Runoff computation of Halda river basin using soil moisture accounting method

Abstract

In this study basin runoff was simulated using the computer model of the Hydrologic Engineering Centre-Hydrologic Modeling System, called HEC-HMS for the upper part of the Halda river basin given rainfall and catchment parameters. The graphical user interface of the HEC-HMS was used to edit, execute and view model data. Data needed for this study included soil, land use, digital elevation model, point elevation, contour line, stream network, rainfall, evaporation, and river discharge. Detailed stream network was created using topographic maps of 1:50000 scale. Required maps and data were collected from Water Resources Planning Organisation (WARPO), Institute of Water Modeling, and Soil Resources Development Institute. The HEC Data Storage System (DSS) was used for storage and retrieval of time series data. Geographic Information System (GIS) tools e.g. Arc/Info 8.02 and Spatial Analyst of ArcView 3.2 were used for delineating watershed and stream network from Digital Elevation Model (DEM). The chosen basin area was divided into sub-basins and the loss rate parameters such as interception, infiltration and base flow were estimated. Most of the parameters i.e. canopy interception and surface storage, soil profile and tension zone depth of Soil Moisture Accounting (SMA) model were calibrated during simulation. No groundwater zone was considered in this study region because percolation rate became zero as model suggested. Flow directions, flow accumulation points, flow path lengths, sub-basin areas were determined from DEM. Hydrologic modeling parameters (i.e. length of longest flow path, SCS curve number, lag time, stream velocity, Muskingum X, Muskingum K etc.) and appropriate rainfall station weightage were estimated using ArcView GIS as input parameter of the model. The rainfall excess was transformed into direct runoff using Soil Conservation Service (SCS) unit hydro graph method. Runoffs from subbasins were combined at the confluences and hydro graphs routed through a channel network using Muskingum channel routing. The simulated daily runoff for the entire year was compared with the observed runoff and that computed using another lumped conceptual model called NAM model of the Danish Hydraulic Institute. The peak flow of simulated runoff was 426.08 m3/s on 24th June and observed runoff was 202 m3/s on 24th June. NAM computed peak flow was 405.61 m3/s on 25th June. Thus HEC-HMS overestimated the peak runoff by 52.3% as compared to observed peak runoff and by 4.8% as compared to peak runoff by NAM model. HEC-HMS and NAM results agreed reasonably well.