Partitioning the Uncertainties in Compound Hot and Dry Precipitation, Soil Moisture, and Runoff Extremes Projections in CMIP6

A comprehensive assessment of compound hot and dry extremes based on different drought conditions (low precipitation, runoff, or soil moisture) and associated uncertainties is necessary to fully understand the possible risks. Here, we analyze changes in the likelihood of compound hot and dry conditions associated with low precipitation, runoff, and soil moisture using Coupled Model Intercomparison Project Phase6 (CMIP6) simulations for present-day climate (+1 degrees C) and additional global warming levels (+1.5 degrees C, +2 degrees C, +3 degrees C). Further, we investigate the contributions of different components (e.g., global warming levels, climate models, copula types) to the total spread in their future projections. Results show the significance of global warming levels in governing risks of rising compound hot and dry extremes. The hotspot regions include the Mediterranean, South Central America, Amazonia, and Sahara. The rising risks are also accompanied by rising uncertainty as the spread in changing likelihood is significantly contributed by Earth System Models (ESMs), global warming levels, their interactions, and the statistical estimation error. The uncertainty due to ESMs spread was observed to be most significant in the case of compound hot and low soil moisture extremes, which also corresponds to some of the most impactful conditions. It was observed that the estimation error dominates the uncertainty in compound hot and low precipitation extremes as compared to the two other combinations. Our findings indicate that the regional likelihood and associated uncertainties of compound hot-dry events in CMIP6 projections are functions of both the selection of drought types and the methodology of deriving the joint extremes.Plain Language Summary Concurrent hot and dry events can have devastating impacts on human health and ecosystems. Recently, it has been realized that compound hot and dry events have become more frequent under increasing global warming. But the projections of compound hot and dry extremes are accompanied by uncertainties at different levels. Therefore, assessment of both the occurrence under different global warming levels as well as the understanding of uncertainty is equally important for deciding effective risk management strategies. The investigation of the effects of human-induced climate for compound hot and dry extremes has been mainly focused on the effects associated with meteorological droughts. Since, the projected changes in the compound occurrences of hot and different drought conditions could lead to varied impacts for different sectors, a more comprehensive approach is required. Here, we estimate the changes in the occurrence of compound hot and dry extremes using the climate model simulations under four global warming levels and identify the major sources of uncertainty in projections. We find that a significant increase in likelihood is expected at increasing warming levels with considerable drought dependent uncertainty which is contributed by the climate models, warming magnitude and the effect of sample size.