Disparities in recovery capacity amplify inequality under consecutive extreme events

Recovery from extreme events remains poorly understood, yet it critically shapes long-term development opportunities. Especially, if the recovery from an extreme event is still ongoing when a subsequent disaster strikes, potentially causing poverty traps. This may become more likely with more intense and frequent extreme events under climate change. Unequal pre-disaster conditions may influence post-disaster recovery capacities and associated inequalities. In this work, we employ an agent-based model that explicitly resolves household-level recovery dynamics to assess the distributional effects of tropical cyclones under different warming scenarios, accounting for changes in cyclone intensity and frequency. The model is constrained using empirical insights from observed changes in nighttime light intensity after historical tropical cyclones, allowing us to link hazard intensity to recovery times across income groups. We drive the agent-based model with future tropical cyclone time series derived from the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP).We assess asset damage, consumption losses, and well-being losses across income groups and countries. Our results show that longer recovery times among low-income households amplify inequality, particularly in terms of well-being losses. Depending on national hazard and income distributions, patterns of poverty risk arising from incomplete recovery vary across countries and warming levels. Our observationally constrained modeling framework enables the explicit incorporation of recovery processes into both historical impact assessments and future risk analyses, resolving losses across different income groups. Moreover, the framework is transferable beyond tropical cyclones to other capital-destroying hazards, such as floods.