Groundwater recharge pathway according to the environmental isotope: the case of Changwu area, Yangtze River Delta Region of China

Groundwater recharge refers to the process by which an aquifer or the water-bearing system obtains water from the outside world (Scanlon et al. 2002; Sophocleous 2002). Recharge sources are wide, including precipitation infiltration recharge, surface water, the condensate, the irrigation water return, the leakage, and the artificial recharge supply (Zhang et al. 2015; Herrmann et al. 2016; Abe et al. 2018; Zhang & Wang 2020). According to the recharge sources of groundwater, Saether & Caritat (1997) defined three main recharge mechanisms that can coexist: direct recharge, indirect recharge, and local recharge (Liu et al. 2019). The recharge of groundwater by these three recharge mechanisms generally occurs in the form of infiltration of precipitation into the soil, which is a complex process of infiltration into the soil and infiltration through the soil matrix.

Ascertaining the recharge mechanism of groundwater is the foundation of the reasonable evaluation and scientific management of groundwater resources (Chatterjee et al. 2009; Dennehy et al. 2015). Since the mid-1980s, the international hydrogeological community has attached great importance to the quantitative study of groundwater recharge (Zektser 2002). The methods of quantitative research on groundwater recharge have gradually become more plentiful, and the application of methods has altered from one to a variety of technical methods. Currently, three types of techniques are applied to the study of groundwater recharge. The first is the physical method, such as the groundwater level dynamic method (Yu & Chu 2012), Darcy's law (Malekani et al. 2018), the zero flux surface method (Russoniello & Michael 2015), the baseflow partitioning method (Bartyzel & Rozanski 2016), and the permeameter method (Masoud et al. 2018; Pourghasemi et al. 2020). The second is the numerical simulation method, such as the water balance model (Yenehun et al. 2020) and the numerical model based on the Richards equation (Hassen et al. 2016; Hollander et al. 2016; Neto et al. 2016). Theoretically, the range of recharge amount estimated by a model is unlimited, but the reliability of a model is unverified by effective measured data. The third is the tracer method. Tracer technology is applied to the study of groundwater recharge and its theoretical method is developing rapidly. Three types of tracers are used in tracer technology: historical tracers (Boutton et al. 1999), environmental tracers (Li et al. 2017; Madrala et al. 2017), and artificial tracers (Clark et al. 2005; Moeck et al. 2017). Chloride and stable hydrogen and oxygen isotopes are the main environmental tracers, and the related research methods involve chlorine conservation methods (Leenheer et al. 2001) and the stable isotopes of the hydrogen and the oxygen method (Mezga et al. 2014; van Geldern et al. 2014; Vystavna et al. 2018). In the investigation of the groundwater recharge mechanisms, the stable hydrogen and oxygen isotopes are primarily employed to evaluate the recharge source and mechanism, and the conservation of the chlorine element can be utilized to calculate groundwater recharge.

The environmental isotopes most commonly used in groundwater studies include the stable isotopes deuterium (D), oxygen-18 ⁠, sulfur-34 ⁠, and the radioactive isotope tritium (T). Stable isotopes D and primarily reflect groundwater sources, hydraulic links, aquifer leakage, and paleoclimate changes (Kumar et al. 2019; Arumi et al. 2020; Kreis et al. 2020; Priestley et al. 2020). is chiefly employed to identify pollution sources and geothermal system water flow (Awad et al. 1995; Lee & Lee 2018). Radioactive isotope T is typically applied for the identification of recent recharges and the vadose transport process (Samborska et al. 2013). Comprehensive analysis of stable isotopes and radioisotopes can effectively reveal the transformed relationship between different water bodies.

Influenced by intensive exploitation, the groundwater level in the Changwu area, Jiangsu Province, has dropped dramatically, and the water circulation within the groundwater system has intensified (Ding et al. 2020). The hydraulic connection and recharge relationship between groundwater aquifers has become more complex, a development which generates difficulties in the evaluation of groundwater resources of each layer. To identify the recharge relationship of the groundwater and surface water in the study area under the current mining conditions, environmental isotopic D, ⁠, ⁠, and T data obtained in 2018 were used in this work. Through comprehensive comparative analysis, the recharge relationship of groundwater and surface water in each layer is determined. This approach provides a scientific basis for correctly delimiting the horizontal and vertical boundary types of different groundwater aquifers and improving the accuracy of groundwater resource evaluation in the study area.

The study area comprises Changzhou City and the Wujin District of Jiangsu Province, located in the hinterland of the Yangtze River Delta. The terrain is flat, the surroundings and the expanse along the Taihu Lake area is slightly higher in the north and lower in the south, with an approximate elevation difference of 1–2 m. The study area covers an area of approximately 1,793 km², including the four municipal districts of Xinbei, Tianning, Zhonglou, and Wujin, which are called the ‘Changwu area’ (Hu et al. 2009). The geographical coordinates are 31°20–32°03′N and 119°40–120°12′E. The geographical location and hydrogeological profile of the study area are shown in Figure 1.

