During the recent years, the city of Mashhad has been growing relatively fast, resulting in significant land use changes. Thousands of hectares of agricultural lands, gardens and pastures have been destroyed and have changed into other uses such as cities, industrial districts, and so on. Following this fact, the streets have been widened and new parks have been created.
The discovery and rehabilitation of urban land use change is one of the most important issues of city planning and management. There are different ways to discover and retrieve changes and to use remote sensing data. In this research, subtraction and division of images, regression, principal component analysis and subtraction bands, as well as utilization of fuzzy logic, comparison of classification and clear probabilities obtained from the maximum probability method have been considered.
To understand and discover the changes, the city of Mashhad has been examined using Landsat TM satellite images of the years 1987 and 1996.
In this research, using posterior probabilities obtained by the maximum probability classification, the type and amount of membership probability of each pixel in change was determined. Then the classes in 1987 and 1996 were subtracted. The results of this study show that:
A. Multi-time satellite images show land use changes in the studied area very well.
B) Using two methods of principal components analysis and the fuzzy logic simultaneously, the place and the degree of membership of the pixels in the changes are recognizable.
C) In the images obtained from reduction of posterior probabilities related to the corresponding classes, the changes are represented by positive or negative numbers and are clearly distinguishable from the areas with no changes.
Therefore, the method based on use of posterior probabilities derived from classification of maximum probability is the most appropriate method, because it can determine the type and nature of the changes and the possibility of membership in changes.