Browsing by Author "Mtamba, Joseph Ochieng"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Item
    Hydraulic modeling of Mara natural wetland using remote sensed data
    (University of Dar es Salaam, 2019) Mtamba, Joseph Ochieng
    In the last ten years climate change has increased concerns in natural wetlands and floodplains. Floods disaster has negative impacts on the infrastructure, livelihoods and environment. Sustainable planning and management for flood disaster management can be achieved through application of hydraulic models. In remote data scarce areas remote sensed data for topography and roughness may be used as an alternative source of data, hence evaluation of its suitability is necessary. Vegetation resistance influences water flow in floodplains. Characterization of vegetation for hydraulic modeling includes the description of the spatial variability of vegetation type, height and density. In this research, the application of dual polarized Radarsat-2 wide swath mode backscatter coefficients (o°) and Landsat imagery to derive spatial hydraulic roughness was explored. Simulations were performed using the FLO-2D hydraulic model to evaluate model performance under three Manning’s coefficient parameterizations which include derived optimum floodplain roughness, constant floodplain roughness referred to as scenario 1, 2 and 3 respectively. The model was used to derive spatial hydraulic indices and flood hazard maps for the floodplain wetland system. The model performance was evaluated using Nash-Sutcliffe model efficiency coefficient (E) and coefficient of determination (R^2), based on water levels measurements and simulated water levels at a gauging station within the wetland. The overall performance of scenario 2 to E = 0.95 and R^2 = 0.95, which was improved in scenario 2 to E = 0.95 and R^2= 0.99. When spatially distributed Manning values derived from SAR relative surface values were parameterized in the model, the model also performed well and yielding E = 0.97 and R^2 = 98. Improved model performance using spatial roughness shows that spatial roughness parameterization can support flood modeling and provide good flood wave simulation over the inundated riparian areas equality as calibrated model. Further, the results can be improved by more accurate elevation data.
  • Thumbnail Image
    Item
    Hydraulic modeling of Mara natural wetland using remote sensed data
    (University of Dar es Salaam, 2019) Mtamba, Joseph Ochieng
    In the last ten years, climate change has increased concerns in natural wetlands and floodplains. Foods disaster has negative impacts on the infrastructure, Livelihoods and environment. Sustainable planning and management for flood disaster management can be achieved through application of hydraulic models. In remote data scarce areas remote sensed data for topography and roughness may be used as an alternative source of data, hence evaluation of its suitability is necessary. Vegetation resistance influences water flow in floodplains. Characterization of vegetation for hydraulic modeling includes the description of the spatial variability of vegetation type, height and density. In this research, the application of dual polarized Radarsat-2 Wide swath mode backscatter coefficients (00) and Landsat imagery to derive spatial hydraulic roughness was explored. Simulations were performed using the FLO-2D hydraulic model to evaluate model to evaluate model performance under three Manning’s coefficient parameterizations which include derived optimum floodplain roughness, constant floodplain roughness and spatial Manning’s coefficients derived with aid of relative surface roughness referred to as scenario 1, 2 and 3 respectively. The model was used to derive spatial hydraulic indices and flood hazard maps for the floodplain wetland system. The model performance was evaluated using Nash-Sutcliffe model efficiency coefficient (E) and coefficient of determination (R2), based on water levels measurements and simulated water levels at a gauging station within the wetland. The overall performance of section 1 was characterized with E=0.75 and R2=0.95, which was improved in scenario 2 to E=0.95 and R2=0.99. When spatially distributed Manning values derived from SAR relative surface values were parameterized in the model, the model also performed well and yielding E=0.97 and R2=0.97 and R2=0.98. Improved model performance using spatial roughness shows that spatial roughness parameterization can support flood areas equally as calibrated model. Further, the results can be improved by more accurate elevation data.