Using cluster analysis for understanding spatial and temporal patterns and controlling factors of groundwater geochemistry in a regional aquifer

doi:10.1016/j.jhydrol.2020.124594

Journal of Hydrology

Volume 583, April 2020, 124594

https://doi.org/10.1016/j.jhydrol.2020.124594 Get rights and content

Highlights

•
Analyzed groundwater geochemistry at the regional scale over a long time period.
•
Analyzed cluster snapshots for spatio-temporal patterns of groundwater quality.
•
Identified controlling factors of spatio-temporal patterns of groundwater quality.
•
Delineated four zones of groundwater geochemistry along regional flow direction.
•
Revealed impacts of Three Gorges Reservoir on regional groundwater geochemistry.

Abstract

Understanding spatial and temporal patterns of groundwater geochemistry at the regional scale over a long time period is challenging, due to the lack of data and effective statistical approaches to characterize complex natural processes and anthropogenic activities. We applied a recently developed cluster analysis method to investigate spatial and temporal patterns and controlling factors of groundwater geochemistry in the confined aquifer of the Jianghan Plain, China. The cluster analysis is applied to a dataset of 13,024 groundwater geochemical measurements for 11 geochemical parameters of 1,184 groundwater samples collected over 23 years from 29 monitoring wells distributed over the Jianghan Plain. The cluster analysis yielded a classification of seven clusters, and the classification was confirmed by using principal component analysis and Stiff and Piper diagrams. Based on the spatial distribution of the seven clusters, the Jianghan Plain is separated into four geochemical zones (i.e., recharge zone, transition zone, flow-through zone, and discharge-mixing zone) along the regional groundwater flow path, which has not been attempted in the past. The temporal changes of groundwater geochemistry are controlled by short- and long-term factors of water-rock interactions and anthropogenic activities. A particular finding of this study is that the Three Gorges Reservoir has a long-term impact on groundwater geochemistry, because the reservoir increased river discharge to groundwater after 2009. This study demonstrates that using the cluster analysis method together with hydrogeochemical analysis can identify spatiotemporal patterns and controlling factors of groundwater geochemistry at the regional scale over a long monitoring period.

Graphical abstract

Introduction

A thorough understanding of groundwater geochemistry is critical for protecting groundwater resources under the conditions of changing climate, growing population, and decreasing freshwater availability (Cloutier et al., 2008, Fendorf et al., 2010, Gorelick and Zheng, 2015, Han et al., 2016, Landon et al., 2011). Groundwater geochemistry is driven by various natural processes and anthropogenic activities (Güler and Thyne, 2004a, Güler et al., 2002, Reghunath et al., 2002, Tóth, 1999, Tóth, 2009), and may vary both in space and time. Understanding spatial and temporal patterns of groundwater geochemistry not only requires designing a monitoring network to collect the right data (in terms of data constitutes, collection locations, collection times, etc.) for tackling the problem of interest, but also requires using appropriate statistical methods to extract the spatial and temporal patterns embedded in measurements of groundwater geochemistry parameters. More importantly, hydrogeochemical analysis is needed to understand the natural and anthropogenic factors that control the spatial and temporal patterns of groundwater geochemistry. While obtaining a large amount of monitoring data for a region-scale study is not uncommon (e.g., Feyereisen et al., 2007, Güler and Thyne, 2004b, Jessen et al., 2017), it is still challenging to reveal the spatial and temporal patterns hidden in the data and to understand the controlling factors of the patterns at the regional scale. This study addresses these two challenges by using a recently developed method of cluster analysis together with hydrogeochemical analysis for a regional aquifer in central China.

