Ali Seif; Hoda Ghasemieh; Hossein Zeinivand; Mehran Zand
Abstract
Today, with the increase in population, land use has been changed to meet the needs, which has made it more important to know the future status of land use. Investigating land use changes plays a fundamental role in environmental studies, modeling and simulation of land use changes and water resources ...
Read More
Today, with the increase in population, land use has been changed to meet the needs, which has made it more important to know the future status of land use. Investigating land use changes plays a fundamental role in environmental studies, modeling and simulation of land use changes and water resources management and helps managers in better land use planning. So, in this research, first, the land use map of Rahim-Abad Basin was extracted for 1999 and 2016, using ETM+ and Landsat 8 sensors in ENVI5.3 software environment, with a kappa coefficient of 95% and overall accuracy of 97% that indicates the high accuracy of the 2016 map.Then, the maps of affecting factors on land use change including maps of distance from the road, distance from the river, distance from the city, distance from the village, distance from the fault, geology, soil texture, rainfall, evaporation, elevation digital model, gradient, groundwater level and the amount of solar radiation were prepared in Arc-GIS 10.6.After, using logistic regression, the role of effective factors on land use was determined and the Relative Operating Characteristics curve (ROC) was used to evaluate the logistic regression. Finally, the land use map of the study basin was simulated for 2026 using the CLUE-s model.Results showed that the area under the ROC curve was 0.9, 0.88, 0.9, 0.92 and 0.91 for grasslands, rain-fed lands, irrigated lands, water and, residential zones, respectively, which expresses the acceptable accuracy of the regression method in investigation of affecting factors on land use.Also, the most changes of land use in 2026 would be related to conversion of rangelands to rainfed lands, and 6.47% rangelands would decrease and 18% rainfed lands would increase.
