Extraction and Analysis of Natural Vegetation Cover and Soil Moisture Using Landsat TM Data in the Arid Oasis of Xinjiang, China

Abstract

The natural vegetation cover and soil moisture have significant impacts on the stability of the Arid oasis. In this paper, the 75 Landsat TM images were processed to extract the natural vegetation cover, obtained by the Normalized Difference Vegetation Index (NDVI), and soil moisture observed by Tasseled Cap Wetness (TCW) during the year 1990–2011 in the Sangong River Basin, Xinjiang, China. Four plots of natural vegetation were selected to measure the average values of NDVI and TCW. The results reflected that the NDVI and TCW kept stable during the past 20 years with a strong correlation between each other. The influences of different geographical locations and different cover types on the change of NDVI mean in four plots were different. There was a close relationship between NDVI (or TCW) and precipitation in a short time. On the other hand, the data of groundwater depth measured from 16 groundwater observation wells year from 1998 to 2002 were used to perform spatial interpolation in the study area. Two shrub land plots in northern oasis were selected to extract the values of groundwater depth, vegetation index and soil moisture at each grid pixel (30 m) for carrying out the correlation analysis. The results demonstrated that there were obvious correlations among them when the groundwater depth varied in a range of 2.5–4 m; otherwise, the growth of vegetation would be inhibited in this area. The remote sensing as an effective way of achieving time series information could provide important data for natural vegetation development and evolution, which could be in favor of oasis’ stability and sustainable development.