices under complex topographic conditions of the Xinglong-Mountain National Nature Reserve featuring mainly natural forest, and serves as data baseline for monitoring the forest eco-system, and comparing to the ecological restoration from tillage retirement. With methods of the ROI（region of interesting） in ENVI4.0, ARC/ INFO AML programming, and atmospheric calibration based on the reflectance of local pond and standardized-bared highway surface, a vector database for compartments was established automatically. It . contains 3 data sets for vegetation indices （VI） namely DVI, NDVI and the Greenness of Tasseled Cap, the climatic statistics in July-August of 2000, 2001 and 2002, and topographic factors respectively, so comparisons among the VI＇s of the woodland divisions and analysis on them to the standing conditions were approached. With a very slight increase （1.11%） of the NDVI as an exception in 2001 to that of the 2000, all the VI＇s in the year of 2001 and 2002 dropped largely ; Annual changes of the three indices for the deciduous are the biggest while conifers or conifer-deciduous mixtures are the second with the NDVI changes amounted only 1/3 tol/7 to the former ones, and the springwood or the afforestable land are the smallest. The conclusion is that the sensitivity and adaptabilities of three main forest divisions to the ecological factors are thus ranked, and great changes in magnitude of VIs may resulted from extremes of ecological factors. Further in-situ investigations of compartments and ecological modeling for RS monitoring ecosystem of this Nature-reserve are needed.