Application Of Remote Sensing Data To Complement Hydrologic Modelling, The Upper Blue Nile Basin

Abstract

It has been recognized that reliable, long-term and well distributed climate information is essential to inform any development policy that aims to address the consequences of climate variability and change on water resources. However, in developing countries planning of such activity is greatly hampered by the lack of a sufficiently dense network of weather stations measuring precipitation. The objective of this dissertation is, therefore, to evaluate the freely available high resolution satellite rainfall estimates in the Lake Tana Basin which has a relatively denser ground rainfall stations network. Rainfall estimates of Tropical Rainfall Measuring Mission (TRMM) 3B42 version 7, EUMETSAT's Meteorological Product Extraction Facility (MPEF), Multi-Sensor Precipitation Estimate-Geostationary (MPEG) and Climate Forecast System Reanalysis (CFSR) are considered. The satellite rainfall is validated by a direct comparison with the gauged rainfall data, and through hydrological modelling to capture the observed flow using a semi-distributed hydrological model Hydrologiska Byråns Vattenbalansavdelning (HBV) and Parameter Efficient Distributed (PED). The result of direct comparison indicated that, the MPEG and CFSR rainfall provided the most accurate rainfall estimates. On average, for 38 stations, 78 and 86 % of the observed rainfall variation is explained by MPEG and CFSR data, respectively, while TRMM explained only 17% of the variation. The hydrological modelling indicated that both the gauged and the CFSR precipitation estimates were able to reproduce the stream flow well for either of the models. The TRMM data was not be able to capture the observed flow through model calibration for both models. Bias corrected MPEG rainfall by the gauged monthly means performed as well as or better than the gauged rainfall data in capturing the observed flow through hydrologic model calibration. This dissertation has also identified potential irrigable areas by considering hydrological and landscape factors that determine lack of irrigation development in the Ethiopian highland. Potential land areas suitable for surface irrigation were determined by using a GIS based MultiCriteria Evaluation (MCE) technique by considering climate characteristics (rainfall and evaporation), land features (soil type, land use and slope), market access (town and road proximity) and proximity to a perennial river. The available water for surface irrigation was quantified by analysing historical river flow data during the dry season of the major rivers in the Lake Tana Basin. The result indicated that the main limitation for surface irrigation in the Ethiopian highlands is the availability of water and not land suitable for irrigation.