1 Introduction

Most existing studies have conventionally treated the permeability of aquifer-aquitard systems as constant. However, recent findings increasingly highlight that permeability is a dynamic parameter, modulated by diverse natural processes, such as earthquakes and rainfall events (e.g., Elkhoury et al., 2006; Liao & Wang, 2018; Liao et al., 2015, 2021, 2022, 2025; Manga et al., 2012; Qi et al., 2023; Shi & Wang, 2016; Y. Shi et al., 2019; X. Sun et al., 2020; Thomas et al., 2023, 2024). Continuous monitoring of water levels within observation wells, interpreted through tidal responses, has become an established methodology to derive permeability characteristics of these groundwater systems quantitatively, providing critical insights into permeability fluctuations associated with seismic and meteorological perturbations (Hsieh et al., 1987; X. Sun & Xiang, 2021; G. Wang, Wang, et al., 2025; H. F. Wang, Yi, et al., 2025; C. Y. Wang et al., 2018; L. Zhang, Huang, et al., 2021; Y. Zhang, Wang, et al., 2021). Despite these advancements, the role of groundwater circulation itself as a controlling factor altering the permeability within aquifer-aquitard frameworks remains insufficiently addressed in previous hydrogeological research.

In the face of global climate change and intensifying human activities, the evolving dynamics within aquifer-aquitard systems have emerged as a central concern in hydrogeological research, primarily due to their profound implications for groundwater resource availability. This issue is particularly pressing in regions with dense populations and extensive industrial and agricultural water demands, exemplified by the North China Plain (Yang et al., 2022; C. Zhang et al., 2020a, 2020b; Zhao et al., 2024). Here, the interplay between climatic variability and human exploitation has substantially disrupted regional groundwater circulation patterns, leading to a historic shift from persistent groundwater depletion to marked recovery. However, research on the patterns of permeability changes in aquifer-aquitard systems due to these altered groundwater circulation conditions remains limited. Particularly, the mechanisms by which long-term climate change and human activities influence groundwater circulation processes and subsequently affect the permeability of aquifer-aquitard systems warrant deeper exploration.

Addressing this research gap, our study utilizes extensive, high-resolution groundwater-level data sets from the Tangshan region. Employing tidal response analysis, we quantitatively investigate how the permeability of the aquifer-aquitard system evolves during recovery of the groundwater level. We analyze the relationship between groundwater level recovery and permeability changes, and explore the underlying mechanisms. Our results show that changes in pore pressure can alter the microstructure of the aquifer-aquitard system, thereby affecting the macroscopic permeability of the system. Our investigation thus sheds light on the physical mechanisms linking pore-pressure variations to permeability dynamics, offering enhanced comprehension of groundwater circulation responses under combined climatic and anthropogenic influences. Additionally, the insights generated from our work present potential guidance for informed and sustainable management of groundwater resources.

2 Observation

The study area is located in the Tangshan area, within the northeastern part of the North China Plain. Geographically, it spans from 38°55′N to 40°28′N latitude and from 117°31′E to 119°19′E longitude. The region is part of the Yanshan foreland alluvial plain, characterized by a flat terrain with elevations typically below 50 m. Shallow groundwater is primarily found in porous aquifers within unconsolidated alluvial deposits, including clays, as shown in Figure 1, while deep groundwater is mainly stored within the fracture network of bedrock formations, primarily Ordovician limestone as penetrated by the wells.

