Maybar Watershed Run off Modeling

Abstract

A study was conducted in Maybar catchment which is located in Amhara regional state in Debub Wello Zone Dessie Zuria Wereda.Maybar watershed cover’s area of 115 ha (1.15km2). Watershed rainfall run-off modeling using Arc SWAT is an important tool in the study of water resources and water management of the watersheds. Watershed rainfall runoff models are mainly used for river flow forecasting for the management of the resource and to minimize the ill effects through early warning measures. In this thesis we use different data to process the SWAT model properly. The input data for the modal includes digital elevation model land use, land cover, soil, hydrological and weather data which are the main input for Arc SWAT model. Digital elevation model, land use, land cover data and soil data are prepared in raster format and must be projected. Weather data are prepared in text format. Arc SWAT model uses daily rainfall, both daily minimum and maximum temperature, daily relative humidity, daily wind speed, daily sunshine hour and also daily flow data. The meteorological data were collected from USGS websites (http://globalweather.tamu.edu) with the geographical reference and the data obtained are in text format. The detail land use land cover and soil data were also obtained from the USGS website (http://www.waterbase.org)and the soil maps were used as per FAO classification. Weather generator was also created to fill-in missing gaps and generates climate data Soil and Water Assessment Tool (SWAT 2012) integrated with Arc GIS 10.2 and. We use to model the watershed which accounts spatial and temporal variation of inputs at HRUs level. To carry out sensitivity analysis, the most sensitive parameters are calculated; hence, Cn_2 and Alfa-Bf are more relatively sensitive for the year 1993 -2013. Calibration is also done for 15 years, that is, 1993– 2007, as presented in the discussion and result section. The same data arrangement steps are used for all these data. Time series plots and the statistical measures of coefficient of determination (R2), Nash-Sutcliffe Efficiency (ENS) and PBIAS were used to evaluate the performance of the model. The results of the model calibration and validation showed reliable estimates of monthly flow yield with R2=0.78, RNS=0.75 and PBIAS=-0.36 during the calibration period and R2=0.70, RNS=0.61 and PBIAS=0.59 during the validation period.