River flow analyses for flood projection in the Kabul River Basin

Abstract

Flooding is one of the critical natural disasters in Afghanistan, causing huge social and economic losses on an annual basis. Due to lack of historical data and long gaps in the recorded data, flood predictions are usually associated with large uncertainties. The available hydrological data are collected before and after the Afghan civil war period. This long gap and climate change effects split the dataset and faces a challenge of using either dataset alone for predicting flood characteristics. In this study, first, the two datasets are compared to find river flow variation in terms of peak and frequency. Next, the river flow variation effects on flood peaks for each return period are analyzed to determine the flood projection. The results show that flood peaks have raised while the mean discharge in the basin is reduced during the second period. The frequency analyses show a change in high and low flow days in the recent period. In addition, the flood recurrence results show that the utilization of single period data for return period flood predictions yield huge variation, while the analyses using the combined dataset show a reasonable estimation of flood characteristics. Furthermore, the comparison of calculated flood peaks based on the first period and combined dataset show that flood peaks have an upward trend.