Flood risk assessment for a medium size city using geospatial techniques with integrated flood models

Abstract

Flood is defined as the high stage of a river that flows across the river banks. Floods are classified based upon the nature of the occurrence, usually as riverine floods, urban floods, lake fluctuations, and rise of sea level. Factors responsible for floods are topography, rainfall intensity, climate change, and improper design of drainage facilities. In this study, we discuss urban flooding, because urban flooding usually differs from other types of flooding where the increase in flood peaks of 1.8 to 8 fold results in flood volume up to 6 fold. An urban peak flood occurs quickly as compared to other types of floods. In urban region the factors responsible for high peak discharge are land use/land cover (LULC) changes, climate change, and increase in rainfall intensity within a given duration. The recent floods in major cities in India have raised the awareness for flood modeling studies. In India, most flood modeling studies are done for major urban river basins passing through cities, but for this study the Koraiyar River basin that passes through the medium-sized city of Tiruchirappalli in South India was chosen. The parameter adopted for analyzing flood volume in this basin is LULC and its impact on urban surface runoff. LULC changes and their impact on surface runoff are studied by integrating remote sensing and geographic information systems (GIS) with hydrologic HEC-HMS (Hydrologic Engineering Centre-Hydrologic Modeling System) and hydraulic HEC-RAS (Hydrologic Engineering Centre-River Analysis System). Runoff is generated by using LULC, slope, hydrologic soil group, curve number (CN) maps, and rainfall intensity. The overall performance of the hydrologic model during calibration was satisfactory, based on the Nash–Sutcliffe Efficiency criteria of with values of 0.5–0.6. The generated peak discharge is used for developing floodplain and hazard maps from 1986 to 2036. The floodplain and hazard maps estimate flood depth and risk in the basin area for changing LULC conditions of different return periods. Flood plain and hazard maps of 2, 5, 10, 50, and 100 years return periods were generated to identify flood extent and hazard level in the basin. The developed maps are used as a tool for effective flood forecasting and warning of flood hazards in the basin.