Comparative Analysis of Groundwater Recharge Dynamics in Granular (Soil) and Fractured Aquifers Using Water Level Fluctuation Monitoring

Abstract Groundwater recharge estimation is fundamental for the sustainable management of water resources, especially in complex hydrogeological systems comprising granular and fractured aquifers. This study investigates recharge dynamics in two sub-basins through the monitoring of Water Table Fluctuation (WTF) in seven wells: five dug wells (DW) located in the granular aquifer (soil) and two tube wells (TW) in the fractured aquifer. Monitoring was conducted at 60-minute intervals between June 2017 and June 2018. Recharge was calculated using the WTF method, evaluating the method’s sensitivity to three distinct approaches for estimating water level variation ( $$\:{\Delta\:}\text{H}$$ ) (Graphical Approach, Master Recession Curve - MRC, and the RISE Method) and alternative specific-yield ( $$\:{S}_{y}$$ ) scenarios based on laboratory measurements, literature values, and an event-based water-balance approach. The event-based approach was applied only to the dug wells, whereas the $$\:{S}_{y}\:$$ values used for the tube wells were treated exclusively as methodological sensitivity scenarios. Statistical analyses of cross-correlation and autocorrelation were employed to define the response time and memory effect of each system. The results from the monitored sites elucidate fundamentally different dynamics: the dug wells generally exhibited rapid water-level responses to precipitation, with maximum rises of up to 1.5 m/day and peak cross-correlation lags of 0–2 days. Autocorrelation decreased below 0.2 within approximately 7–9 days in four dug wells, whereas DW5 and both tube wells remained above 0.2 at the maximum evaluated lag of 16 days. The tube wells showed attenuated responses, mean water-level rise rates below 0.02 m/day, and weak correlations with individual rainfall events. Estimated apparent recharge scenarios varied widely, ranging from 4.17% to 22.35% of valid precipitation for the TWs and from 13.20% to 108.85% for the DWs, demonstrating high sensitivity to the adopted methodological assumptions. It is concluded that the two domains exhibit distinctly contrasting hydraulic response times, emphasizing the need to treat shallow and fractured systems separately in groundwater management and highlighting the methodological uncertainties inherent in estimating absolute recharge. The findings underscore the importance of considering hydrodynamic and methodological characteristics in the sustainable management of groundwater resources, as well as highlighting the importance of using recharge assessment methods that utilize aquifer response data, such as the WTF method. The lack of an independent recharge validation method (e.g., environmental tracers or numerical modeling) remains a limitation of this study, emphasizing that the calculated rates represent methodological envelopes reflecting structural uncertainty rather than absolute deterministic metrics. Comparisons with water-balance or baseflow-separation estimates should account for differences in spatial scale, temporal resolution, storage representation, and the hydrogeological domain represented by each method.