Hydrological Studies of Mereb-Gash Basin in the Context of Water Management: Fid-Bau Portal

Hydrological Studies of Mereb-Gash Basin in the Context of Water Management

Kozlov, Dmitry / Ghebrehiwot, Anghesom

The present work attempts to obtain unique hydrological information and effective tools for streamflow modelling from insufficiently gauged river basin. To this end, physically-based semi-distributed and conceptual models are investigated using climate reanalysis datasets and geomorphologic catchment characteristics. Prior to feeding model forcing variables, their statistical and spatial patterns, methods of potential evapotranspiration estimations, and basin drought conditions are studied. Two or more approaches have been employed for each of these tasks. The findings from statistical and spatial analyses indicate presence of predominantly significant monotonic trends: precipitation and humidity tend to decrease, whereas temperature and potential evapotranspiration tend to increase with considerable spatial variability in the basin. Similarly, the drought indices exhibit predominantly humid conditions from 1979 through the end of 1990s but persistent dry conditions over the period 2000 to 2013. Out of the three methods for the estimation of potential evapotranspiration, Hargreaves and Penman–Monteith produced consistent estimations, the latter being with the highest values. Concerning the rainfall-runoff models, the goodness-of-fit results for single-event-based hydrographs indicate that GIUH-Nash model has the potential to produce acceptable values in most cases irrespective of sources and resolutions of digital elevation models, but it is found to be sensitive to the type of algorithms selected for stream network generations. The findings of the continuous simulation models suggest considerable overestimation of most of the precipitation in the reanalysis datasets, which in turn has a significant effect on other variables, such as potential evapotranspiration and thereby leading to a substantial discrepancy between simulated and observed. While there is no substitute for restoring and enhancing ground observation networks, further research that aims at identifying effective mechanisms to properly utilize reanalysis datasets are suggested.