%0 Journal Article %T Investigation of Land Use Changes in Karkheh Watershed during 1990 and 2020 Using Google Earth Engine Platform and Landsat Satellite Images %J Iranian Journal of Soil and Water Research %I University of Tehran %Z 2008-479X %A Sadian, Ali %A Shafizadeh-Moghadam, Hossein %D 2021 %\ 12/22/2021 %V 52 %N 10 %P 2569-2580 %! Investigation of Land Use Changes in Karkheh Watershed during 1990 and 2020 Using Google Earth Engine Platform and Landsat Satellite Images %K Karkheh watershed %K Land use %K Google Earth Engine %K Random forest %R 10.22059/ijswr.2021.330075.669068 %X Karkheh is one of the most important watersheds for water resources management and croplands in Iran, where the largest dam in Iran and the Middle East is located there. Karkheh is considered as Iran’s food basket and exploring land-use changes in this watershed has highly strategic. In the present study, land-use changes during 1990 and 2020 in the Karkheh basin have been extracted and evaluated using Landsat satellite images and random forest algorithm in the Google Earth Engine platform. In this paper, the changes of 11 classes, including forest, shrubland, grassland, irrigated, rainfed, garden, barren, water body, wetland, urban, and riparian have been quantified. The largest area of the region was belong to grassland and barren. In this research, the classification process has been done separately for each Landsat image scene in the Karkheh basin, and finally, all the scenes have been mosaic together. Using this method, most of the images in a scene are used, and the time series of indexes specific to each class of each scene is used for classification, which achieves more accurate results than the method of classifying the whole area in one place. The results show urban areas have increased by 113%, water bodies by 149%, garden by 163%, riparian by 39%, irrigated by 122%, wetland by 10% and rainfed by 34%. However, forest 22%, barren 20%, and shrubs 20% were reduced. As a result, this statistic indicates an expansion of agriculture and reduction of grassland. The accuracy assessment of the classified images confirmed the overall accuracy and kappa coefficient as being 96% and 95% for 1990, 94%, and 93% for 2020. These indices show the appropriate accuracy of classification maps and the validity of the results. %U https://ijswr.ut.ac.ir/article_85537_7e0eac166ff37d86d596dc2ab3d724d2.pdf