Rainfall-Runoff Modeling for Jamuneswari Watershed Using SWAT Software

Abstract

For a flood prone country like Bangladesh, early warning system for flood is very important to aware public and protects their property from significant damage. This requires accurate development of watershed model and proper prediction of the water level and discharge for the desired river. In this study, a rainfall-runoff model was developed for the Jamuneswari River catchment in the North-West part of Bangladesh using Soil and Water Assessment Tool (Arc SWAT) software. Arc SWAT incorporates GIS information with attribute database is used to estimate the discharge at catchment outlet. The developed model was calibrated and validated with soil and land-use dataset, topography and global weather dataset at Badarganj station which is the outlet of Jamuneswari River catchment. The model calibration and validation task was undertaken using SWAT CUP engine for the period of 2007 to 2013 and 2014 to 2017 respectively. The statistical criteria’s such as Co-efficient of Determination (R2) and Nash Sutcliffe Efficiency (NSE) was found within acceptable limit by comparing observed and simulated discharge. The developed model exhibits satisfactory performance for rainfall-runoff modelling for the Jamuneswari River catchment. The model can be used to predict the future flood discharge so that property damage and life losses can be minimized effectively.