Using cluster analysis together with hydrogeochemical analysis has advanced our understanding on spatial and temporal patterns of groundwater geochemistry (Cloutier et al., 2008, Kim et al., 2003, Nguyen et al., 2015, Shrestha and Kazama, 2007, Simeonov et al., 2003, Wang et al., 2015). This is always done in two steps. In the first step, cluster analysis is conducted to classify groundwater geochemical data into a number of clusters, each of which reflects its own composition of groundwater geochemistry. Subsequently, hydrogeochemical analysis is conducted for the clusters to investigate spatial and temporal patterns in groundwater geochemistry. When dealing with spatiotemporal data obtained from a long-term monitoring network, many clustering methods have two limitations on how clusters are classified. One limitation is that cluster classification is conducted only for temporal means (i.e., the means over the entire sampling period) (e.g., Qian et al., 2007, Sayemuzzaman et al., 2018). While this kind of cluster analysis can help identify spatial patterns, it cannot be used to understand temporal patterns. The other limitation is that cluster classification is conducted separately to data of different sampling times or hydrological conditions (e.g., Hussain et al., 2008, Thyne et al., 2004). This kind of cluster analysis may be inadequate to simultaneously reveal spatial and temporal patterns when groundwater geochemistry substantially changes over time, because the number of clusters and the geochemical characteristics of the clusters can be dramatically different before and after the changes. Addressing the two limitations requires a cluster analysis method that deals with data in space and time simultaneously and can reveal spatial and temporal patterns.

Pacheco-Castro et al. (2018) recently developed a cluster classification method that has the potential to overcome the two limitations discussed above. Instead of conducting cluster classification for temporal means or data of different sampling times, the cluster analysis of Pacheco-Castro et al. (2018) is conducted for the entire monitoring data collected at all monitoring wells and sampling times. If groundwater samples are collected from m monitoring wells at s sampling times and p geochemical parameters are measured for each of m × s groundwater samples, the measurements form a matrix that has m × s rows and p columns. The cluster classification method of Pacheco-Castro et al. (2018) yields a column vector of m × s elements of clustering index for the combination of sampling wells and sampling times. This vector leads to s snapshots of clusters in space. Since the cluster classification is conducted for all monitoring data, the number of clusters and their geochemical characteristics are the same for each snapshot. This makes it possible to jointly study the spatial and temporal patterns of groundwater geochemistry. Linking the snapshots to geological, hydrogeological, and geochemical knowledge of the site of interest can help understand the controlling factors of groundwater geochemistry.

This can be illustrated in the following two scenarios:

(1)
If the natural and anthropogenic factors controlling groundwater geochemistry are similar at several monitoring wells, the geochemistry at the wells should be similar, and the wells should belong to the same cluster. Studying the spatial distribution of the clusters can reveal large-scale patterns of groundwater geochemistry, which in turn helps identify the controlling factors of groundwater geochemistry in space.
(2)
If groundwater geochemistry changes at the wells, the wells may belong to another cluster. Therefore, studying the temporal changes of the clusters can reveal temporal patterns of groundwater geochemistry, which in turn helps identify the factors that control groundwater geochemistry in time.

Pacheco-Castro et al. (2018) applied their cluster classification method to 288 groundwater samples collected over three years (2009–2011) from a karst aquifer in in Yucatan, Mexico. They found the following: (1) groundwater geochemistry at the west and in the coastal area of the site is controlled by seawater intrusion and sulfate-rich groundwater, (2) groundwater geochemistry at the middle and east part of the site is controlled by water-rock interactions and annual precipitation, respectively, and (3) groundwater geochemistry at two local areas of the site is controlled by anthropogenic activities. They also found that temporal variation of groundwater geochemistry at the site is caused by groundwater dilution due to changes in the amount and spatial distribution of precipitation. Since the cluster analysis of Pacheco-Castro et al. (2018) was for a small dataset collected over a short monitoring period, it is necessary to further evaluate their method for a large dataset over a long monitoring period.

