Assessment of land use changes into dry land using satellite images and Geographical information system (GIS)

The aim of this stydy is to proced the trend process of land cover changes as a result of land use changes into dry land using satellite images, remote sensing and GIS softwares and to find out the type and magnitude of changes by the image sensor Landsat ETM in 2000 and sensor of Landsat OLI in 2014 using ENVI®5 software and unsupervised classification methods in Dehsard and kohsefeid, Kerman province. Finally, land use classification map including rangelands, agricultural lands and gardens was specified. In order to produce the training points, field visits, GPS and Google Earth software were utilized. Supervised classification was used to assess the accuracy of classification images. Afterwards, KAPA coefficie was applied to calculate the precision of produced maps. Precision matrix was created for each map. For the detection and assessment of land use changes as compared to the others, Crosstab was used. The results of land use changes in two regions showed that in this regard, the rangelands had the most changes estimated as 77% and 73% for Dehsard and Kohsefid watersheds, respectively. The precision of classification maps was given as 98% for both watersheds. The results of this study showed that the expansion of agricultural activities concerning the rangeland ecosystems could change the rangelands into the lands with low efficiency and in two mentioned regions, 9% and 20% of changes occurred in the low efficiency lands in Dehsard nd Kohsefid watersheds, respectively.