基于NDVI像元二分模型的矿区植被覆盖动态监测

Vegetation dynamic monitoring in mining area based on NDVI serial images and dimidiate pixel model

运用泗顶铅锌矿区TM/ETM1991年、1999年、2005年和2010年时间序列影像提取NDVI值,以像元二分模型由NDVI值反演出各时相植被覆盖度,在E NVI软件下进行植被覆盖度分级并计算各级覆盖度转移概率矩阵,进而监测矿区地表覆被变化过程。结果表明:2 0年间矿区植被覆盖度整体处于较高覆被,且呈现先快速后缓慢增加的趋势,其中:大部分中低覆被类型向较高覆被类型转移,较高覆被类型向高覆被类型转移。该方法可以快速、定量地反映矿区植被覆盖及变化情况,为矿区生态环境评价和治理提供技术支持。

In order to monitoring the dynamic changing course of vegetation coverage in Siding Zinc-lead Mine, TM＆ETM time serials images from the years 1991, 1999, 2005 and 2010 were used as information source. How to preprocess the satellite image, extract the NDVI data, inverse the vegetation coverage by using dimidiate pixel model and analysis vegetation coverage state transition matrix were studied, furthermore, the changing process of the mine area＇s vegetation coverage was monitored. The results show that the vegetation coverage in overall Mining area was at a higher level in past 20 years. The vegetation coverage variation tendency was increasing rapidly during 1991 to 1999 but slowly from 2005 to 2010, and transferring gradually from lower level to higher level. This method can be used for ecological environment evaluation and management in lager mining areas because it can monitor and analyze on vegetation coverage and its changes rapidly and quantitatively.