Monitoring of soil salinity using ASTER satellite data to evaluate desertification in the Marvast plain, Yazd province

Desertification relates to the both the process and end state of drylands degradation. Salinization and alkalinization are two indicators of soil degradation in arid and semi-arid regions. The main objectives of this research is monitoring of soil salinity using high spectral and spatial resolution of remote sensing to assess desertification in the Marvast plain، Yazd province. Two images of Terra satellite، ASTER synchronous to 2003 and 2010 are used. After preprocessing and analyzing of the images، relationship between parameters of soil salinity (i. e. SAR and EC) and spectral reflections were determined and، both two satellite images were classified using maximum likelihood method. Then، the surface area of each class and the amount of its changes were calculated. Results showed that during the period of 7 years (2003-2010)، area of non-saline lands has decreased while، the area of saline land has increased، which leads to the salinization of agricultural lands، reduction of its yield and also extent of desertification in this region. Accuracy of EC map classification for 2003 and 2010 images are 87. 5% and، 82. 5%، respectively. Kappa coefficients for both images are 0. 83 and 0. 76. Accuracy of SAR map classification for 2003 and 2010images are 87. 5% and 87. 5%، respectively. Kappa coefficients for these two images are 0. 81 and 0. 77، respectively. Generally، it can be conclude that using of remote sensing data، especially ASTER images has high efficiency for change detection analysis in soil salinity and natural resources management.