Machine-learning models of coral cover and life histories reveal that climate refugia for coral reefs persist into 2050

Climate change is accelerating the decline of coral reefs, yet some locations may retain conditions that support persistence under future warming. We compiled 45,091 coral field observations (1960–2025) and 42 climate, biophysical, and human-pressure predictors to train machine-learning ensembles that estimate coral cover in 2020 and 2050, and life-history composition in 2020, across a global high-resolution (250 m) dataset of suitable coral reef habitat (8.9 million pixels; 554,443 km2). We then used a multi-objective optimization framework to identify reef areas that jointly maximize current and future coral cover, life-history diversity, spatial cohesion, and model certainty. Our results reveal 166,364 km² of climate-resilient reefs spanning 72 countries and 100 territories and jurisdictions, adding 30 countries and 54 jurisdictions beyond the original 50 Reefs assessment (Beyer et al. 2018). More than half of this resilient habitat (61%; 100,775 km2) occurs within five countries with extensive reef systems: the Bahamas, Cuba, Australia, Indonesia, and the Philippines. These findings highlight the uneven but widespread distribution of potential climate refugia and emphasize the opportunity for countries to strategically focus national policies, conservation finance, and management actions on climate-resilient areas most likely to sustain coral reef futures.