Future climate change increases the risk of wheat yield loss due to agricultural drought in southeastern Australia

Agricultural drought poses a significant threat to food security and human sustainability by reducing crop yields, and it is projected to intensify in the future due to ongoing global warming and increasing rainfall variability. As a key contributor to the worldwide food supply, the New South Wales (NSW) wheat belt in southeastern Australia is highly exposed to drought-related risks due to its prevailing dry climate and reliance on the rain-fed cropping system. Yet the future impacts of agricultural drought on regional wheat yields remain poorly quantified under climate change. This study aims to evaluate the risk of wheat yield losses induced by agricultural drought under different climate scenarios, focusing on its spatial distribution and temporal evolution across the NSW wheat belt. We integrated a process-based crop simulation model with a probabilistic approach to assess wheat yield loss risk under future climate scenarios. Agricultural Production System sIMulator (APSIM) model was forced with climate data from multiple Global Climate Models (GCMs), enabling the simulation of long-term wheat yield and plant available water (PAW). The simulated PAW values were standardized to derive SPAWI (Standardized Plant Available Water Index), which was used to characterize drought conditions. Copula functions were then utilized to construct the joint probability distribution between wheat yield and SPAWI, allowing the calculation of yield loss probabilities and the identification of drought trigger thresholds. There was a rising trend in agricultural drought frequency across the wheat belt, especially under the Hot/Dry scenarios. Regional results showed elevated wheat yield loss probabilities in the future, approaching 10 % in the drier and warmer areas. Moreover, the drought index thresholds for triggering wheat yield loss were higher over dry areas but lower in the wet region. Uncertainty attribution analysis identified GCM selection as the primary source of yield loss probability change in arid regions, while in wet regions, the choice of copula function played a more critical role. Our findings show a rising risk of wheat yield loss in the NSW wheat belt under future climate scenarios and reveal substantial spatiotemporal heterogeneity in yield impacts. The results offer critical geographic insights for supporting localized adaptation strategies and evidence-based agricultural planning under drought conditions in the future.