Water Resource Engineering

Permanent URI for this collection

http://etd.aau.edu.et/handle/12345678/189

Browse

Browsing Water Resource Engineering by Subject "BCC-ESM1-MR"

Now showing 1 - 1 of 1
  • No Thumbnail Available
    Item
    Spatio-Temporal Variability of Historical and Future Climate Extremes and Their Impacts on Hydroclimatic Extremes in Ethiopia
    (Addis Ababa University, 2026-02-01) Daniel Berhanu Afrasso; Tena Alamirew; Solomon Gebreyohannis
    Climate variability and change have profound impacts on the livelihoods and economies of the East African region, including Ethiopia due to its strong reliance on rain-fed agriculture, high population exposure, and limited adaptive capacity. Ethiopia has historically experienced recurrent droughts and floods, and these extremes are expected to intensify under climate change. This study evaluated model performance, addressed biases, and assessed future projections of precipitation and hydroclimatic extremes across Ethiopia. Together, these studies highlight both the potential and the limitations of CMIP6 models for supporting climate impact assessments and adaptation planning in the country. An assessment of 45 CMIP6 models against ENACTS rainfall data revealed that while most models successfully captured the spatial distribution of mean and extremes, they could not reproduce the magnitude of indices, especially in highland regions such as the northwest, west and southwest. Many models introduced a pronounced dry bias in high-elevation areas that typically receive substantial rainfall during the June–September (JJAS) main rainy season. Extreme indices such as consecutive dry days (CDD), total wet-day precipitation (PRCPTOT), heavy precipitation days (R10), very heavy precipitation days (R20), simple daily intensity index (SDII), very wet days (R95pTOT), extremely wet days (R99pTOT), maximum consecutive five-day precipitation (Rx5day), and maximum one-day precipitation (Rx1day) were generally underestimated. However, consecutive wet days (CWD) were often overestimated in highland and high-rainfall areas, particularly in the western parts of tropical, subtropical, temperate, and alpine agro ecological zones (AEZs). Conversely, in lowland and drier areas such as east and northeast parts of desert and tropical AEZs, model biases were smaller. Evident seasonal and regional differences in model skill suggest that model performance depends heavily on the choice of indices, and metrics, as well as on the season and geographic context. Using Taylor Skill Scores (TSS) and Comprehensive Rating Metrics (CRM), the most reliable models for Ethiopia during the JJAS season were identified as CESM2-WACCM-FV2, NESM3, NorESM2-LM, NorESM2-MM, CMCC-ESM2, IPSL-CM6A-INCA, and E3SM-1-0. Of the 45 models evaluated against the observed dataset, NorESM2-MM performed best during the FMAM season. The next best performers were HadGEM3-GC31-MM, GFDL-CM4, NorESM2-LM, GFDL-ESM4, BCC-ESM1-MR, and MPI-ESM-1-2-HAM. Nevertheless, ensembles of those top ranking models consistently provided better alignment with observed indices compared to either individual models or the full model ensemble. These results emphasize that model weighting and sub-ensemble selection are critical steps in applying CMIP6 outputs to climate risk assessments. Projections of future precipitation extremes offer a more detailed view of potential risks. Using bias-corrected outputs from top-performing CMIP6 models under SSP245 and SSP585 scenarios, studies indicate that, by the 2050s and 2080s, there will be an increase in PRCPTOT and an intensification of extremes such as R10, R20, R95p, R99p, and SDII in northwest, west, and southwest Ethiopia during the JJAS period. In FMAM, however, the historical declining trend in rainfall may persist in some regions, particularly in the northeast and the Awash basin, though increases are projected in the north (Tekeze basin). The northeast is projected to face more intense extremes during JJAS but decreased rainfall during FMAM. These divergent patterns highlight growing intra-annual and regional variability in precipitation. At the basin scale, the Awash basin illustrates the severity of these challenges. Historical analyses reveal distinct spatial contrasts, with relatively stable precipitation in the upper basin but greater variability downstream. Using bias-corrected outputs from seven top-performing CMIP6 models, hydrological modeling with SWAT+ demonstrates significant changes in precipitation, streamflow, and extremes across the basin. Projections suggest increases in JJAS rainfall and CWD, particularly in the lower basin, but declines in FMAM rainfall and streamflow in the Mojo catchment, where flows may decrease by 80–100%. Meanwhile, JJAS streamflow is expected to rise by up to 70% in response to increased rainfall. Temperature increases of about +2 °C by the 2080s will further exacerbate evapotranspiration pressures. Both drought severity (+36%) and flood frequency (+48%) are projected to rise, underscoring a dual challenge of managing water deficits and surpluses. These findings highlight the urgent need for integrated climate change resilience water resource management strategies that can accommodate both extremes within the same basin. The findings of these studies emphasize the following important conclusions. First, while CMIP6 models are valuable tools, they require careful evaluation, bias correction, and selective ensemble construction to provide actionable insights. Second, Ethiopia’s climate future will likely be characterized by heightened variability and intensification of extremes. Third, adaptation and resilience strategies must be informed by both the potential for increased rainfall and flooding in some regions and declining rainfall and water stress in others. Finally, uncertainty in climate model projections must be explicitly recognized in planning. Acknowledging these uncertainties allows for the adoption of flexible and adaptive management strategies that remain robust under a range of possible futures.