Porous groundwater in the Changwu area is mainly stored in quaternary loose sediments. The porous phreatic water and the I, II, and III confined aquifers are distributed from top to bottom. Porous phreatic water and the I confined water are mainly supplied by atmospheric precipitation, surface water, and agricultural irrigation water and are chiefly discharged by exploitation and evaporation. The II and III confined aquifers are primarily employed for development and utilization, and the main mining layer of the groundwater is the II confined aquifer.

The II confined aquifer belongs to the Quaternary Middle Pleistocene aquifer, which is distributed in the Qidong formation and divided into the upper (II₁) and lower (II₂) aquifers. The II₂ aquifer is the most important aquifer, with extensive thickness and good water yield. From the mid-1980s to 2000, the II confined aquifer was in a long-term over-exploitation state, and the water level dropped dramatically, thereby forming a large-scale water level depression cone. As a result, the groundwater seepage field is seriously unbalanced and a series of environmental geological problems (such as land subsidence, ground fissures, and land collapse) was induced. This condition seriously threatens the safety of residents’ lives and properties, destroys the ecological environment, and affects the sustainable development of the local economy. Accordingly, local governments at all levels have formulated several management regulations on groundwater mining restriction and prohibition since 2000. This endeavor curbed the continuous decline of the groundwater level, which has alleviated the rate of land subsidence. Ascertaining exploitable groundwater resources is the foundation of reasonable exploitation and the protection of the geological environment. This study delimits the boundary type of the groundwater flow numerical simulation area in view of the recharge relationship between the environmental isotopes in each layer of groundwater, surface water, and groundwater.

According to the isotopic characteristics of groundwater, the environmental isotopes of tritium (T), sulfur-34 ⁠, deuterium and oxygen-18 were selected to study the recharge relationship of groundwater in each layer of the Changwu area. A total of 28 samples was collected in March 2018. The hydrological and groundwater quality monitoring data in this paper are all derived from the groundwater environment monitoring reports of Wujin District, Changzhou City, Jiangsu Province, as well as the monitoring, series testing and test data of this study in recent years. Four samples from the I confined aquifer, 19 are from the II confined aquifer, one sample from the III confined aquifer, two samples from the limestone aquifer, and two samples from the sandstone aquifer. The spatial distribution of samples and the environmental isotope test data are shown in Figure 1 and Table 1, respectively. To compare the environmental isotopic contents of ⁠, ⁠, and in the limestone water and the I confined aquifer, four limestone water samples were collected in the typical limestone area of Nanjing, and one of the I confined water samples was collected in the Suzhou area. The measured values of the environmental isotopes of the comparative samples are shown in Table 2.

The content and correlation of the stable isotopes ⁠, ⁠, and can qualitatively characterize the source and mechanism of groundwater recharge. Radioisotope T can be employed to identify the age of groundwater recharge and assess the renewal capacity of groundwater.

The values of and in groundwater can indicate its recharge source, temperature during recharge, and hydrochemical process experienced by groundwater. There is a good linear relationship between and in atmospheric precipitation (Ribeiro et al. 2020). Global meteoric water line ⁠. The degree of deviation of the and values of groundwater from the local meteoric water line can be utilized to analyze the strength of the geophysical and chemical processes in the recharge and runoff processes of groundwater. The greater the deviation, the stronger the geochemical process (Ette et al. 2020).

Tritium is a radioactive isotope of hydrogen and mainly exists in the environment in the form of HTO (Wigger & Van Loon 2018). Tritium is unstable and decays into helium with a half-life of 12.33 ± 0.33a. HTO molecules decay continuously in the natural water cycle, such that the tritium content in water lowers over time. Therefore, according to its decay law and the measured tritium content in groundwater, the recharge time and mixing mechanism of groundwater can be approximately estimated. The recovery methods of tritium concentration in precipitation are mainly related to the interpolation method, hyperbola method, unitary linear correlation method of different influencing factors, and factor analysis method (Grahek et al. 2016; Kashiwaya et al. 2017; Larionova et al. 2017; Ansari et al. 2018; Vu et al. 2020). The tritium recovery curve of the atmospheric precipitation in the Nanjing area established by Zhang et al. (2011) is shown in Figure 2.

In 1963, the tritium concentration in the precipitation in the Nanjing area was as high as 1,764 TU, while that in the Nanjing area exceeded 200 TU between 1963 and 1968. By the 1970s, the tritium concentration in the precipitation of the Nanjing area was generally between 50 and 150 TU. After the 1980s, the level gradually returned to the natural one. According to the tritium recovery curve of the atmospheric precipitation in the Nanjing area, the average tritium concentration of groundwater formed by atmospheric precipitation in different periods can be recovered. The evolution path of tritium in groundwater can then be deduced on the basis of the tritium decay cycle.

