Prediction of future climate in semi-arid catchment under CMIP6 scenarios: A case study of Bahi (Manyoni) catchment in Internal Drainage basin (IDB), Tanzania

Abstract

Prediction of future climate is vital, especially for areas with low rainfall amounts, rapid population growth and high groundwater dependency. Bahi (Manyoni) is among the driest catchment hosting the new capital city of Tanzania which depends heavily on groundwater resources. Therefore, future climate prediction was done using the Coupled Model Intercomparison Project Phase 6 (CMIP6) scenarios of SSP126, SSP245 and SSP585. Linear Scaling (LS) bias-corrected on European center Reanalysis 5th generation (ERA5) was used to supplement temperature data. Out of 15 Global Climate Models (GCMs), CNRM-CM6-1-HR, INM-CM4-8 and INM-CM5-0 for rainfall and AWI-CM-1-1-MR, MRI-ESM2-0, EC-Earth3 and EC-Earth3-Veg for temperature were selected based on their highest performance in Taylor Skill Score (TSS) and subsequently their ensemble used. TSS of 0.83, 0.90, and 0.65 were obtained for rainfall, minimum temperature (Tmin), and maximum temperature (Tmax) ensembles respectively. LARS-WG 6.0 was applied for downscaling climatic data. Findings predicted that in 2080s for SSP126 and SSP585 scenarios, catchment mean annual rainfall to decrease between 13.8% and 4.3% in northern part, and to increase between 14.5% and 25.4% in southeast with an overall average Tmin increase between 0.2 °C and 4.5 °C and Tmax between 0.8 °C and 2.8 °C. This indicated relatively warmer and drier climate in northern compared to southern part of the catchment. These differences in microclimates could not be captured in large-scale studies such as IPCC, 2021. These findings highlighted potential climate change impacts on crop production and groundwater recharge, which contributes to Tanzania's Nationally Determined Contributions and National Adaptation Plans against future climate change.