The evaluation was conducted in this study by using a large amount of monitoring data collected from the confined aquifer of Jianghan Plain, an alluvial plain located in central China with an area of approximately 27,400 km² (Fig. 1). The Yangtze River, the world’s third longest river, flows through the plain, and the Three Gorges Dam, the world’s largest hydroelectric project, is about 80 km upstream of the plain (Fig. 1). Groundwater is mainly of Ca(Mg)-HCO₃ water type due to dissolution of carbonate minerals (Gan et al., 2014, Yu et al., 2017, Zhou et al., 2012). This area is undergoing extensive agricultural activities, and 72% of the land is for agricultural use (Fig. 1). Anthropogenic activities have affected groundwater geochemistry of the regional aquifer (Yang et al., 2018). To monitor groundwater quality, a monitoring network was established in 1990, and a large amount of monitoring data have been collected during the past three decades for a total of 21 groundwater geochemistry parameters specified by the National Quality Standard for Ground Water of China (Ministry of Environmental Protection of the People's Republic of China, 1994). The 21 parameters are pH, temperature, alkalinity, total dissolved solids (TDS), hardness, concentrations of seven major ions (Ca²⁺, Mg²⁺, K⁺, Na⁺, Cl^-, SO₄^2-, and HCO₃^–), and concentrations of minor ions and trace constituents (NO₂^–, NO₃^–, NH₄⁺, Fe³⁺, Fe²⁺, Fe, Mn²⁺, F^-, and total As). Groundwater contaminants reported in literature include arsenic (Duan et al., 2015, Duan et al., 2017, Gan et al., 2014) and wastewater infiltration (Niu et al., 2017) in the confined aquifer and nitrate (Yang et al., 2017, Yang et al., 2018) in the shallow aquifer of Jianghan Plain. However, there are few published analyses of groundwater geochemistry at the plain scale, except the paper of Niu et al. (2017), who reported the temporal trends of multiple groundwater geochemistry (e.g., Cl^-, SO₄^2-, and NO₃^–) but did not analyze spatial patterns of groundwater geochemistry. The overall spatial and temporal patterns of groundwater geochemistry and their controlling factors at the plain scale are still unknown.

This study used the groundwater geochemistry data collected over the period of 1992–2014. The dataset consists of a total of 13,024 geochemistry measurements for 1,184 groundwater samples, each of which has 11 selected groundwater geochemistry parameters (pH, Ca²⁺, Mg²⁺, K⁺, Na⁺, Cl^-, SO₄^2-, HCO₃^–, NH₄⁺, F^-, and Fe). This large dataset enables us to evaluate whether the cluster analysis method of Pacheco-Castro et al. (2018) can help better understand the spatial and temporal patterns of groundwater geochemistry and their controlling factors for a regional aquifer over a long period of 23 years. Based on the cluster analysis and hydrogeochemical analysis, we delineated for the first time four zones of groundwater geochemistry at the site which are related to aquifer recharge, regional groundwater flow, water-rock interactions, and anthropogenic activities. The four zones should be of value for future study of detailed groundwater geochemistry at the local scale. We also identified the long-term impacts of the Three Gorges Reservoir on groundwater geochemistry downstream of the reservoir. While the impacts of dam construction and operation on river hydrology have been well documented (Guo et al., 2018, Nilsson et al., 2005, Zhou et al., 2013), little is known about the impacts on groundwater geochemistry. This study provides a new insight on using cluster analysis for understanding long-term evolution of groundwater geochemistry at the regional scale.

Section snippets

Study area

Jianghan Plain is a semi-closed basin with high elevation (~350 m) in the northwest and low elevation (~25 m) in the southeast portion of the plain, and the regional groundwater flow is from northwest to southeast. Based on precipitation data from the Jingzhou meteorological station (provided by the China Meteorological Data Service Center, available at http://data.cma.cn/), monthly average precipitation for the period of 1992–2014 ranges from 24 mm in December to 161 mm in July, with 41% of

Data and statistical procedures

Groundwater samples were collected from 29 monitoring wells (Fig. 1) twice a year during 1992–2014, one in the dry season from January to February and the other in the wet season from July to August. Since not all the wells were sampled regularly over the 23 years, the total number of samples is 1,190, less than 1,334 = 29 × 23 × 2. For the 21 groundwater geochemistry parameters, 11 parameters were selected for the cluster analysis, and they are pH, Ca²⁺, Mg²⁺, K⁺, Na⁺, Cl^-, SO₄^2-, HCO₃^–, NH₄⁺,

