In order to manage mapping errors propagated through the analyses of change, all the digital datasets were resampled to 56 m before proceeding for analysis. For change analysis, raster maps of each year were used. Change matrices were generated for the study periods. This was done by comparing the number of pixels falling into each category of land use at one time period and the category of the same pixels in the previous time period determining the change class. Matrix model available in ERDAS Imagine was used for this purpose.