Detection of change in vegetation in the surrounding Desert areas of Northwest China and Mongolia with multi-temporal satellite images

Vegetation monitoring is an important step in developing a better understanding of land use and its changes, due to the sensitivity of surface vegetation to changes in the global climate and environment. In this study, normalized difference vegetation index (NDVI) of the area surrounding the Gobi Desert in North Asia was multi-temporally interpreted by analyzing time-series Satellite Pour l'Observation de la Terre (SPOT) Vegetation (VGT) data, over a roughly nine-year period from January 1999 to November 2007. The study area was classified into eight classes, and compared to classified Moderate resolution Imaging Spectrometer (MODIS) global land-cover data to select desertification-sensitive areas. The study focused on three classes (barren land, open shrubland, grassland) due to their high sensitivity to climate change. The results showed significant extension of the barren land class from 1992 to 1999, with 47.8% of the open shrubland transformed into barren land. Among five terms (1999-2003, 2003-2005, 2005-2007, 1999-2005, 1999-2007) which are carefully selected from variations of the annual NDVI mean for each class over nine years, significant changes were observed for barren land from 1999-2003, and for open shrubland and grassland from 2005-2007. An analysis of the positive change (the change from sparse vegetation to dense vegetation) and negative change (or desertification) was conducted over the study period; the number of pixels corresponding to a positive change for barren land was similar to the number of negative change pixels. Human activity and afforestation over the study area were also captured in multitemporal satellite imagery. For open shrubland and grassland, the negative change area was bigger than the positive change area. Precipitation data over the nine-year period exhibited a pattern similar to that for the vegetation data, as expected.