Classified seven clusters

This section presents the results of the hierarchical cluster analysis and the results of statistical analysis and geochemical analysis that were used to evaluate whether the classified seven clusters are reasonable. Fig. 2 shows the dendrogram of the hierarchical cluster classification for the 1,184 groundwater samples, and the phenon line drawn at the linkage distance of 25 led to seven clusters, denoted as Cluster C1–C7. To understand the geochemical characteristics of the seven clusters,

Spatial patterns and controlling factors

The hierarchical cluster analysis gave an array with 1,184 elements of cluster index (C1–C7) for the 1,184 groundwater samples. The cluster indices for each sampling time were extracted and plotted as a snapshot to show their spatial distributions at the well locations. The snapshots are the basis for investigating the spatial and temporal patterns of groundwater geochemistry.

Four zones of groundwater geochemistry

Based on the highest frequency data listed in Table 2 and the understanding of the controlling factors of groundwater geochemistry discussed above, four groundwater geochemical zones were delineated and shown in Fig. 8. The figure also plots spatial distribution of sediment outcrops at land surface, which was also used for delineating the four zones. The delineation of the four zones is a major contribution of this study for understanding groundwater geochemistry at the study site. The zone

Temporal variation and controlling factors

Although the groundwater geochemistry is generally stable over the 23 years, temporal variations are observed by examining the snapshots of the spatial patterns of the clusters (Fig. 6), recalling that changes of clusters indicate changes of groundwater geochemistry. A better way of examining the cluster changes is to plot the temporal variation of cluster index for each well, as shown in Fig. 9. The seven wells were selected because of their low relative frequency (less than 80%) of belonging

Comparison with conventional cluster analysis

The cluster analysis method used in this study enabled us to simultaneously identify the spatial and temporal patterns and controlling factors of the groundwater geochemistry. The simultaneous identification may not be achieved by using conventional methods of cluster analysis. This is illustrated by comparing the results of our cluster analysis with those obtained by using a conventional cluster analysis that uses temporal averages to identify spatial patterns. For each groundwater

Conclusions

A recently developed cluster analysis method of Pacheco-Castro et al. (2018) was used in this study for understanding the spatial and temporal patterns and controlling factors of groundwater geochemistry in the regional aquifer of the Jianghan Plain, China. This cluster analysis method can handle the large dataset with 13,024 measurements of 1,184 groundwater samples collected over a long period of time (1992–2014) from 29 monitoring wells distributed over the Jianghan Plain. The major

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 research was supported by China Geological Survey grants 1212011121142 and 1212011120084, and National Natural Science Foundation of China grant 51629901. The first author was supported by the Fundamental Research Funds for National Universities, China University of Geosciences (Wuhan) for his research at the Florida State University. The second author was supported by National Science Foundation grant EAR-1828827. We thank Barbara Mahler and two anonymous reviewers for their thoughtful

References (67)

V. Cloutier et al.
Multivariate statistical analysis of geochemical data as indicative of the hydrogeochemical evolution of groundwater in a sedimentary rock aquifer system
J. Hydrol.
(2008)
K. Deng et al.
Three Gorges Dam alters the Changjiang (Yangtze) river water cycle in the dry seasons: evidence from H-O isotopes
Sci. Total Environ.
(2016)
Y. Duan et al.
Temporal variation of groundwater level and arsenic concentration at Jianghan Plain, central China
J. Geochem. Explor.
(2015)
Y. Duan et al.
Arsenic speciation in aquifer sediment under varying groundwater regime and redox conditions at Jianghan Plain of Central China
Sci. Total Environ.
(2017)
Y. Gan et al.
Hydrogeochemistry and arsenic contamination of groundwater in the Jianghan Plain, central China
J. Geochem. Explor.
(2014)
O. Ghesquière et al.
Scenarios of groundwater chemical evolution in a region of the Canadian Shield based on multivariate statistical analysis
J. Hydrol.: Reg. Stud.
(2015)
C. Güler et al.
Hydrologic and geologic factors controlling surface and groundwater chemistry in Indian Wells-Owens Valley area, southeastern California, USA
J. Hydrol.
(2004)
L.C. Guo et al.
How have the river discharges and sediment loads changed in the Changjiang River basin downstream of the Three Gorges Dam?
J. Hydrol.
(2018)
D. Han et al.
Deep challenges for China's war on water pollution
Environ. Pollut.
(2016)
M. Hussain et al.
Cluster analysis and quality assessment of logged water at an irrigation project, eastern Saudi Arabia
J. Environ. Manage.
(2008)