The values of H13 and H19 in the I confined water samples are approximately 20‰ and belong to precipitation recharge. Meanwhile, the value of H16 is at 10 ‰, an amount which is affected by acid rain which, in turn, is consistent with the high value.

The concentrations of H21 and H22 in the limestone aquifer samples are approximately 30 ‰, are close to the 28‰ to 30 ‰ range of the typical limestone distribution areas in Nanjing. Analysis revealed a direct spatial contact relationship between the limestone aquifer and the II confined aquifer in the Changwu area. Additionally, the groundwater level of limestone is lower than that of the II confined aquifer which, in turn, recharges the limestone aquifer. Given the high concentrations in the limestone water, the minerals are still dissolved after the II confined aquifer is recharged into the limestone, and the gypsum is mainly responsible for the increase in concentration.

The concentrations of H25 and H2 in the sandstone aquifer samples are 14.1 ‰ and 27.6 ‰, respectively. As H25 is located in the sandstone outcrop area and is affected by modern acid rain, the concentration is approximately 14 ‰. However, H26 is located in the buried sandstone area and contacts limestone aquifer, thereby resulting in the concentration of 27.6‰, a value which is very close to that of the limestone distribution area.

The II confined aquifer includes 19 samples. When T ≤ 3 TU, two samples (H27 and H28) in this set of data belong to the reduction of and lead to the increase in concentration. When T > 3 TU, 13 samples with concentrations are present. The spatial distribution of this group of samples revealed that the concentration of H9 is 0.02 mmol/L, and the concentration is 42.7‰. These outcomes can be summed up as the reduction in ⁠, thereby resulting in the increase in concentration. The other 12 samples can be divided into two groups: when ⁠, four samples (H14, H15, H17, and H20) showed no relationship between T and ⁠. When TU, eight samples showed the trend of T and in inverse proportion, as shown in Figure 3.

The concentrations of the samples H13, H16, and H19 in the I confined aquifer are stable in the range −5.6 ‰ to −5.8 ‰, and the concentrations are in the range −43 ‰ to −46 ‰. Given that H26 belongs to concealed sandstone aquifer, in the H25 and H26 samples of the sandstone water body, the concentration was −8.3‰ and the concentration was −67‰ (values which were lower than those of the other samples taken) because H26 belongs to the concealed sandstone aquifer. The relationships of the remaining samples can be divided into the following two groups (Figure 4). One group of concentrations is stable between −7.8‰ and −7.9‰, this generally does not change with the change of T concentrations, and includes H3, H4, H5, H6, H1, H20, and H21. The concentrations of the other group of samples were between −6.8‰ and −7.8‰, and the distribution of concentrations and T concentrations was approximately proportional.

The relationships for the remaining samples can be divided into the following two groups (Figure 5). One group of concentrations was stable between −60‰ and −65‰, and generally does not change with the change of T concentrations, and includes H7, H8, H9, H10, H11, H12, H14, and H15. The concentrations of the other group of samples were between −53‰ and −59‰. The concentrations and T concentrations were inversely proportional to each other.

Figure 6 indicates that the H13, H16, and H19 samples of the I confined aquifer were relatively concentrated in the upper right corner and were located below the atmospheric water line ⁠. The H21 and H22 samples of the limestone aquifer are relatively concentrated in the middle near the atmospheric water line. For the H25 and H26 samples from the sandstone aquifer, the H25 is close to the meteoric water line. This situation arises because H25 is located in the sandstone outcrop area, and the recharge environment of concealed sandstone aquifer and other samples vary, such as the disparity in the different average temperatures or times.

According to the geographical location and hydrogeological conditions of water samples, the II confined aquifer can be divided into the following three groups. The first group includes nine samples close to the meteoric water line, including H1, H2, H3, H4, H5, H6, H20, H27, and H28. The second group consists of eight samples, including H7, H8, H9, H10, H11, H12, H14, and H15. These samples are located under the meteoric water line and are distant from one another. The third group consists of four samples at the transition zone, including H17, H18, H23, and H24.

Figure 7 indicates that in the D domain of the I confined aquifer sample, the water has an evaporation effect and modern water recharge factors, thereby leading to the deviation from the atmospheric water line. The F domain of the II confined aquifer sample in the middle of the figure includes limestone aquifer samples H21 and H22 in the Henglin area. The limestone aquifer is proven to be recharged by the II confined aquifer because of the same distribution range observed. The E domain of the II confined aquifer samples include the six samples H1, H2, H3, H4, H5, and H6 from the area along the Yangtze River. The relationship of the group of this data proves that does not change with the change of tritium concentration, so a recharge of the Yangtze River water occurs in terms of the recharge environment. Only the mean value of the Yangtze River is relatively stable, and the tritium concentration can change with time. Thus, the water of the E domain has a recharge under the modern environment.