The water level data utilized in this study encompass both high-frequency and low-frequency sampling data, sourced from two distinct institutions. To effectively monitor seismic precursory information, the Tangshan Seismological Bureau installed two observation wells, TS05 and MJGK wells, in the Tangshan area. These wells penetrate Ordovician limestone aquifers at depths of approximately 150–240 m and 736–920 m, respectively. Both wells are screened over the full aquifer intervals to capture groundwater flow in fractures. Since 2015, these wells have been equipped with automated instruments for continuous water level monitoring, featuring a sampling interval of 1 min and a resolution of 1 mm. The high-frequency water level data obtained are primarily used to study changes in the permeability of the aquifer system. Additionally, to meet the need for continuous monitoring of groundwater resources, the China Geological Survey established a groundwater level monitoring network covering the Tangshan area, comprising 40 confined monitoring wells (well distributions as shown in Figure 1). These wells monitor confined aquifers at depths primarily ranging from 190 to 250 m, with the aquifer lithology also being Ordovician limestone. The wells are equipped with automated water level monitoring instruments for continuous observation, with a sampling frequency of once per month. The collected well water level data are primarily used to investigate the spatiotemporal evolution of groundwater levels in the Tangshan area.

3 Theoretical Model

According to this model, several factors influence the tidal response of well water levels, including: well characteristics, such as the borehole diameter; hydrogeological parameters of the aquifer-aquitard system, including the storativity ($$S$$), transmissivity ($$T$$), and leakage ($$u$$); geological structure, such as the thickness of the aquifer and aquitard. These factors collectively determine how the water level responds to tidal forces, providing a comprehensive framework for understanding and modeling the interaction between the groundwater level and the Earth tide. Considering that confined and unconfined aquifers are special cases of leaky aquifers, corresponding respectively to $$u\to 0$$ or $$u\mathit{\ll }T$$, and $$u\to \infty $$ or $$u\mathit{\gg }T$$, C. Y. Wang et al. (2018)'s model is applicable not only to the quantitative study of tidal responses of leaky aquifers, but also to that of unconfined and confined aquifers.

4 Results and Analysis

We utilized the widely adopted Baytap-G (Tamura et al., 1991) software for tidal analysis (for detailed information, see Text S2 in Supporting Information S1). The primary parameters for the tidal analysis were configured as follows: a span of 720 hr and a shift of 24 hr. Through this analysis, we extracted the amplitudes and phases of various tidal constituents in the well water levels, including the principal components such as M₂ and S₂ waves. In the tidal components of well water level, the M₂ wave is primarily influenced by the Earth tide, while the S₂ wave is mainly affected by the barometric tide (Hsieh et al., 1987; Rojstaczer, 1988; C. Y. Wang et al., 2018). Given that the M₂ wave is the predominant tidal constituent influenced by solid Earth tides and exhibits the highest signal-to-noise ratio among the tidal components, we employed the amplitude and phase of the M₂ wave to invert for the permeability of the aquifer-aquitard systems and for subsequent analysis of permeability evolution.

Using the tidal response model (Equations 1 and 2), we inverted the amplitude and phase for the M₂ wave (see Figure 3a) to obtain the transmissivity ($$T$$), leakage ($$u$$), and storativity ($$S$$) of the aquifer (see Figure 3b). Considering the quantitative relationships between the aquifer's transmissivity and leakage and their respective hydraulic conductivities ($$T=Kb$$, $$u={K}^{\prime }/{b}^{\mathit{\prime }}$$, where $$K$$ and $$b$$ are the horizontal hydraulic conductivity and the thicknesses of the aquifers, respectively; $${K}^{\mathit{\prime }}$$ and $${b}^{\mathit{\prime }}$$ are the vertical hydraulic conductivity and thickness of the aquitard, respectively), Figure 3b reflects the temporal changes in the permeability of the aquifer-aquitard system. The horizontal permeability of the aquifer and the vertical permeability of the aquitard both showed a trend of increasing from 2019 or 2020. From comparing Figure 4a–4d, it is evident that the trends of leakage and transmissivity change almost completely synchronously after 2019 or 2020, consistent with the trend changes in the well water level and its tidal responses. Additionally, the leakage and transmissivity exhibited coseismic changes following both distant and nearby earthquakes (see the selected earthquake catalog in Table S1 in Supporting Information S1), followed by a gradual recovery to pre-seismic levels within typically less than 6 months. Critically, the storativity remains largely constant, which rules it out as a driver of the observed tidal response changes. This strengthens the conclusion that the variations are predominantly caused by the changes in permeability (or transmissivity and leakage) of the aquifer-aquitard system.

