Geospatial techniques for flood inundation mapping

摘要

In India, most of the rivers form big size natural islands due to change in its course. However, identification of suitable river island for construction of Eco-friendly parks/tourist destination is a very challenging task since these are exposed to river flooding. River islands which are least vulnerable to the impact of severe flooding can be a suitable place for construction of tourism destination such as eco-friendly Parks, Hotels etc. The study involves a two step approach viz. automatic extraction of river islands and model development for flood inundation mapping for extraction of eco-friendly tourism destinations. In this study, automatic extraction of the river islands has been carried out using texture based classification approach. The satellite data acquired by the Indian Remote Sensing Satellites sensors such as LISS-III and Cartosat-1 DEM have been used for analyses. In the first step, satellite imagery has been broadly categorized into 6 landuse/cover classes viz. Water, Sand, Islands, Settlements, Agriculture and Cropland. Extraction of such islands which remain unaffected during severe flooding has been accomplished with the flood inundation mapping which has been carried out in HEC-GeoRas with in GIS environment. The model utilizes the primary 4 inputs viz. geometry of the river (DEM, slope), time series data of water surface elevation, landuse/cover, and location of rain gauge station for flood inundation mapping. This paper also investigates the applicability of the eco-island concept to include protection of wetland, management of land-resources, sustainable use of natural resources and construction of ecological park/hotels. The output of the study will be very useful for Government authorities in stabilizing economy, and enhancing the tourism infrastructure in a better way.