Introduction: Drought is a gradual, multi-stage phenomenon in the hydrological cycle, beginning with precipitation deficits and, after a certain time lag, leading to reductions in surface runoff and groundwater resources. Despite the importance of this temporal lag, many studies have examined the relationship between meteorological and hydrological droughts without considering system response delays. Drought monitoring using standard indicators is a tool for sustainable water resource management and adaptation to climate change. The reconnaissance drought index (RDI) has a special place in drought studies due to its quantitative calculation, comparability in different assessments, and use of long-term data. The RDI enables multi-scale drought analysis by allowing it to be used in different periods of short-term droughts (RDI-3 – RDI-1), medium-term droughts (RDI-6), and long-term droughts (RDI-12 – RDI-24). Another standard indicator for assessing water scarcity in various studies is the stream drought index (SDI). Recent studies show that the intensity and timing of drought transitions are strongly dependent on climatic characteristics, watershed physical conditions, water storage capacity, and timing of the indices. In particular, hydrological systems typically respond with a time lag of several months, reflecting the role of cumulative storage such as infiltration, soil, and recharge aquifers. This study aims to quantify the time lag in drought propagation from meteorological to hydrological systems in the Damghan Watershed and to identify the dominant response delay of river flow to precipitation deficits.
Methodology: The Damghan Watershed, covering approximately 13,000 km², is located in a semi-arid climate, with annual rainfall ranging from 160 to 446 mm and elevations between 1281 and 3744 m. The climatic, hydrological, and geological characteristics have made Damghan Watershed a suitable example for analyzing the multiscale behavior of meteorological and hydrological droughts and examining the response of groundwater systems to climate change. Monthly precipitation and streamflow data from 1990 to 2020 were analyzed, and data quality was controlled using statistical tests (Grubbs test at 95% confidence). Meteorological drought was assessed using the RDI, while hydrological drought was evaluated using the SDI at 1-, 3-, 6-, 12-, and 24-month timescales. Standardized RDI and SDI series were calculated for multiple timescales to capture temporal variations in drought conditions. Pearson correlation analysis was applied to evaluate the relationship between meteorological and hydrological droughts, and lagged correlation analysis was performed to identify the dominant response delay of river flow to precipitation deficits. Long-term trends and changes in drought severity were assessed using the Mann-Kendall test and Sen’s slope estimator, while aggregation of indices over different timescales (1–24 months) was conducted to examine the effect of temporal scale on drought propagation and the cumulative behavior of the system.
Results and Discussion: Pearson correlation analysis between standardized RDI and SDI for 2008–2018 (120 months) showed a moderate concurrent correlation (Lag 0, r ≈ 0.48 at the 3-month scale), while lagged correlation peaked at a three-month delay (r = 0.71, p < 0.001), indicating a three-month hydrological memory in the system. Short-term scales (1–3 months) exhibited high variability in SDI (+2.1 to −2.4) and RDI (+1.8 to −2.0), whereas long-term scales (12–24 months) were more stable, with severe droughts (SDI < −1.5) persisting longer. Mann-Kendall and Sen’s slope analyses revealed a significant decline in wet conditions and an increasing frequency of moderate to severe droughts at the 24-month scale. Scale analysis further showed that correlation generally increased with aggregation period, but the three-month lag consistently displayed the highest correlation, highlighting its key role in water stress propagation. These results indicate that the Damghan Watershed requires approximately three months to convert a meteorological drought signal into reduced surface runoff, a lag influenced by soil moisture storage, infiltration, aquifer recharge, and the region’s geological structure. Overall, the findings suggest that drought is not an instantaneous event but a cumulative, time-dependent process, with maximum intensity transfer at medium-term scales, and understanding this delay can enhance early warning systems as well as guide water resource planning and reservoir management.
Conclusion: By integrating multi-scale drought index analysis with lagged correlation assessment, this study provides a quantitative framework to understand the temporal propagation of drought. The identification of a dominant three-month response delay highlights the watershed’s hydrological memory and can improve river flow forecasting, supporting climate-adaptive water management policies in semi-arid regions. Despite the significant results, it should be noted that the hydrological drought analysis in this study was based on data from a single hydrometric station; therefore, the findings related to SDI are not representative of the behavior of the entire Damghan Watershed and their generalization to the entire region should be done with caution. However, the analytical framework presented in this study, especially the use of time-lag correlation analysis, has the potential to be applied to other watersheds and similar studies. Overall, this study, focusing on time-lag analysis, provides a quantitative framework for explaining the transition from meteorological drought to hydrological drought, which can help improve drought monitoring, better understand the response of hydrological systems, and support decision-making in water resources management in semi-arid regions. It is suggested that future research, by utilizing data from multiple hydrometric stations and conducting spatial analyses, will enhance the generalizability of the results and provide a more comprehensive picture of hydrological drought behavior at the basin level. |
