Observational constraint on climate model projections of global compound hot–dry events and the socioeconomic risks under climate change

The frequency of compound hot–dry events ( f HD ) is projected to increase significantly with future warming, yet associated uncertainties remain considerable and poorly constrained. In this study, we constrain future projections of f HD (2070–2099) using observations of recent trends in temperature ( T ) and precipitation ( P ) (1980–2014) during the warm-seasons. The physical mechanism is that the variance of f HD across climate models is dominated by their projected changes in P (Δ P ), which can be constrained by recent trends in T and P . Compared to the raw projections, the observationally constrained f HD is reduced by 9.68%–18.74%, with uncertainty narrowed by 3.79%–10.66% under the high emission scenario. The highest decline of f HD is located in regions with low population and gross domestic product (GDP), and globally, population and GDP exposures to f HD are reduced by 6.02%–10.73% and 6.51%–12.03%, respectively. The observationally constrained f HD with lower uncertainty provides more reliable information for risk management under climate change.