Floodplain Mapping Using HEC-RAS and ArcGIS: A Case Study of Kabul River

Abstract

This paper describes the application of HEC-RAS model to the development of floodplain maps for the part of Kabul river that lies in Pakistan. The intent is to assist policy makers and planners in the development of flood mitigation measures for the Khyber Pakhtunkhwa Province, which experienced unprecedented floods in July/August 2010 exposing the vulnerability of the province to this natural catastrophe. Owing to its reasonable accuracy and free availability, shuttle radar topography mission digital elevation model was chosen for the extraction of geometrical data for the river. Conventional flood frequency analysis, involving log-normal, Gumbel’s, and log-Pearson type III (LP3) distributions, was used to calculate extreme flows with different return periods. Using Kolmogorov–Smirnov (KS) test, LP3 was found to be the best distribution for the Kabul River. The peak floods from frequency analysis were input into HEC-RAS model to find the corresponding flood levels expected along river reaches extending through Warsak dam to Attock. Results obtained with HEC-RAS model were used in combination with ArcGIS to prepare floodplain maps for different return periods. Through floodplain maps, areas that are vulnerable to flooding hazards have been identified. Analysis of floodplain maps indicated that more than 400 % area is likely to be inundated as compared to the normal flow of the river. Most of the area found to be vulnerable to flooding is currently used for agriculture. Comparison of simulation of 2010 flood with the image of the flood taken by MODIS clearly shows a close agreement between the two.