C. Liu et al.

Application of factor analysis in the assessment of groundwater quality in a blackfoot disease area in Taiwan

Sci. Total Environ.

(2003)

T.T. Nguyen et al.

Clustering spatio-seasonal hydrogeochemical data using self-organizing maps for groundwater quality assessment in the Red River Delta, Vietnam

J. Hydrol.

(2015)

B. Niu et al.

Temporal variations of groundwater quality in the Western Jianghan Plain, China

Sci. Total Environ.

(2017)

R.R. Pant et al.

Spatiotemporal variations of hydrogeochemistry and its controlling factors in the Gandaki River Basin, Central Himalaya Nepal

Sci. Total Environ.

(2018)

R. Reghunath et al.

The utility of multivariate statistical techniques in hydrogeochemical studies: an example from Karnataka, India

Water Res.

(2002)

M.V. Schaefer et al.

Redox controls on arsenic enrichment and release from aquifer sediments in central Yangtze River Basin

Geochim. Cosmochim. Acta

(2017)

S. Shrestha et al.

Assessment of surface water quality using multivariate statistical techniques: a case study of the Fuji river basin, Japan

Environ. Modell. Soft.

(2007)

V. Simeonov et al.

Assessment of the surface water quality in Northern Greece

Water Res.

(2003)

Q. Tang et al.

Flow regulation manipulates contemporary seasonal sedimentary dynamics in the reservoir fluctuation zone of the Three Gorges Reservoir, China

Sci. Total Environ.

(2016)

G. VanTrump et al.

The U.S. geological survey RASS-STATPAC system for management and statistical reduction of geochemical data

Comput. Geosci.

(1977)

H. Wang et al.

Hydrogeochemical characterization of groundwater flow systems in the discharge area of a river basin

J. Hydrol.

(2015)

Y. Wang et al.

Strontium isotope characterization and major ion geochemistry of karst water flow Shentou, northern China

J. Hydrol.

(2006)

S. Yang et al.

Downstream sedimentary and geomorphic impacts of the Three Gorges Dam on the Yangtze River

Earth Sci. Rev.

(2014)

C.A.J. Appelo et al.

Geochemistry, Groundwater and Pollution

(2005)

Z. Cheng et al.

Chemical and isotopic response to intensive groundwater abstraction and its implications on aquifer sustainability in Shijiazhuang, China

J. Earth Sci.

(2017)

C.A. Crandall et al.

Hydrochemical evidence for mixing of river water and groundwater during high-flow conditions, lower Suwannee River basin, Florida, USA

Hydrogeol. J.

(1999)

J.C. Davis

Statistics and Data Analysis in Geology

(1990)

Y. Du et al.

Sources and fate of high levels of ammonium in surface water and shallow groundwater of the Jianghan Plain, Central China

Environ. Sci. Process Impacts

(2017)

Duan, Y., 2016. Seasonal variations of groundwater arsenic concentration in shallow aquifers at Jianghan Plain,...

S. Fendorf et al.

Spatial and temporal variations of groundwater arsenic in south and southeast Asia

Science

(2010)

G.W. Feyereisen et al.

Long-term water chemistry database, Little River Experimental Watershed, southeast Coastal Plain United States

Water Resour. Res.

(2007)

Y. Gan et al.

Groundwater flow and hydrogeochemical evolution in the Jianghan Plain, central China

Hydrogeol. J.

