Fig. 3: Projected hydropower exposure to drought and losses from drought-induced excess PM_2.5.

a–f Percent change in the mean fraction of hydropower generation affected by drought 2020–2059 relative to the mean 2000–2019. Each box plot represents one of three Shared Socioeconomic Pathways (SSPs) paired with Representative Concentration Pathways (RCPs) (SSP-RCP scenarios). The distribution of runoff projections from the 22 climate and earth system models determines the box plot’s spread. Box plots indicate median (middle line), 25th, 75th percentile (box), 1.5 times the interquartile range (whiskers), outliers (single points), and mean values (triangles). The IEA sub-regions are: b Central America and Mexico, c Southern South America (Argentina, Bolivia, and Chile), d Andean Region (Colombia, Ecuador, and Peru), e Rest of South America (Brazil, Venezuela, Paraguay, and Uruguay), and f Caribbean. g, h Plot the observed and projected evolution of premature deaths and the corresponding monetized losses between 2000 and 2059. Estimates of future losses are derived from demographic, climate, and economic projections under the three SSP-RCP scenarios. The retirement schedule of combustion power plants follows electricity sector scenarios from IEA⁶⁹. The Announced Pledges Scenario (APS) assumes retirements in line with existing country pledges. The Stated Policies Scenario (STEPS) assumes retirements in line with the IEA's current assessment of the region’s energy direction. The Reference Electricity Scenario (RES) assumes no retirements. All 66 projections (22 models  × 3 scenarios) also assume that additional electricity demand is met with non-combustion power, that no new combustion power plants are introduced, and that the concentration-response function is constant (no adaptation). Bold lines are LOESS curve fits. Source data are provided as a Source Data file (sourcedata.xlsx). The data and code used to obtain the estimates are available at https://www.openicpsr.org/openicpsr/project/217201.