Research papersAssessing the impact of tunnelling on karst groundwater balance by using lumped parameter models
Introduction
Karstic aquifers are characteristic of complex flow geometry and strong hydrogeological heterogeneity, caused mainly by double or triple porosity structures consisting of pores, fissures, fractures and conduits of various scales (White, 2002, Ford and Williams, 2007, Hartmann et al., 2014). Consequently, karst groundwater typically consists of base flow through fissure network and rapid flow through enlarged fractures and conduits (Jukić and Denić-Jukić, 2009, Hosseini et al., 2017). These characteristics give rise to significant challenges in groundwater flow modelling in karstic formations. Further complexity arises when anthropogenic activities, e.g., tunnel construction, are involved, where the groundwater flow paths and discharges may be changed drastically within a short period (Hao et al., 2016, Lv et al., 2020), even resulting in geological and environmental hazards such as mud and water inrush, spring depletion, and ground subsidence (Raposo et al., 2010, Pu et al., 2011, Mahler and Bourgeai, 2013, Grimmeisen et al., 2016, Han et al., 2017, Jiang et al., 2018, Liu et al., 2019, Chen et al., 2020, Zhou et al., 2021).
A variety of hydrological models (e.g., black-box, distributed, and lumped) have been proposed to confront the challenges and approximate the groundwater budget changes in karst formations (Beven, 1989, Fleury et al., 2007, Le Moine et al., 2008, Makropoulos et al., 2008, Jukić and Denić-Jukić, 2009, Hartmann et al., 2012). The black-box models treat an aquifer as a black-box system and focus on establishing an overall relationship between input (infiltration) and output (discharge), completely disregarding the structure and the physical mechanisms of the aquifer (Dreiss, 1983, Mangin, 1984). The distributed models solve flow equations with temporal and spatial discretization over the whole aquifer, hence requiring detailed information on the aquifer structure and hydrological properties for producing an acceptable simulation (Zhang et al., 2011, Zheng et al., 2020, Saller et al., 2013). The lumped models provide a compromised tool for karst groundwater simulation in which linear or nonlinear reservoirs/compartments are used to represent different physical elements of the aquifer structure and groundwater circulation (Fleury et al., 2007, Fleury et al., 2009, Jukić and Denić-Jukić, 2009). Each reservoir relies on a production function and a transfer function to simulate the process from input to output in each hydrogeological unit, and hence the lumped models are suitable for aquifer simulation with limited observation data and insufficient field characterization.
Spring discharge observations and trace element data are commonly used to calibrate the lumped models (Wagener et al., 2003, Hartmann et al., 2013, Ye et al., 2014, Mudarra et al., 2019). Particular concerns have been focused on the temporal variability in groundwater recharges and the effect of karst aquifers drainage induced by climate change or human activities (Jukić and Denić-Jukić, 2009, Chang et al., 2017, Hosseini et al., 2017). The construction of deep-buried tunnels in karst regions can profoundly change the water budget in the karst system. Hence there is a need to revisit the structure and parameters of the lumped model, for incorporating this strong impact of human activity.
This study is motivated by tunnel excavation-induced karst groundwater budget changes in the Jiayan Water Diversion Project (JWDP), located in Guizhou Province, Southwest China. In this work, a modified lumped model is presented to quantitatively estimate the impact of tunnel construction on the karst aquifers, by integrating a channel drainage compartment that represents the drainage from the karst aquifer to the tunnel during construction. The model structure is systematically investigated through comparison of simulations with varying levels of simplification by merging or removing some of the flow components. The dominant hydrogeological processes that control the aquifer discharge are identified by performance assessment of the models and sensitivity analysis of the input parameters. The impact of tunnel construction on the aquifer discharge is finally quantified with the calibrated models, which contributes to decision-making for appropriate engineering treatments during tunnel construction.
