Spatial modelling of flood impact and community resilience in Eastern Uganda
MetadataShow full item record
Since 1900 to date, increasing hydro-meteorological disaster events have often left lives and livelihoods in vulnerable communities affected by their impact. Far East Uganda i.e. Elgon, Teso and Bukedi sub-regions, has continuously experienced extreme floods. This area has over the years registered huge losses of lives mainly from floods and landslides, wreaking havoc on the existing communities, damaging houses and crops, and in several cases leading to loss of lives. With limited resources directed towards building community resilience, identifying spatial patterns of various susceptibilities and adaptive capacities, and modelling the underlying relationships between resilience and impact is crucial in producing information to support strengthening resilience. The aim of the study was to explore the influence of community resilience on flood impact in hotspot areas. Primary data was collected through the Uganda Red Cross Society’s branch structure that was requested to capture the general impact/ damage level as observed by the communities in the study area on a scale of 1 – 5. Field visits were also conducted in the study locations and discussions done with communities and their local leaders. The study utilized the exploratory spatial data analysis, hotspot and geographically weighted regression tools to explore data, analyze flood impact hot-spot, cold-spots, outliers, regions of no change, and examine relationship between vulnerabilities, adaptive capacities, and flood impact. Hotspot clusters were identified in Katine, Arapai and Kamuda in the Teso sub-region. These areas were therefore noted to have concentration of higher flood impact compared to other parts of the study area and thus experience significantly high flood impact. Cold spots were identified in Bukhalu and Bunambutye in Elgon sub-region. Therefore, these areas showed concentration of lower flood impact compared to other parts of the study area thus they generally experience low flood impact. Regions of no clustering were identified in Butaleja, Budaka, Naboa, Merikit within the Bukedi sub-region. Therefore, the flood impacts experienced in this areas are not concentrated and thus happen with varying magnitudes that portray no association or commonality. An under prediction in Teso and over prediction in Elgon indicates the variation between the model values and the actual data values and thus how the data fits onto the model. Since that was not conclusive, a Moran’s I spatial autocorrelation was run on the GWR residuals to test for bias. The Z-score for the residuals was -0.356810 that fell within the region of randomness and no bias. Therefore, the residuals were normally distributed depicting a good GWR model and hence acceptable regression analysis between the dependent and independent variables. From the study, it was noted that community susceptibilities and adaptive capacities (resilience indicators) greatly influence how Eastern Uganda (Elgon, Teso, Bukedi) is impacted by floods especially in flood hotspots. The characteristics of building material strongly influence the effect of floods on houses/ dwellings. Other Socio-economic characteristics equally play a great role in influencing flood impact. Information about resilience and flood impact produced by GIS analyses in the study therefore provides scientific evidence on their relationships and is useful in informing resilience strengthening mechanisms. For better modelling of spatial relationships disaster data should be collected for every flood incident and proper record keeping should be emphasized among stakeholders/actors in Disaster Risk Reduction (DRR). Further research can be conducted to probe or develop specific regression algorithms to model flood impact in Eastern Uganda. Community awareness/ sensitization on flood risk should be scaled up by community leaders and government to support inclusive community-based resilience approaches in Eastern Uganda. Dissemination of weather forecast information and advisories at community level should be done to inform resilience building mechanism at that level.