(2018)

S.M. Gorelick et al.

Global change and the groundwater management challenge

Water Resour. Res.

(2015)

Cited by (57)

The enhanced visible-light-driven porous O/P-C<inf>3</inf>N<inf>4</inf> for persulfate photoactivation: Enhanced removal of refractory pollutants and lignin valorization
2024, Chemical Engineering Journal
In this research, an enhanced visible light response of oxygen and phosphorus-doped porous g-C₃N₄ (HAPA-CN) was prepared by thermo-polymerization of urea, hydroxyacetic acid and phytic acid. Its internal structure was verified by solid-state nuclear magnetism (NMR) and secondary ion mass spectrometry (SIMS). The surface electron density on HAPA-CN was enhanced by the inclusion of oxygen and phosphorus. The boosted photocatalytic activity was attributed to the high spectral utilization of sunlight and meliorated charge separation efficiency. The experimental results showed that the 0.05 HAPA-CN/persulfate (PS) system exhibited a higher efficiency in the photodegradation process of bisphenol A (BPA) and 2-mercaptobenzothiazole (MBT). Compared with ordinary g-C₃N₄, the integration of photocatalysis and persulfate oxidation led to an increase in the removal of BPA and MBT pollutants, which increased by 68.11 and 11.18 times, respectively. 0.05 HAPA-CN also exhibited a certain photodegradation of BPA under long-wavelength light irradiation such as blue, green, and red light. In the photocatalytic conversion of sodium lignosulfonate (SL), the 0.05 HAPA-CN/PS system achieved a maximum yield of 134.34 mg/g_SL vanillic acid at 30 min, which was significantly better than that of g-C₃N₄. Synergistic gas production (H₂, CO, CH₄, C₂H₄, C₂H₆) was also significantly enhanced. Combined with theoretical calculations, the mechanism was also analyzed. This work provides new insights for exploring high efficiency photocatalyst that combine porous structure, oxygen and phosphorus co-doping and PS activation technology for removal of organic pollutants and lignin valorization.
Towards a high-performance anode for zinc metal batteries: A tri-functional nitrogen-defective graphitic carbon nitride material for anode protection
2023, Journal of Colloid and Interface Science
As an important potential candidate for large-scale energy storage, rechargeable zinc metal batteries have become a research hotspot. Zn metal anodes are an extremely crucial component of rechargeable Zn metal batteries. However, the Zn dendrites, the evolution of hydrogen and side reactions on the surface of the Zn metal anodes severely hinder its commercial use. Here, we report a modified Zn metal anode with a layer of nitrogen-defective graphitic carbon nitride (NDCN) nanosheets coated onto a commercial Zn foil by a simple spraying method. The NDCN coating on the anode’s surface not only provides zincophilic sites, but also drives the deposition of zinc with a particular crystallographic orientation. At the same time, the formation of zinc dendrites and the evolution of hydrogen were suppressed, which enhanced the reversibility of the anode. Thus, the symmetric cells with the NDCN-protected zinc foil (NDCN@Zn) electrodes remained stable for 2000 h at a current density of 5 mA cm⁻². More importantly, the long-term cycling performance of the full cell tested at 1 A g⁻¹ retained approximately 88 % of its capacity even after 2300 cycles. This simple but effective method will provide a reference for future studies on protecting the anodes of zinc metal batteries.
Characteristics and evolution of karst collapse columns in the Huainan coalfield
2023, Science of the Total Environment
Karst collapse columns (KCCs) seriously affect the mining safety of deep coal seams. This study systematically summarizes and analyzes the development of KCCs, and classifies the different development stage to identify their development stages and evolutionary process in the Huainan coalfield. The evolution models for KCC development are given, combining with the exploration strata data from boreholes, the groundwater flow data of regional field, the hydrogeochemical data from the relative aquifers of KCC, and the hydrodynamic parameters. The results show that first types of KCCs are discovered in Liuzhuang and the Pansan mine, which is broken and disorganized, with a high degree of filling and cementation, and with the low permeability and velocity, and lower storage capacity. The KCCs in the Xieqiao and Zhangji mine have various morphologies and size differences. Their internal rocks are broken and semi-cemented, with a coefficient of permeability between 0.2 and 0.5 m/d and a specific discharge between 0.1 and 0.2 L/s•m. The KCCs are located in the cone of depression and are part of a sink area with moderate water-richness, where upper and lower aquifers have close hydraulic connections with mixed water quality. The internal core of the KCCs in the Gubei Mine is close to the center of the cone of depression which is highly broken and disorganized, with a high permeability, and a specific discharge more than 1.5 L/s•m. The intensity of runoff increases as its vertical depth also increases, especially the local areas are a higher hydraulic conductivity. Based on comprehensive hydrogeological characteristics, the KCCs development is divided into three stages: growing stage, declining stage, and dead stage. According to their characteristics of different stages, a series of evolutionary processes are established. Combining the sedimentological, karstological, and hydrogeological theories, some measures have been taken for prevention and control of mine water hazard in the various developmental stages. Hence, this research not only provides a new classified approach for KCC stages, but also an essential reference for a better understanding the mechanism of water inrush of KCCs in Northern China.