Section snippets
Study site and data
The study area is located in Guizhou Province, Southwest China (Fig. S1a in Supplementary material), where karst landscapes cover about 62% of the land area (Chen et al., 2020); it is bounded by the Baifu River in the east, the Liuchong River in the south, and a groundwater divide in the north. The terrains range from 1300 to 2300 m a.s.l., sloping from northwest to southeast. The study area lies in the subtropical monsoon climate zone with a marked seasonal pattern in the annual variations of
Model development
As depicted conceptually in Fig. 1, in the Shuchang HU, the rainfall infiltrates into the karst system by both diffuse flow through soil cover and quick flow through sinkholes and vertical shafts, where the recharge area often varies as the groundwater level fluctuates (White, 2002, Ford and Williams, 2007). The quick flow directly feeds into the underground river, while the diffuse flow recharges the epikarst zone characteristic of relatively lower degree of karstification at the site (
Model structures and performance
Both four- and three-compartment models (MLM1–8) are used to simulate the groundwater response in the Shuchang hydrogeological unit. The SCEM-UA method with the NSE criterion is used to calibrate the parameters of each model, after a warm-up period of one month to eliminate the influence of initial conditions on the result. As an example, Fig. 4 presents the box-whisker plot of the calibrated parameters for the model MLM1 having the most complex structure. Fig. 5, Fig. 6 plot the spring
Discussion
In Section 4, we have identified the dominant hydrogeological processes in the Shuchang HU, by comparing the performance of a series of models with different degree of simplifications and by screening analysis of parameter sensitivities. We find that the four-compartment models consistently perform better than the three-compartment counterparts. But as a result of the thin thickness of the soil cover, the calibrated soil storage capacity Vsmax is around 2 mm (MLM1), and is less sensitive to the
Conclusions
This study presents a series of modified lumped models with different structures or flow components to characterize the mechanisms and relationships between recharge and discharge in karst aquifers. A particular concern is placed upon the impact of tunneling activities on the aquifer discharge, by integrating a channel drainage compartment in the models. The lumped models are applied to an independent hydrogeological unit (HU) where a deep-buried tunnel is being constructed. By comparing the
CRediT authorship contribution statement
Bing-Qi Zhou: Investigation, Methodology, Formal analysis, Validation, Writing - original draft. Zhibing Yang: Investigation, Methodology, Formal analysis, Validation, Writing - review & editing. Ran Hu: Formal analysis, Validation. Xian-Jin Zhao: Investigation, Methodology. Yi-Feng Chen: Conceptualization, Supervision, Formal analysis, Resources, Funding acquisition, Writing - original draft, Writing - review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
This work was financially supported by the National Natural Science Foundation of China (Grant No. 51925906), the National Key R&D Program of China (Grant No. 2018YFC0407001), and the Guizhou Province Key Science and Technology Project (Grant No. [2017]3005-1).
References (49)
Water exchange and pressure transfer between conduits and matrix and their influence on hydrodynamics of two karst aquifers with sinking streams
J. Hydrol.
(2010)Changing ideas in hydrology – The case of physically-based models
J. Hydrol.
(1989)- et al.
Identification of the dominant hydrological process and appropriate model structure of a karst catchment through stepwise simplification of a complex conceptual model
J. Hydrol.
(2017) - et al.
Conceptual hydrogeological model of flow and transport of dissolved organic carbon in a small Jura karst system
J. Hydrol.
(2012) - et al.
Non-Darcian flow effect on discharge into a tunnel in karst aquifers
J. Rock Mech. Min. Sci.
(2020) Linear unit-response functions as indicators of recharge areas for large karst springs
J. Hydrol.
(1983)- et al.
Structure and water storage capacity of a small karst aquifer based on stream discharge in southwest China
J. Hydrol.
(2016) - et al.
Modelling of the functioning of karst aquifers with a reservoir model: application to Fontaine de Vaucluse (South of France)
J. Hydrol.
(2007) - et al.
Modelling the hydrologic functions of a karst aquifer under active water management–the Lez spring
J. Hydrol.
(2009) - et al.
Weekly variations of discharge and groundwater quality caused by intermittent water supply in an urbanized karst catchment
J. Hydrol.
(2016)