Statistical downscaling of climate variables for local forecasts and applications to improve climate change prediction in upper Blue Nile Basin

Abstract

Climate change studies are indeed crucial for sustainable and resilient development, especially in vulnerable regions. These studies help in understanding local climate change impacts under different scenarios, which is essential for water resource management, disaster mitigation, and agricultural development. Statistical Downscaling Model (SDSM), in common with CanESM5 (CMIP6) and counterparts CanESM2 (CMIP5) was used to predict temperature and rainfall in the study basin. CanESM5 (CMIP6) predictions were higher than CanESM2 (CMIP5) for both rainfall and temperature. Both CanESM2 and CanESM5 outputs fit well with observed data (R² values of 0.8–0.9) suggest good model performance. This is consistent with other studies that have found GCMs, including CanESM models, to be effective in simulating climate parameters. Projection of increased rainfall (up to 120 mm and 250 mm monthly for CanESM2 and CanESM5 respectively) with some areas showing reduction (up to 50 mm) aligns with the general trend of increased variability in precipitation patterns under climate change scenarios. The projected temperature increases of 0.5–2 °C is consistent with global warming trends for maximum temperature. The variation in minimum temperatures was not significant at most stations, with some showing up to 1 °C increase, is noteworthy and may have implications for local ecosystems and agriculture. Statistical downscaling works well for average predictions but struggles with extreme events is an important limitation to note. This aligns with the general challenges in climate modeling, where capturing extreme events remains a significant area for improvement.