ESTIMATING VEGETATION COVERAGE CHANGE USING REMOTE SENSING DATA IN HAIDIAN DISTRICT, BEIJING

Vegetation coverage, the ratio of vegetation occupying a unit area, is a very important parameter in the development of climate and ecological models.On-ground fieldwork surveys of vegetation coverage are time consuming and expensive and produce low-precision results.Estimation of vegetation coverage from remote sensing data may be a more efficient approach.This study explores the potential of deriving vegetation coverage from normalized difference vegetation index (NDVI) using remotely sensed data.As each pixel of a TM image represents a mosaic of structures on the ground, sub-pixel models for vegetation coverage estimation are proposed.The utility of different sub-pixel models for vegetation coverage estimation based on land cover classification is also described.The accuracy of vegetation fraction estimation using this model is high.Vegetation coverage of Haidian district in 1975, 1991 and 1997 were modelled using this method in Beijing.The spatio-temporal patterns of change in vegetation coverage and its driving processes during this period are also discussed in this paper