The D domain of the II confined water samples includes eight samples around the northwest of Gehu Lake where an evaporation effect occurs in the water, so it deviates far from the atmospheric water line. The biggest difference between the G samples and the E and F domain samples is that the water body of the G domain samples is evaporated and then recharged to groundwater. One of the samples in the E and F zones may be recharged by direct infiltration of precipitation, and the other may be recharged by infiltration of the Yangtze River water. Therefore, the G domain water body cannot receive the supply of the Yangtze River water, but only that of the lake water, because the lake water body is heated and evaporated. Regarding the location of the G domain, no evaporation of the lake water ensues under modern climate conditions and the lake water should be in a colder environment than that the modern climate. Therefore, no modern groundwater recharge happens in the G domain.

The water in the F domain is close to the atmospheric water line. Moreover, the lake water in this region is not recharged by the evaporation effect, a feature which may arise from natural precipitation or the Yangtze River water. However, compared with modern precipitation, Suzhou is located in the A domain, Nanjing is located in the B domain, Changwu should be located between them, and the I confined aquifer of Changwu is located in the D domain. If the effect of evaporation is subtracted, Changwu should be in the A domain or the B domain, which proves that the F domain cannot be the modern precipitation supply. Compared with the E domain which receives the water supply from the Yangtze River, the F domain is closer, but this relationship does not exist in the geographical location of the sample site. According to the ancient geographical location of the Changwu area, the present Henglin area is equivalent to the ancient Furong Lake low-lying land 2000 years ago (Zhao et al. 2007; Miao et al. 2016), and Furong Lake was connected to the Yangtze River in history. The predecessor of the Furong Lake is speculated to be a bend of the Yangtze River, that later degenerated into an ancient Furong Lake because the bend was gradually silted up. Historical research proves that the silting of the Yangtze River bend occurred four thousand years ago (Chen & Yang 1991; Min & Li 1992). Therefore, the water of the F domain is presumed to be recharged by the Yangtze River about four thousand years ago, and no recharge exists now.

With the development of the Holocene, the south bank of the Yangtze River (Changwu section) continuously deposited new sedimentary layers, and the river bank continued to move northward. Furthermore, a new aquifer was formed in the Weicun–Ligang–Shengang–Xiagang area along the river and received the recharge of the modern Yangtze River water.

A stable isotope reflects the change process of the climate and environment. At present, the commonly used comparison curve in the world is the ice core data of Greenland and Antarctica (Wunsch 2003; Landais et al. 2015) and includes the change and enrichment of the climate in the past 130,000 years, for which the climate in the past 8,000 years is relatively stable. According to environmental isotope, the recharge relationship of groundwater in the Changwu area is as follows: the shallow aquifer and the I confined aquifer are mainly recharged by atmospheric precipitation and surface lake water. The II confined aquifer along the Yangtze River is recharged by modern precipitation. The II confined aquifer in the Henglin area was recharged by the ancient Yangtze River before 4,000 years ago, and no recharge relationship exists now. The recharge condition of the II confined aquifer around the northwest of the Gehu Lake is in the climate environment of 8,000 years ago and was recharged by the surface depression lake water at that time.

The II confined aquifer in the transition zone of Gehu and the Henglin area does not have the recharge under the present climatic conditions and mainly accepts part of the leakage recharges from the I confined aquifer. Moreover, the concealed limestone aquifer is mainly supplied by the II confined aquifer, and the concealed sandstone aquifer supplies the II confined aquifer. The formation of the II confined aquifer in Changwu is directly related to the changes of the Yangtze River channel. The right channel of the Yangtze River erodes the south bank of the river. When the channel erodes to an aquifer at a certain depth, the groundwater in this aquifer drains into the Yangtze River, thereby causing vertical infiltration recharge of water above the aquifer.

The hydrological and groundwater quality monitoring data in this paper are all derived from the groundwater environment monitoring reports of Wujin District, Changzhou City, Jiangsu Province, as well as the monitoring, series testing and test data of this study in recent years. In total, 28 samples were collected in March 2018.

Liang He: Conceptualization, Methodology, Validation. Junru Zhang: Data Curation, Writing – Original Draft. Suozhong Chen: Writing – Review & Editing, Funding acquisition. Manqing Hou: Resources, Visualization. Junyi Chen: Software, Supervision.

We are grateful for the funding support from the National Natural Science Foundation of China (Grant No. 41571386), the Priority Academic Program Development of Jiangsu Higher Education Institutions (Grant No. 1612206002), and the Postgraduate research and innovation plan project in Jiangsu Province (Grant No. KYCX21_1334).

All relevant data are included in the paper or its Supplementary Information.