Challenges in modelling the impact of frost and heat stress on the yield of cool-season annual grain crops

Jonathan Richetti^1*Victor Oscar Sadras²Di He¹Brenton Leske³Pengcheng Hu¹Yacob Beletse¹C. MARIANO COSSANI²Ha Nguyen¹David Matthew Deery¹Bangyou Zheng¹M. Fernanda Dreccer¹Jeremy Whish¹Julianne Lilley¹

• ¹Commonwealth Scientific and Industrial Research Organisation (CSIRO), Canberra, Australia
• ²South Australian Research and Development Institute, Adelaide, South Australia, Australia
• ³Department of Primary Industries and Regional Development of Western Australia (DPIRD), South Perth, Western Australia, Australia

Frost and heat events at critical growth stages could cause large yield losses. These temperature extremes are increasing in frequency and intensity due to climate change in many parts of the broadacre cropping regions globally, presenting challenges to food production. For cool-season grain-growing regions, where summers are already too hot, heat and frost risks can limit adaptation options. Capturing these stresses in crop models accurately is increasingly important for evaluating the timing, severity, and yield consequences of extreme events. However, most existing process-based models were not designed to simulate short-duration temperature extremes, limiting their ability to assess climate risk and inform adaptation to frost and heat. Yield responses to heat and frost are associated with pollen sterility, grain abortion, accelerated senescence, and grain filling. Six challenges limit current modelling approaches: (1) inadequate spatial and temporal resolution of extreme events, (2) threshold-based and non-linear crop responses, (3) interactions between phenology and management, (4) cumulative and interacting stress effects across development stages, (5) limited representation of genotype-specific sensitivities, and (6) reliance on daily temperature data. Addressing these challenges requires improved use of sub-daily climate data, incorporation of physiological damage mechanisms, and enhanced crop-and genotype-specific parameterisation. These developments are critical for improving crop yield predictions under extreme temperatures in the context of climate change.