Gabiri, Geofrey: Multi-scale modeling of water resources in a tropical inland valley and a tropical floodplain catchment in East Africa. - Bonn, 2019. - Dissertation, Rheinische Friedrich-Wilhelms-Universität Bonn.
Online-Ausgabe in bonndoc:
author = {{Geofrey Gabiri}},
title = {Multi-scale modeling of water resources in a tropical inland valley and a tropical floodplain catchment in East Africa},
school = {Rheinische Friedrich-Wilhelms-Universität Bonn},
year = 2019,
month = feb,

note = {This study investigated the dynamics of hydrological processes at the wetland-catchment scale through field scale-based analysis, point scale modeling using Hydrus-1D model along a floodplain transect in Tanzania and wetland-catchment modeling with SWAT model in an inland valley in Uganda. The impact of different land use management options and the projected climate change on the water resources of the inland valley were also evaluated using a hydrological response unit (HRU)-based (ArcSWAT2012) and a grid-based setup (SWATgrid) of the SWAT model. The inland valley is located in Namulonge, central Uganda, and it is one of the headwater catchments of Lake Kyoga basin. The inland valley catchment covers an area of 31 km2 with a wetland area of 4.5 km2. The floodplain is located in Kilombero district, Southern Tanzania and the catchment area is 40,240 km2 and the study area in a wetland is 96 km2. Both sites reflect the prevailing diversity of wetland attributes and uses.
Monitoring of hydro-meteorological data for both sites was conducted for two hydrological consecutive years of 2015 and 2016. The cross–section of the wetland transect was subdivided into three major hydrological positions defined as riparian zone, middle, and fringe. Hydrological instrumentation and data collection for soil moisture, soil properties, depth to shallow groundwater was conducted along these hydrological positions for both wetland systems. In addition, there was data mining from other sources.
Following the field-based analysis at a wetland scale in the inland valley, the spatial and temporal variability in soil moisture increased significantly (p<0.05) towards the riparian zone, however, no significant difference was observed between middle and riparian zone. The distribution of soil hydrological regimes, saturated, near and non- saturated regimes does not correlate with the hydrological positions. Precipitation strongly controlled the temporal variability while microscale topography, soil properties, distance from the stream, anthropogenic factors, and land use controlled the spatial variability of soil water availability in the inland valley
In the Kilombero floodplain, Hydrus-1D model was successfully calibrated (R2 = 0.54–0.92, RMSE = 0.02–0.11 cm3/cm3) using measured soil moisture content. Satisfying statistical measures (R2 = 0.36–0.89, RMSE = 0.03–0.13 cm3/cm3) were obtained when calibrations for one plot were validated with measured soil moisture for another plot within the same hydrological zone, indicating the transferability of the calibrated Hydrus-1D. The hydrological regimes correlated with the hydrological positions in the floodplain. Soil moisture dynamics is controlled by overbank flow, precipitation, and groundwater control at the riparian and middle zone, while it is controlled by rainfall and lateral flow from mountains at the fringe during the long rainy seasons. In the dry and short rainy seasons, rainfall, soil properties, and atmospheric demands control soil moisture dynamics at the riparian and middle zone.
For the wetland-catchment scale hydrological modeling in the inland valley, good model performance was achieved from the calibration and validation of daily discharge (R2 and NSE > 0.7) for both model setups (ArcSWAT2012 and SWATgrid). The annual water balance indicates that 849.5 mm representing 65% of precipitation is lost via evapotranspiration. Surface runoff (77.9 mm) and lateral flow (86.5 mm) are the highest contributors to stream flow. Four land use management options were developed in addition to the current land use system, with different water resources conservation levels (Conservation, Slope conservation, Protection of the headwater catchment, and Exploitation). There is a strong relationship between the first three management options with decreasing surface runoff, annual discharge and water yield while the fourth option will increase annual discharge and total water yield.
The future climate change in the inland valley was analyzed using climate scenarios RCP4.5 and 8.5 of six GCM-RCM models from the CORDEX-Africa project. Compared to the reference period of 1976-2005, a general increase in temperature of +0.9 0C to +1.9 0C over the period of 2021-2050 is projected by the model ensemble. A mixed change signal in annual precipitation (-30 to 43.9%) is projected among the six climatic models. However, on average, the models show an increase in annual precipitation of +7.4% and +21.8% under RCP4.5 and 8.5, respectively.
The application of the climate model ensembles in SWAT showed future discharge change similar to the projected precipitation change. The six climate models showed uncertainty in the annual discharge change ranging from -44 to 149% although on average, the climate models project an increase of +16% and +29% under RCP4.5 and 8.5, respectively. Wet and dry seasons are expected to get wetter and drier, respectively in the future. Compared to land use management options, climate change will have a dominant impact on the water resources in inland valleys. Adoption of Conservation, Slope conservation and protection of the headwater catchment options will significantly reduce the impacts of climate change on the total water yield and surface runoff and increase evapotranspiration and water availability in the inland valley.},

url = {}

The following license files are associated with this item: