Selection of appropriate loss methods in HEC-HMS model and determination of the derived values of the sensitive parameters for un-gauged catchments in Northern Ethiopia

The estimation of peak flood and runoff is important in planning water development and flood mitigation/adaptation works. This can be simplified by using calibrated hydrological models, yet in many developing country situations there is a paucity of primary flow data. Against this background, the current study was undertaken in an un-gauged catchment in Northern Ethiopia. The study was aimed at deriving values of the sensitive parameters of the HEC-HMS model that could be transferred to similar geographical settings. To calibrate the model, observed data were collected using a rated current meter for flow and manual rain gauges for precipitation. The spatial characteristics of the study area were processed using HEC-GeoHMS, as an ArcGIS extension. To simulate runoff volume and peak flow, two kinds of loss methods in HEC-HMS model were tested, and the parameters CN (curve number) and constant loss rate were found the most sensitive parameters. From this, improved simulation results (NSE of 88% and R-2 of 0.89) were obtained in a model set contained initial and constant loss method and SCS unit hydrograph. It is suggested that the methodology and results of the study could be further used in simulation of runoff in the catchment and other nearby catchments as an aid to planning.