فهرست مطالب aliasghar ardashirpei

The purpose of this research is to evaluate the land use of Ghouchan city by using object-oriented and pixel-based classification as well as predicting these changes using the CA Markov model until 2030. In this research, Landsat satellite images of ETM and OLI sensors for the years 2000 and 2018 (August) were used. After the images were taken, radiometric corrections were applied to the images, and then using the ground-object and object-oriented pixel methods, a land use map was extracted. In order to evaluate the classification accuracy, general accuracy and Kappa coefficients were used. The results obtained in the object-oriented classification in both of the general accuracy and kappa coefficients were 94% and 97%, respectively, which is more accurate than the pixel-based method. Most of the area in the region, using Object Oriented Classification in 2000, is related to land use and mountainous land use, and to base land use and land use, respectively. According to the Classification for 2018 using Object Oriented Classification, most of the area had the weakest land use and dry land use. Using the CA Markov modeling and considering the two land use maps, the probability matrix was calculated, and the CA mapping prediction map for the next 12 years, 2030, was obtained, and the area and percentage of each Uses were calculated separately. The results showed that the greatest increase in the variation among the users would be for poor pasture users in 2030, which is also increased by 2019-27449 hectarrs. The largest reduction in area will be for dense pasture users with a total area of 236666 hectares. Man-made human consumption will grow at 62.530 hectares during this 12-year period. By predicting user variations, the extent to which resources can be expanded or degraded can be guided, and this can be done by directing these changes to appropriate paths.