The Holy City of Mashad, a changing environment

Remote sensing change detection techniques have shown their abilities to monitor and detect type, magnitudes and geographical location of changes over a certain period of time. Sometimes, the changes that reported from digital image processing techniques are not the actual changes. In this research work, an attempted has been made to study the urban land use/land cover changes over a period of ten years from 1987 to 1998 for the holy city of Mashad, Iran. The PCI's Geomatica image processing software was used to register two Landsat TM images of the study area. Using maximum likelihood classifier both images were classified and a change map was produced. Statistics from change map showed an innumerous change to this urban area, which were not the actual changes. Due to this, through a procedural image processing techniques and based on fuzzy logic an image was created to show magnitude of the changes. The by masking areas of less changes over the change map, a new image created. This image holds most reliable changes over the study area and after extensive ground truth investigations using a handheld GPS, its statistics was offered to the city planners. It is recommended here not to use the widely used classification techniques such as MLC for urban landuse/cover classification. The SAM technique is recommended in this case.