Rainfall-Runoff Modeling Considering Soil Moisture Accounting Algorithm, Case Study: Wabe Catchment, Omo Gibe Basin, Ethiopia.

Abstract

Hydrologic studies on rainfall-runoff have been extensively applied by water resource planners to simulate the hydrological response in many regions around the world to fulfill various desirable needs with a purpose of effective and proper planning and managing of water resources for present and future uses. Therefore, the main objective of this study was to use a continuous soil moisture accounting (SMA) algorithm in a Hydrologic Engineering Center-Hydrologic Modeling System (HEC-HMS) to model and predict stream flow in the Wabe catchment, Omo-Gibe River Basin in Ethiopia. Long term daily rainfall data from 4 rain gauging stations from1985 to 2016 years, daily river flow of 1 stream gauging station from 1987 to 2007 years, land use and soil data of the watershed, and DEM were obtained from relevant sources. These data were then analyzed and interpreted, and used to set up the HEC-HMS model. In this study, soil moisture accounting loss method, Clark unit hydrograph transformation method, linear reservoir base flow method and Muskingum routing method were adopted. In order to fix the horologic parameters of each watershed, first the sensitivity analysis was carried out with the base data, and then the model calibration was performed using data from 1987 to 1999 and validation for the period from 2000 to 2007 at a daily time step. The sensitivity analysis of different model parameters were ranked according to their sensitivity in terms of percent change in simulated runoff volume, peaks and Nash Efficiency. Sensitivity analysis helped to understand the behavior of the model and relationships between the key model parameters and the variables. The model performance was evaluated based on computed statistical parameters and visual checking of plotted hydrographs. For the calibration period, the performance of a continuous model ranges from good to very good with a coefficient of determination R2 = 0.727, Nash-Sutcliffe Efficiency NSE = 0.711, percentage error in volume PEV=2.36%, percentage error in peak PEP = 6.25%, Percent Bias (PBIAS) = 2.35 and Root Mean Squared Error (RMSE)/observations’ standard deviation ratio—RSR = 0.5. Similarly, the continuous model performance for the validation period ranges from good to very good with R2 = 0.861, NSE = 0.807, PEV = 0.42%, PEP = -2.91%, PBIAS = -0.42 and RSR = 0.4. The performance results obtained showed that, the SMA model in the HEC-HMS was found to give a good prediction of stream flow in the Wabe Catchment. Finally, the peak flood results of the HEC-HMS model were compared to the statistical distribution models results of two methods those selected based on their ranks of goodness of fit. The forecasted peak flood by HEC-HMS, General Pareto distribution (GDP) and General Extreme Value (GEV) distribution methods, at 2, 5, 10, 25, 50, 100, 200 and 500 year return periods were 479, 644.7, 755.5, 896.8, 1003, 1110.2, 1219.3 and 1367.5; 407.13, 618.23, 722.17, 812.34, 856.74, 887.52, 908.86 and 927.38; and 542.89, 711.47, 817.22, 944.48, 1034.53, 1120.39, 1202.59 and 1306.14 m3 /s respectively. Then, these predicted peak flood will help in water resources and flood management for this study area.