Modelling crop yields and water balances for Ethiopia with LPJmL

Ethiopia faces climate changes that could impact its food and water security. This report describes the modelling of crop production and water balances for two separate studies on Ethiopia using the Lund Potsdam Jena managed Land (LPJmL) model. For the project Soy10 we modelled the effect of a 10% shift in protein source from cereals to soybean on land use and water demand, and explored how resilient these crop’s yields would be to climate change. For the project Multiple Scales and Extreme Events (MSX) we modelled Ethiopia’s crop yields at the rainfed potential level and the corresponding water demand compared to current levels. Results for Soy10 indicate that the assumed protein shift could free up around 2% of Ethiopia’s physical crop area without negatively impacting water availability, and that soybean yields would be more climate resilient than cereals under the selected climate scenario. Results for MSX suggest that Ethiopia has the potential to greatly improve its current yields through production intensification. Intensification would lead to a higher total agricultural water use, but also to greater water use efficiency. Note, however, that results should be interpreted with care, as there are still some points to improve in the modelling process which could affect the results.