PROJECT DESCRIPTION

Mountainous regions are disproportionately affected by the ongoing global climate crises. In complex landscapes such as the Ethiopian, coarse-resolution climate datasets poorly explain the spatial details and hence limited capabilities for climate adaptation/mitigation studies. In addition, the CMIP6 datasets are large (e.g., global), not easy to download, and bias-correct for Ethiopian users. To this end, the AMUClim project aims to generate a high-resolution (4x4 km2), bias-corrected, and open-access climate data repository at Arba Minch University (AMU), Ethiopia. We use AI models to downscale and bias-correct the recent IPCC s Coupled Model Intercomparison Project Phase 6 (CMIP6) at coarse-resolution (>100x100 km2) climate projection model outputs. In AMUClim, >10 climate models, a baseline (1981-2014), and four future projected scenarios (2015-2100) will be used as inputs to the AI model. To test the AI model, we use the Enhancing National Climate Services (ENACTS) high-resolution, blended (satellite-surface observation) datasets covering Ethiopia. The trained and tested AI model will be used to generate high-resolution CMIP6 model outputs. These datasets will be processed, stored at the AMU s cutting-edge computing (HPC) facilities, and freely shared with all global end-users to facilitate open science. The AMUClim dataset will be used for teaching and climate applications in universities and research institutes across the nation. To our knowledge, AMUClim dataset will be the first openly and locally available climate repository in Ethiopia. Our project is scalable not only to the East African mountainous region but also to sub-Saharan Africa. It is also sustainable because AMU established the State-of-the-art computing facility with a dedicated Climate and Atmospheric Research Center (C-CAR) with research staff. First, AMU graduate and staff members and staff members of the Ethiopian Meteorology Institute (EMI) will be trained in applying AI models for complex climate datasets. Second, a high-resolution and high-quality dataset will be developed and available for end-users to facilitate the formulation of cutting-edge research that can influence policies in the areas of agriculture, health, water, and the environment. The AMUClim project contributes to the GEO Mountains aims/objectives and to the realization of the UN s agenda 2030 (Goals 4-5 and 13).