An Investigation of the Impervious Surface Changes in Esfahan City by Using Remote Sensing

Change detection is the basic assessment requirement in natural resources management. Maps of land use/land cover changes can be made based on multi-temporal remote sensing imagery. Several methods have been proposed to detect land changes. The subject of this study is to investigate changes in impervious surfaces within the Esfahan city by using MSS and AWiFS images and the comparison of post-classification method during the 1987-2008 period. Therefore, the images were corrected geometrically by using the geometric algorithms. RMSe of MSS and AWiFS images were obtained from 0.58 and 0.56 pixels, respectively, indicating that the spatial error was less than one pixel. Then, two images were provided with 5 classes including water (river), residential area, bare land, plant cover and roads using the training samples and multilayer perceptron neural network. The overall accuracy of MSS and AWiFS images were obtained to be 90.79 and 92.86, respectively. In the next phase, for change detection, post classification comparison method was performed and the changes map was produced. The results show that impervious surface area has increased 6829 ha and the vegetation area has decreased over 527 ha during this period. The principal reasons for this change are rapid population growth rate of 62% and industrial growth during this period.