Investigating the Effects of Land Use Changes on Trend of Desertification Using Remote Sensing (Case Study: Abarkooh Plain, Yazd, Iran)

Abstract

More than one-third of the earth is characterized by arid and semi-arid climate and desertification phenomenon in these areas has been intensified in recent decades. This study aims to investigate the trend of desertification using vegetation indices and Iranian Model of Desertification Potential Assessment (IMDPA) in Abarkooh Plain, Yazd province, Iran. The satellite images (Landsat images in May 1987, 2000 and 2013) were used to map the land uses in the region. After the image preprocessing with a supervised classification method (support vector machine) that had the highest accuracy in all years, the map of land use changes was prepared in five kinds of land uses such as residential and industrial lands, agriculture and horticulture lands, arid and desert lands, playa and saline soils, and mountainsT .hen, the desertification map of Abarkooh Plain was prepared by calculating the Normalized Difference Vegetation Index (NDVI( and Modified Soil Adjusted Vegetation Index )MSAVI( according to IMDPA model. Three criteria, urban and residual lands to agricultural land ratio (Ur/Ag), plant canopy density and vegetation renewal indicators were used in IMDPA model. On the basis of the Ur/Ag ratio, 5.74%, 6.74% and 87.52% of the region were characterized by severe, moderate and low desertification, respectively due to land-use change factors. In the viewpoint of aboveground canopy density indicator, desertification in the area was classified into four classes including very severe, severe, moderate, and low desertification. In this regard, 99.06%, 0.76%, 0.13% and 0.05% of the areas are put in very severe, severe, moderate and low desertification classes, respectively. According to vegetation renewal indicator, 92.51%, 3.21% and 4.28% of the area were characterized as unchanged, low and very severe desertification, respectively.