Study on coupling of typical elements in surface water and groundwater in the middle reaches of the Yangtze River, China
2023, Journal of Hydrology
Understanding the coupling relationship of typical elements between surface water (SW) and groundwater (GW) is helpful to the prevention of regional water environment pollution. In this study, the coupling of typical elements (C, N, P) of SW and GW in the middle reaches of the Yangtze River was comprehensively analyzed using multiple statistics and hydrogeochemical modeling methods. SW chemical characteristics was found similar with that of GW in the middle reaches of the Yangtze River with cluster analysis, and all sampling points was divided into two principal components based on principal component analysis. One principal component was more affected by the water-carbonate mineral interaction, the other one was greatly affected by anthropogenic activities. More interestingly, no fixed replenishment and discharge relationship was found between SW and GW in different seasons, which laid a foundation for the coupling of typical elements between SW and GW. Cl^- in SW and GW had the same major source. In addition, almost identical water-carbonate reactions was found occurred in phreatic water (PGW) and confined water (CGW). As a link between SW and CGW, PGW may accept the supplies of the two waters at the same time, resulting the concentrations of δD, Cl^-, DOC and HCO₃^– in PGW ranking between SW and CGW. Furthermore, Nitrogen and phosphorus were found to migrate between SW and GW, and a potential ammonium source in shallow aquitard which generated a large amount of ammonium. The coupling effect of nitrogen source input and transformation occurred in the process of nitrogen coupling between SW and GW, which increased the ecological risk of water environment. This study provided a scientific basis for the control of SW-GW pollution in the middle reaches of the Yangtze River, and supported the security of the water ecological environment in the Yangtze River Basin.
Detection of leakage in the plunge pool area at Xiluodu arch dam with an integrated approach
2023, Journal of Hydrology
The Xiluodu arch dam is one of the highest dams in the world. A noticeable leakage having a maximum rate of 192 L/s and accounting for 46 ∼ 75 % of the total discharge at the site has been observed in the plunge pool area since impoundment. For long-term safety of the dam, an integrated approach of groundwater temperature analysis, hydrochemical data clustering, cross-correlation function and inverse modelling is presented to detect the sources and flow paths of the leakage. It is found that the leakage is originated from three major sources, i.e., the confined limestone aquifer (underlain by the unconfined basalt aquifer), the reservoir and the downstream river. The vertical leakage from the limestone aquifer occurs through unplugged boreholes drilled during site investigation. Inverse modelling shows that induced by particle filling, clogging and rock deformation, the effective permeability of the unplugged boreholes follows the same exponential decay with the near-field basalt rocks upstream the impervious system. Consequently, the amount of the vertical leakage decreases annually, which leads to an annual decrease of the discharge in the plunge pool area. The leakage from the reservoir and the downstream river is dynamically stable in magnitude, which occurs through the basalt aquifer by bypassing the grout curtains in the dam and subsidiary dam foundations, respectively. During the past seven years of operation, the proportions of leakage from the above three sources account for, on average, 44 %, 31 % and 25 %, respectively.
How does a combination of numerical modeling, clustering, artificial intelligence, and evolutionary algorithms perform to predict regional groundwater levels?
2022, Computers and Electronics in Agriculture
The prediction of groundwater levels in arid and semi-arid regions is of great importance to tailor the best water management strategies. In this study, we propose a new approach that combines simulation, clustering, and optimization tools for groundwater level prediction. This approach simulates groundwater levels (GWL) using the MODFLOW method, clusters the study aquifer into different clusters using the k-mean method, and predicts regional GWL using the artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) methods that were optimized by the Harris Hawks Optimization (HHO), Whale Optimization Algorithm (WOA), and Particle Swarm Optimization (PSO). The efficacy of our approach was evaluated via a case study in northwest Iran. The MODFLOW method simulated the distribution of GWL across the study area with R² = 0.99, root mean square error (RMSE) = 0.97 m (m) and mean absolute error (MAE) = 0.82 m. The k-means method clustered the aquifer into seven clusters based on the hydraulic conductivity, storage coefficient, groundwater level, groundwater depth, groundwater withdrawal, and aquifer saturation thickness parameters. The prediction of groundwater level for each cluster demonstrated the accurate performance of all optimized models with mean RMSE = 0.6 m and mean absolute percentage error (MAPE) = 0.23 m. The prediction phase identified groundwater level in the previous month (obtained from the MODFLOW method), withdrawal of aquifer, precipitation, temperature, and evaporation as the most influential variables for groundwater levels in different clusters. We recommend the methodology proposed here for the prediction of groundwater levels in different aquifers with heterogeneous characteristics that pose computational burdens and uncertainties.

