Regime-based drought dynamics: a data-driven differential equation framework for state transitions, weak seasonality, and irregular variability

Abstract This study provides a comprehensive assessment of drought dynamics based on 576 monthly observations from multiple stations in the Wadi Isser basin. Descriptive analyses revealed that near-normal conditions dominate the climatic profile, averaging 18–21 stations per month, although sporadic episodes of severe drought and intense wetness highlight the system’s vulnerability to extreme anomalies. Cluster analysis identified three distinct drought regimes: (1) a stable, near-normal state with minimal anomalies; (2) wet-dominated months characterized by widespread rainfall and frequent extreme wet events; and (3) drought-dominated periods marked by severe and extensive water scarcity. Time-series decomposition confirmed that near-normal conditions are structurally persistent, lacking strong seasonality or long-term shifts, yet prone to abrupt short-term disruptions triggered by irregular climatic shocks. Complementing these insights, a data-driven SIR modeling framework revealed self-limiting dynamics: extreme events dissipate rapidly; near-normal states remain resilient but fragile when overly dominant; and moderate conditions act as transitional phases shaped by interactions between normal and extreme states. Collectively, these findings underscore a climate system that is predominantly stable yet punctuated by rare, intense, and spatially widespread drought or wet episodes. The results enhance understanding of drought variability and offer a robust analytical framework for anticipating hydro-climatic extremes in semi-arid Mediterranean basins.