Integrating GIS and Hydrological Model for Flood Forecasting

Abstract

Floods cause significant loss to life and property by way of civilian casualties, more people are killed annually than by any other natural disaster, annual economic losses consist of property worth crores of Rupees damaged or lost. Natural disasters caused by floods challenge scientists to forecast the magnitude and timing of peak flood discharges more accurately. Hydrological models combined with GIS have become important tools for flood forecasting. In the present study, SCS-CN model has been used to estimate daily direct runoff and Huff’s disintegration model has been applied to distribute the daily direct runoff into hourly rainfall excess, which has been convoluted with the synthetic unit hydrograph derived by SCS method to generate hourly discharge at specified location. The model comprises of three components, first development of rainfall-runoff relationship, second computation of 1-hr Synthetic Unit Hydrograph (SUH) and third, convolution of the rainfall excess with SUH. The Uri river watershed of lower Narmada river basin has been selected for the study. The average daily rainfall and curve number has been used to estimate daily direct runoff volumes, which are convoluted with Unit Hydrograph to get hourly discharge for some selected events during the monsoon months in the year 2005- 2007. The model results have been compared with the observed hourly flows and found very good correlation between them. In order to predict floods in the river basin for next 24 hours, prediction of 24 hr. expected rainfall is required. By continuous correction in predicted rainfall is carried out by replacing the actual rainfall to obtain predicted flood hydrograph.