We explored how disasters that occur before recovery from earlier events is complete can complicate recovery, using examples from science and the real world across different societal domains. We then reflect on why recovery under repeated disasters is more complex than simply returning to normal, and how the way we define recovery and the system we study shapes our understanding. These insights can help us better understand, study, and manage recovery from disasters.

Dust storms threaten health, travel and farming, but the conditions that prepare them may change with season. We studied eight 2024 storms in the Taklimakan Desert using two towers, weather records and air-path tracking. Spring storms were mainly prepared by strong regional pressure differences, while summer storms showed stronger early ground heating, followed by rising winds just before they began. Local terrain altered these signals. The results may improve seasonal dust-storm warnings.

A forecasting framework was developed to river flow prediction in data-scarce tropical regions. The study extracted hidden rainfall patterns from satellite precipitation records and integrated them with large-scale ocean–atmosphere climate signals into a statistical model, resulting in improved predictive accuracy. These findings can support more effective water-resources management during dry periods and strengthen early warning systems in regions with limited hydrometeorological monitoring.