View all citing articles on Scopus

View full text

Research papersUsing cluster analysis for understanding spatial and temporal patterns and controlling factors of groundwater geochemistry in a regional aquifer

Highlights

Abstract

Graphical abstract

Introduction

Section snippets

Study area

Data and statistical procedures

Classified seven clusters

Spatial patterns and controlling factors

Four zones of groundwater geochemistry

Temporal variation and controlling factors

Comparison with conventional cluster analysis

Conclusions

Declaration of Competing Interest

Acknowledgments

J. Hydrol.

Sci. Total Environ.

J. Geochem. Explor.

Sci. Total Environ.

J. Geochem. Explor.

J. Hydrol.: Reg. Stud.

J. Hydrol.

J. Hydrol.

Environ. Pollut.

J. Environ. Manage.

Sci. Total Environ.

J. Hydrol.

Sci. Total Environ.

Sci. Total Environ.

Water Res.

Geochim. Cosmochim. Acta

Environ. Modell. Soft.

Water Res.

Sci. Total Environ.

Comput. Geosci.

J. Hydrol.

J. Hydrol.

Earth Sci. Rev.

Geochemistry, Groundwater and Pollution

Chemical and isotopic response to intensive groundwater abstraction and its implications on aquifer sustainability in Shijiazhuang, China

J. Earth Sci.

Hydrochemical evidence for mixing of river water and groundwater during high-flow conditions, lower Suwannee River basin, Florida, USA

Hydrogeol. J.

Statistics and Data Analysis in Geology

Sources and fate of high levels of ammonium in surface water and shallow groundwater of the Jianghan Plain, Central China

Environ. Sci. Process Impacts

Spatial and temporal variations of groundwater arsenic in south and southeast Asia

Science

Long-term water chemistry database, Little River Experimental Watershed, southeast Coastal Plain United States

Water Resour. Res.

Groundwater flow and hydrogeochemical evolution in the Jianghan Plain, central China

Hydrogeol. J.

Global change and the groundwater management challenge

Water Resour. Res.

Research papers
Using cluster analysis for understanding spatial and temporal patterns and controlling factors of groundwater geochemistry in a regional aquifer