Food security is threatened by compound events (extreme events like heat and drought occurring together), intensifying with climate change. Crucial for studying their impact on crop yield variability is the setting of temperature and precipitation thresholds. While relative thresholds (e.g., the 95th percentile) can hardly be justified concerning plant physiology, absolute thresholds (e.g., 30 °C) are expected to differ substantially between plant-level and large-scale assessments. As this contradiction has not yet been addressed, suitable relative and related absolute thresholds for the prominent crops grain maize and winter wheat are examined in this study. With these, it is analyzed whether extreme or compound events explain yield variability better and which development phase is sensitive to them. Also novel in the approach is to compare defining heat with daily mean and maximum temperatures and drought over 10 and 30 days. The analysis covers the years 1983 to 2021 and the 96 administrative districts of Bavaria, Germany, which are located in central Europe and exhibit a considerable precipitation gradient. Relative thresholds vary over this gradient, yet lead to similar absolute thresholds. This indicates that absolute thresholds are more suitable to explain crop yield variability. The discovered thresholds for daily maximum temperatures are at least 28 °C for grain maize and 24 °C to 25 °C for winter wheat, being lower than in plant-level analyses. Compound events have more impact on grain maize compared to individual extreme events. Yet, this effect was not revealed for winter wheat yields, showing the greatest sensitivity to individual heat events. During the vegetative phase, grain maize was most sensitive to heat. During the reproductive phase, grain maize was most sensitive to drought and winter wheat to heat. These results can be used in the methodology of further studies and for developing measures that buffer the impact of compound events on crop yields.
Bogenreuther et al. (Wed,) studied this question.