Persistence of hydrological droughts in Slovakia: a Markov-chain analysis of multi-season river flow variability

Background/introduction Drought persistence across physiographically and climatically diverse river basins remains poorly quantified, limiting probabilistic water-resource planning under a warming climate. This study provides the first such assessment for Slovakia, analyzing 17 gauging stations with long near-natural discharge records spanning Carpathian headwaters to lowland reaches. Methods Daily discharges were aggregated into winter (October–March) and summer (April–September) seasons; seasons with negative anomalies were classified as dry, and others as wet. Conditional dry-to-dry (Pdd) and wet-to-wet (Pww) transition probabilities were estimated at each station, and two-state first-order Markov chains were fitted. Long synthetic series were generated to compare observed and simulated spell-length distributions, derive 100-year maximum dry and wet spells, and estimate probabilities of exceeding multi-year drought thresholds over 5- and 10-year horizons. Results Drought persistence varies markedly across the study domain. High-elevation forested headwaters exhibit high seasonal variability but low persistence, whereas lowland and larger basins in southern and western Slovakia have Pdd > Pww and dry spells of 13–19 seasons. First-order Markov chains reproduce observed spell-length distributions well at most stations. Simulations indicate 100-year meteorological droughts of approximately 3-4 years in alpine basins but up to a decade in some lowland rivers, with a 3–5-year drought likely within a 10-year horizon. Discussion/conclusion At several sites, the maximum observed wet spells exceed modelled 100-year events, revealing asymmetric wet-dry persistence not fully captured by first-order Markov chains. These findings offer the first probabilistic framework for multi-season drought duration in Slovakia, with direct implications for water-resource planning under ongoing climate warming.