5 Discussion

The observed trends in permeability changes align with the trends in groundwater level fluctuations. As shown in Figure 4, there is a strong correlation between the aquifer's hydraulic conductivity and the well water level. This correlation indicates that the permeability of the aquifer and aquitard is influenced by groundwater head or pore pressure. The aquifer is overlain by a continuous aquiclude or aquitard, causing the groundwater within both the aquifer and aquitard to experience significant hydrostatic pressure proportional to the height of the groundwater level. Consequently, an increase in groundwater level results in heightened hydrostatic pressure within the aquifer and aquitard, enhancing the permeability of the aquifer-aquitard system.

This phenomenon—that changes in pore pressure affect rock permeability—has been consistently observed in laboratory studies (e.g., Li et al., 2023; Palmer & Mansoori, 1998; B. J. Sun et al., 2023; Walsh, 1981; G. Wang, Wang, et al., 2025; Z. C. Wang, Yi, et al., 2025; Xiao et al., 2022; L. Zhang, Huang, et al., 2021; Y. Zhang, Wang, et al., 2021). Experimental results report a wide spectrum of permeability variations, including minor changes of around 5%–10% (G. Wang, Wang, et al., 2025; Z. C. Wang, Yi, et al., 2025), more substantial changes of 80%–90% (L. Zhang, Huang, et al., 2021; Y. Zhang, Wang, et al., 2021), and extreme variations up to 200%–300% (Li et al., 2023). The permeability increase observed in our field study—approximately 34%–89% (see Figure 4)—fall within this range and align with laboratory findings, confirming that the magnitude of permeability change driven by pore pressure under natural conditions is consistent with experimental predictions. The agreement between the magnitude of our field observations and controlled laboratory experiments suggests that the same fundamental mechanical processes (e.g., fracture dilation) governing permeability changes at the core scale are also operative at the formation scale in the field.

The mechanisms responsible for enhancing the permeability of aquifer-aquitard systems mainly consist of fracture unclogging (Elkhoury et al., 2006, 2011; Liu & Manga, 2009; Y. Shi et al., 2019), increased fracture aperture (Allègre et al., 2016), and fracture extension or the formation of new fractures (Liao et al., 2015; C. Y. Wang et al., 2016, 2018). These mechanisms are relevant for explaining permeability changes in aquifer-aquitard systems induced by both earthquakes and hydrological cycles (e.g., Liao & Wang, 2018; Liao et al., 2022, 2025; X.-L. Sun & Xiang, 2020; X. Sun & Xiang, 2021; C. Y. Wang et al., 2018). In this study, we posit that the variations in fracture aperture caused by poroelastic response under varying hydrostatic pressures or groundwater heads are the dominant factor influencing the permeability changes observed in the two wells monitored in the Tangshan area.

Firstly, the trend-driven changes in permeability parameters significantly surpass those triggered by earthquakes. The majority of earthquake-induced increase in permeability are below 5%, while the trend-driven changes exceed 30%. Given that earthquake-related impacts on permeability are predominantly linked to fracture unclogging, the scale of the trend-driven changes is too substantial to be accounted for by fracture unclogging alone. This implies that the trend-driven variations in permeability are not contingent on fracture unclogging.

Secondly, the change in well water level ($${\Delta }h$$) is relatively small—approximately 20 m of water column height—which translates to an effective strain value of $$e=\rho g{\Delta }h/{BK}_{u}\sim {10}^{-5}$$, where $$\rho {=10}^{3}\,{\text{kg}/\mathrm{m}}^{3}$$ is the density of groundwater, $$g=9.8\,{\mathrm{m}/\mathrm{s}}^{2}$$ is the acceleration due to gravity, B is the Skempton coefficient of the aquifer or aquitard, and $${K}_{u}$$ is the undrained bulk modulus of the aquifer or aquitard, $${BK}_{u}$$ is typically approximately 10 GPa for limestone (H. F. Wang, 2000). This magnitude of strain is insufficient to cause the extension of existing fractures or the formation of new fractures within the aquifer and/or aquitard.

