Geoiformatics based spatial decision support system for flood disaster management in urban and periurban areas

Abstract

Effective management of flood inundation is imperative to avoid loss of life, properties and infrastructures. The present study addresses this issue by considering the study area between Ukai reservoir and Surat city that experienced normal and unprecedented floods over the years. A suitable modelling scheme, that takes into account the local terrain topography, soil, channel flow, land use/cover, rainfall etc.was adapted and combined with input generated using Geoinformatics techniques to develop a Spatial Decision Support System (SDSS). Previous studies on flood modelling and management for this region lacked such an integrated approach. In this study, several models were considered and evaluated and the coupled HEC-HMS and HEC-RAS model was selected based on the evaluation score. Hydrological model HEC-HMS is used for the estimation of rainfall based runoff for the various junctions of the Tapi river within the study area. These runoff contributions were used in the hydrodynamic model as lateral inflow and then unsteady river flow simulations were done to study the water surface profile and inundation pattern along the bank of the major river Tapi, flowing through a stretch of peri-urban and thickly populated urban areas in the large industrial city of Surat, and several mid-sized and small towns and villages. Steady flow simulations for same area and for the different return period flow ranging from 2 to 500 years were also done to study the water surface profile and inundation pattern. A Spatial Decision Support System based on the model simulations and for the management of anticipated flood induced damages in the city of Surat and surrounding region is developed. Calibration results of HEC-HMS indicate a high value of Nash Sutcliff coefficient of 0.85. Percentage of simulated runoff from the rainfall varies from 42% (2002) to 57% (2006) considering low to high rainfall years. Similarly, integrated HEC-RAS model was also calibrated with Nash Sutcliff coefficient value of 0.86. Water inundation pattern simul