Observed and CMIP6 simulated occurrence and intensity of compound agroclimatic extremes over maize harvested areas in China

Understanding the changes in the frequency and intensity of compound agroclimatic extremes is important for studying the resilience of the food system under anthropogenic warming. However, the spatiotemporal variation of compound agroclimatic extremes for specific crops, and the performance of the Coupled Model Intercomparison Project Phase 6 (CMIP6) simulations in reproducing them, have largely been under-addressed. Herein, we have investigated the spatiotemporal variation of the occurrences and intensities of four compound agroclimatic extremes (hot-wet, hot-dry, cold-wet, cold-dry) for maize harvested areas in China during 1990–2014 and examined the capability of the CMIP6 to capture these variations. The results did not reveal any significant trends but there was a pronounced interannual volatility in both the occurrence and intensity of compound agroclimatic extremes. Seventy one percent of the maize harvested areas in China experienced at least one compound extreme event during the study period. Hotspots for high occurrence and high intensity included the major plains in the Huang-Huai-Hai and southern China maize zones. In general, five general circulation models (GCMs) from the CMIP6 relatively poorly captured the variation of the occurrence and intensity, and reported lower interannual volatility and less regional disparity compared to the observations, and showed weakness in capturing time-specific events. At the national scale, the five GCM ensemble mean outperformed any single model, and no single GCM outperformed any other individual GCMs. The best model differed according to the specific region of interest and the type of event. Our results highlight the need to improve field management decisions and adopt adaptation strategies in accord with local conditions to reduce the potential impacts on food systems.

Abstract licence: CC BY-NC

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.