Compound hot and dry extremes might be underestimated in CMIP6 projections

Abstract

Future projections of compound hot and dry extremes vary significantly across different ESMs in CMIP6. This affects our understanding of possible future risks of such events specially under high warming-high emission scenarios. We analyze the compound hot and dry extreme events in the SSP5-8.5 projections in 35 different ESMs contributing to CMIP6 for present-day climate (+1°C) and additional global warming levels (+1.5°C, +2°C, +3°C). It is found that the inter-model differences in the projections is quite significant. Investigating this spread, we establish a statistical emergent relationship between future changes in compound hot and dry extremes and the correlation between hot and dry events in the recent past. This robust association indicates that future changes in the global average likelihood of compound hot and dry extremes could be significantly underestimated. The emergent relationship also shows strong regional constraints, and areas which exhibit substantial underestimation of  future changes in compound hot and dry extreme projections are the North America, Amazonian, Mediterranean, West Africa and Mediterranean regions. The results contribute to understanding and reducing uncertainties in future projections of compound hot and dry extremes and therefore aid the formulation of effecting risk management and climate mitigation strategies.