Development of a hydrologic model of Kagera river basin using remote sensing data

Model-based decision support tools are important for the management of natural resourcesspecially water resources. Specifically in developing countries, there is a lack of local data(ground collected data) and information to build, calibrate and validate the underlyinghydrological models. We should therefore use alternatives source of data that such asremote sensing data or global data.There are various archives of global datasets available online that could be used forhydrological modeling. They range from topographical data (SRTM) to soil data (FAO soilmaps) and land use (Global land use maps) as well as climatic data. The most interestingare climatic datasets, which include precipitation and temperature. Some data have beencollected by different satellites such as TRMM, NCEP/NCAR Reanalysis, others providestatistical data. An example is the CRU monthly datasets for rainfall, temperature andnumber of wet days per month that can produce daily data in as stochastic manner, usingthe weather generator DWGA.In this research, remote sensed data and global data have been used and evaluated todevelop a hydrological model for Kagera River basin. The Kagera catchment is located ineast-central Africa. This is a sub-catchment of the Nile River basin and it is distributed infour countries: Burundi, Rwanda, Tanzania and Uganda. The model was developed usingSoil and the Water Assessment Tool-SWAT, which is a continuous time and hydrologicalmodel used for river basin scale modeling. Satellite datasets and global datasets were usedto generate the inputs for simulating daily/monthly flows for the Kagera basin.All the precipitation datasets had almost the same temporal and spatial distribution;however the precipitation depths varied from one data set to another. The simulated flowsshowed clear differences of model response to those precipitation data sets. The SWATmodel performance was low, due to precipitation data quality, model resolution andvariability in topography, climate and landform of the study area