Crop modelling with AquaCrop during climate change in the Ancash region of the Peruvian Andes

Peruvian Andean rural farmers often have precarious livelihoods and already experience less predictable weather conditions than in recent decades. With the goal of investigating hydrological and agricultural resilience in a region with an uncertain climate future (with regard to both temperature and precipitation), we present here the results obtained from using the AquaCrop software to model both crop growth and the consequent harvest yields in the valleys of the Peruvian Andes, including the Rio Santa Valley in the Ancash region. The crop models are presented for 1970-2099 (the historical and the future during climate change), using RCP2.6 & RCP8.5 Regional Climate Models (RCMs) from CORDEX at a spatial resolution of 0.22 degrees. We chose the CORDEX RCM data that was dynamically downscaled from the CMIP5 GCMs instead of the CHELSA statistically-downscaled data, since the downscaling of the CORDEX RCM data produces more locally-heterogeneous climate averages, which are more consistent with the variable topography. The CORDEX RCM model data has subsequently been bias-corrected to monthly CHIRPS precipitation and monthly ECMWF ERA-Interim temperature extremes from 1981-2005 for locations in the Ancash region, including Yungay and Aija. For the various crops that we modelled (maize/corn, potatoes, dry beans, quinoa, wheat), we find significant interannual variability of the dry yields from crop harvest (without irrigation or fertilizers), particularly when the climate is transitioning to a warmer one for those crops that prefer warmer climates. Without the consideration of irrigation or fertilizers, the possibility of high yield interannual variability could make it difficult for the Peruvian Andean farmers to plan ahead, and maintaining a diversity of crops within the Rio Santa Valley and the wider Ancash region could be advantageous for these farmers.