The research attempts to demonstrate the potential of remote sensing in base mapping and change detection in urban land-use in Delhi, with the aim of understanding the urban hydrology. The study uses IRS LISS III sensor and ERDAS Imagine image processing system to establish the potential of remote sensing techniques in obtaining synoptic and repetitive coverage of the city to monitor the patterns of urban growth and urban fringe activity.
A good level one map, in terms of accuracy of classification is obtained from digital image processing of the satellite data of the urban area. According to the study, a supervised classification map is better than the unsupervised one for the urban sprawl class. However, such a map with a good accuracy cannot be generated in a single iteration. The class designation change can be prompted in the accuracy checking exercise.
The most useful sensor is IRS - LISS III due to the presence of short, wave infra red channel and owing to its better spatial resolution. In the final classification using LISS I and LISS III data, the Kappa accuracy is 55 % and the overall accuracy is 74 %.
The remote sensing techniques used could meet the spatial, temporal and multi-spectral database requirements of urban areas. Runoff estimation for studying the impact of urbanization can be done successfully employing SCS Curve Number (CN) technique by using average CN from the land-use map thus generated.
Download the report here: