Geostatistical and geochemical analysis of surface water leakage into groundwater on a regional scale: a case study in the Liulin karst system, northwestern China

https://doi.org/10.1016/S0022-1694(01)00376-6Get rights and content

Abstract

The Liulin karst system is typical of hydrogeological systems in northern China, with a group of springs as the dominant way of regional groundwater discharge. Surface water leakage into groundwater has been observed in six sections of the rivers in the study area. To extract hydrogeological information from hydrochemical data, 29 water samples were collected from the system. On a trilinear diagram, most of the groundwater samples are clustered around the surface waters, indicating the effect of leakage on their chemistry. R-mode factor analysis was made on seven variables (Na, Ca, Mg, SO4, Cl, HCO3, and NO3) of the samples and three principal factors were obtained: the F1 factor is composed of Ca, Mg and SO4 , the F2 of HCO3 and NO3, and the F3 of Na and Cl. These factors are then used as regionalized variables in ordinary Kriging for unbiased estimates of the spatial variations of their scores. Considering regional hydrogeological conditions, the hydrogeological implications of the spatial distribution of the factor scores as related to the effects of the surface leakage are discussed. To evaluate the geochemical processes, the geochemical modeling code NETPATH was employed. The modeling results show that mixing commonly occurs in the system and dolomite dissolution is more important than calcite dissolution. Dedolomitization (calcite precipitation and dolomite dissolution driven by anhydrite dissolution) is locally important, in the western flank of the system where the surface water leakage has the least effect.

Introduction

Surface–subsurface water interaction plays an important role in controlling groundwater chemistry and dynamics in karst terrains. Understanding these processes is the basis for sustainable development, management and protection of water resources.

Losses of streamflow into groundwater systems are major processes of such interactions. Under natural conditions, there are basically three pathways of the losses. The first and the second refer to the transformation of surface runoff into subsurface either via swallow holes at the margin of the karst, or via sinkholes as internal flow (White, 1988). The third is the leakage through carbonate bedrock beneath the river bed, as often observed in the 19 major karst groundwater systems in Shanxi Province, northwestern China (including Liulin for the present study), which are the most important source of water supply for the province (Han et al., 1993). Such leakage problems are difficult to be quantitatively studied, not only because of the extreme complexity of the dissolution cavities, but because of the regional scale at which the leakage takes place.

Artificial tracer methods using purposefully drilled wells have been widely applied to detect the bypass routes of reservoir water around dams built in carbonate terrain (Silar, 1988). These methods are unsuitable for regions of several thousand square kilometers, where groundwater flows in anisotropic and heterogeneous aquifers, as commonly observed in the karst systems in Shanxi. As an alternative approach to this problem, hydrochemistry of groundwaters as a natural tracer to demonstrate the leakage process is applied by us and discussed in the present paper.

As a major diagnostic tool in groundwater hydrology, hydrogeochemical data have been used to identify recharge zones and flow patterns, calculate recharge rates or mixing ratios, and to discern hydraulic connections between aquifers (Hem, 1989, Mayo et al., 1992, Mazor et al., 1993, Panno et al., 1994, Wang and Khaustov, 1997). Geostatistical analysis of geochemical data can often give some insights into the underlying factors controlling hydrogeological processes. For instance, Kriging has been found to be especially useful for analyzing regional scale hydrochemical data (Goovaerts et al., 1993).

The objectives of the present study are: (1) to characterize the hydrogeochemical features of the Liulin karst groundwater system as a typical case in northern China; and (2) to show the effectiveness of combining geostatistical and geochemical analysis techniques to extract hydrological information about the leakage processes from hydrogeochemical data.

Section snippets

Geostatistical methods

Factorial analysis and ordinary Kriging are combined in this study to extract hydrogeologic information related to surface runoff leakage in Liulin. Factorial analysis has been widely used in geoscience studies. Also, as one of the most widely used geostatistical methods in the hydrology community in recent years, Kriging is a powerful interpolating approach in unbiased estimation of the field variables at a given point (Rizzo and Dougherty, 1994). The concept of a regionalized variable is the

Hydrogeology of the Liulin karst system

Located in a semi-arid region, the Liulin karst system is hydrogeologically typical in northern China: a group of springs are the predominant way of regional groundwater discharge. Along a distance of less than 2 km of the Sanchuan River valley outcrop more than 100 springs with a total discharge of 2.3–6.1 m3/s. These springs outcrop in five groups and are collectively called the Liulin springs (Fig. 1).

The northern and eastern boundaries of the system are composed of Archaen metamorphic rocks,

River water leakage

Flow meter measurements over years have shown surface runoff leakage along six sections of the four rivers (Fig. 2 and Table 1). The largest leakage was observed at No. 3 leakage section along the Dadongchuan River, with 55–92% of the river water leaking via fissures and pores in carbonate rocks of the riverbed.

Hydrochemistry

Locations of the 29 samples are shown in Fig. 2. The major ion concentrations of the 29 samples are listed in Table 2. The hydrochemical properties and trace element contents of the 16

Conclusions

The results obtained in the present study show that coupling geostatistical techniques with hydrogeochemical analysis is an effective approach in regional karst hydrology studies. The major conclusions of the study are as follows:

  • 1.

    Surface water leakage into the Ordovician carbonate aquifers occurs in most parts of the Liulin karst groundwater system. In places such as No. 3 leakage section along the Dadongchuan River, as much as 55–92% (up to 0.235 m3/s) of the surface runoff leaked into the

Acknowledgements

The research was financially supported by National Natural Science Foundation of China (No.49832005) and the Ministry of Science and Technology of China (Grant No.95-pre-39). Dr L. Niel Plummer of USGS is appreciated for sending us the NETPATH code. Dr Eric J. Reardon of University of Waterloo is acknowledged for giving valuable supervision and providing computing facilities when Yanxin Wang stayed in Waterloo as a visiting scientist and prepared the manuscript in 1998–1999. The manuscript

References (17)

  • I.I. Chebotarev

    Metamorphism of natural waters in the crust of weathering

    Geochimica et Cosmochimica Acta

    (1955)
  • P. Goovaerts

    Factorial Kriging analysis: A useful tool for exploring the structure of multivariate spatial soil information.

    J. Soil Sci.

    (1992)
  • P. Goovaerts

    Geostatistics for Natural Resource Evaluation

    (1997)
  • P. Goovaerts et al.

    Factorial Kriging analysis of springwater contents in the Dyle River Basin, Belgium

    Water Resour. Res.

    (1993)
  • X. Han et al.

    Karst groundwater systems: case studies on karst springs in Shanxi

    (1993)
  • Hem, J.D., 1989. Study and interpretation of the chemical characteristics of natural waters. USGS Water Supply Paper...
  • A.L. Mayo et al.

    The use of solute and isotopic chemistry to identify flow patterns and factors which limit acid mine drainage in the Wasatch Range, Utah

    Ground Water

    (1992)
  • E. Mazor et al.

    Hydrochemical implications of groundwater mixing: an example from the southern Laramie Basin, Wyoming

    Water Resour. Res.

    (1993)
There are more references available in the full text version of this article.

Cited by (123)

  • Hydro-biogeochemical processes of surface water leakage into groundwater in large scale karst water system: A case study at Jinci, northern China

    2021, Journal of Hydrology
    Citation Excerpt :

    These five samples are located along the Fenhe River channel, where widely distributed soluble carbonates of middle Ordovician are exposed to the surface, thereby providing favorable conditions for surface downward leakage. If no mixing happens to these karst groundwater samples along its flow path, their sodium and chloride concentrations should be between those in groundwater from the recharge and discharge areas, following a normal evolution trends of major ion chemistry under isolated hydrogeological conditions (Chebotarev, 1955; Wang et al., 2001). The most reasonable explanation for the enrichment of sodium and chloride as well as salinity contents (Figs. 6 and 7(a–f)) is that these karst groundwater are inordinately recharged by Fenhe River waters via faulting structures and/or river channel leakage.

  • Natural and anthropogenic factors driving groundwater resources salinization for agriculture use in the Campania plains (Southern Italy)

    2021, Science of the Total Environment
    Citation Excerpt :

    In this study the factor scores belonging to each groundwater monitoring wells have been considered as input variables to interpolate and consequently display the range and the degree of groundwater salinization influenced by the common factors. This procedure has been commonly worldwide utilized with excellent results (Busico et al., 2018, 2020; Shyu et al., 2011; Wang et al., 2001). The F1 explains the 34% of the total variance and shows a positive correlation between Na+, SAR, Cl−, B+ and EC highlighting the presence of a persistent salinization probably due to phenomena of actual SWI, and to the water rock interaction with shallow marine sediments and reworked volcanic materials (Busico et al., 2018).

View all citing articles on Scopus
View full text