Furthermore, the water level-permeability relationships illustrated in Figures 3 and 4 exhibit negligible hysteresis. This indicates that the hydraulic responses are predominantly poroelastic, with minimal contribution from non-elastic processes such as irreversible fracture propagation or clogging. The linear correlation further supports that permeability changes are largely recoverable and governed by reversible fracture aperture adjustments under varying hydrostatic pressures or groundwater heads.

Considering the above analysis, we can rule out fracture unclogging, fracture extension, and new fracture formation as significant mechanisms influencing permeability changes. Instead, the changes in permeability are primarily attributed to variations in fracture aperture caused by pore pressure changes or deformation of the rock matrix, as well as potential deformation of clogging materials. For open fractures (such as shear or tensile fractures), increased pore pressure enlarges the fracture aperture, enhancing hydraulic conductivity. This can also open previously closed or partially closed micro-fractures, thereby increasing the permeability of the aquifer-aquitard system. For filled fractures, increased water pressure can compress the filling materials (e.g., clay minerals or debris), increasing the flow paths and area available for groundwater movement, which enhances the permeability of existing fractures. Both mechanisms likely coexist and contribute to the observed increase in permeability.

Since 2019, the groundwater levels near the two high-frequency observation wells in the Tangshan area have shown a trend of stabilization and recovery, followed by a sustained upward trend (see Figure 5). This can be attributed to two main factors: firstly, the annual rainfall in the Tangshan area has increased yearly since 2019 (see Figure 2); secondly, the local government has actively implemented water-saving policies and measures to effectively manage and conserve groundwater resources (refer to: Hebei News Network, 2023; https://ts.hebei.com.cn and Huanbohai News Network, 2023; https://www.huanbohainews.com.cn). We believe that the significant changes in the groundwater circulation within the study area, driven by both climate change (increased rainfall) and human activities (restricted extraction), have led to a substantial increase in groundwater storage. Consequently, the pore pressure in aquifers and aquitards has continuously increased, thereby enhancing the permeability of the aquifer-aquitard system.

The observed changes in permeability during the rise in groundwater levels, driven by both natural hydrological processes and human activities, highlight the need to consider the dynamic nature of permeability parameters when studying the aquifer-aquitard systems and related groundwater resource, environmental, and hazard issues. Furthermore, this study suggests that permeability changes in aquifer-aquitard systems due to rising groundwater levels in the North China Plain may be a widespread phenomenon influenced by climate change and groundwater management, warranting further comprehensive investigation.

6 Conclusion

This study reveals that groundwater level recovery in the North China Plain has led to a substantial increase in the permeability of aquifer-aquitard systems. Tidal response analysis of high-frequency well records shows that both aquifer transmissivity and aquitard leakage have increased since 2019, in parallel with rising groundwater levels. These permeability changes are primarily driven by fracture aperture expansion caused by elevated pore pressure, rather than by seismic events or fracture unclogging. This confirms a direct causal relationship between pore pressure changes and permeability dynamics, demonstrating that permeability adjusts in response to pore pressure fluctuations driven by groundwater level changes. Furthermore, this implies that the bedrock aquifer-aquitard systems are intricate systems with fluid-solid coupling characteristics, influenced by both natural hydrological processes and human activities. These findings underscore that permeability is not static, but dynamically regulated by groundwater conditions—calling for its explicit treatment as a variable in hydrogeologic models. This has broad implications for managing groundwater resources and assessing crustal fluid pathways in regions undergoing recharge from climate variability or policy interventions.

Conflict of Interest

The authors declare no conflicts of interest relevant to this study.