object based

By evaluating land use change and land cover, it is possible to plan and land manage to reduce the impact of alteration and degradation of exploitations. The purpose of this study was to evaluate land use changes in Ardabil province between 1987 and 2015 using Thematic Mapper (TM) sensor of Landsat 5 and Operational Land Imager (OLI) sensor of Landsat 8 satellite images. Firstly, the efficiency of the object-based classification method was evaluated using ground control points, error matrix, and comparing total accuracy and kappa coefficients for the resulting maps. Total accuracy and Kappa coefficients for 1987 were 82.3% and 0.70, and 2015 were 94% and 0.90, respectively. Change detection results showed that from 1987 to 2015, rangelands decreased by 4.2%, equivalent to 75039.3 hectares and forests decreased by 0.5%, equivalent to 9393.7 hectares compared to the total province area and irrigation farming increased by 3.1%, equivalent to 55276.8 hectares, residential areas increased by 0.8%, equivalent to 14989.7 hectares and dry farming increased by 0.7%, equivalent to 12632.4 hectares compared to the total province area. The most significant changes that occurred from 1987 to 2015 were the conversion of rangelands to dry farming with 13.8%, equivalent to 141794.2 hectares and irrigation farming lands with 4.8%, equivalent to 49836.9 hectares of rangelands, irrigation farming to dry farming with 7.7%, equivalent to 11464.8 hectares and forests to rangelands with 6.1%, equivalent to 1059.9 hectares which indicate that the requirement of a serious revision of the management policies.