How important is adapting regional climatic projections to the local environment? A procedure for microclimatic corrections makes the difference for crop growth in a virtual experiment

Climatic conditions drive plant metabolism and growth and their projections are used as drivers in modelling experiments for the prediction of crop yields. However, these are generally issued at regional scale, and do not consider microclimatic variations. In this study we question the impact of taking into account microclimate adjustments on simulated crop yield. A procedure for the correction of temperature and humidity to the local microclimate is proposed and applied on climatic predictions for a site, resulting in modifications of about 2% in mean daily relative humidity and 2.5°C in mean daily maximum temperature. Crop yield is predicted with a processbased agroforestry model, for the pure crop and in the alley-cropping system, using both climatic series. Predicted crop yield differed by up to 58% in individual years and overall by 22% (CV RMSE) across climatic series. A significant trend in crop yield disappeared after corrections. This study highlights the importance of taking into account microclimatic corrections when using climatic projections to predict crop growth on realistic sites.