Engineering and Applied Science Research (Mar 2021)
Application of a rainfall-runoff model for flood generation in the Huai Sangka catchment, Thailand
Abstract
Estimation of the rainfall-runoff relationship is essential for solving and managing water resources. This paper presents the use of an event-based rainfall-runoff model built with HEC-HMS version 4.3 to simulate runoff in the Huai Sangka catchment (193 km2), a small agricultural watershed. The rainfall-runoff model was conducted by a semi-distributed modelling approach, where parameter values of sub-basins were estimated based on spatial and temporal data such as topography, land use, soil and climatic variables. The modelling approach includes Soil Conservation Service (SCS) Curve Number, SCS unit hydrograph and Muskingum routing methods for simulating losses, runoff transformation, and routing in the rainfall-runoff system, respectively. The rainfall-runoff model was used to simulate runoff with corresponding to four extreme rainstorm events (in September 2007, September 2008, August 2011 and July 2017). The present study aimed at testing the applicability of the modelling approach to ungauged basins. Thus, only travel time parameter K of the Muskingum routing method was adjusted to achieve reasonable reproductions of major flood hydrographs. The other parameters, which were initially estimated from the spatial data, were remained constant. On the basis of event-based runoff simulation results, the largest percentage error in peak was 12.96%, considered satisfactory. The comparison of the observed and simulated hydrographs confirmed the reliable performances of the rainfall-runoff model with NSE values, ranging between 0.61 and 0.74 and with RSR values, ranging between 0.51 and 0.57 for the selected rainstorm events. Thus, the model was considered suitable for flood simulations in the study catchment. Furthermore, it was found that the travel time parameter K is strongly related to physical characteristics of channels. In summary, the modelling approach is applicable to ungauged watersheds since model parameters can be estimated on the basis of physical characteristics.
