Environ Earth Sci (2014) 72:3901–3913 DOI 10.1007/s12665-014-3279-5

ORIGINAL ARTICLE

A study of the interrelation between surface water and groundwater using isotopes and chlorofluorocarbons in Sanjiang plain, Northeast China

Bing Zhang • Xianfang Song • Yinghua Zhang • Dongmei Han • Changyuan Tang • Lihu Yang • Zhongliang Wang • Tingyi Liu

Received: 14 June 2013 / Accepted: 7 April 2014 / Published online: 27 April 2014 Ó Springer-Verlag Berlin Heidelberg 2014

Abstract Surface water and groundwater are the main recharged from Songhua river. The combination of stable water resources used for drinking and production. Assess- isotopes, tritium, and CFCs was an effectively method to ments of the relationship between surface water and study the groundwater ages and interrelation between sur- groundwater provide information for water resource man- face water and groundwater. Practically, the farmlands near agement in Sanjiang plain, Northeast China. The surface the river and under foot of the mountain could be culti- water (river, lake, and wetland) and groundwater were vated, but the farmlands in the central plain should be sampled and analyzed for stable isotopic (dD, d18O) controlled. composition, tritium, and chlorofluorocarbons concentra- tions. The local meteoric water line is dD = 7.3d18O–6.7. Keywords Hydrogen and oxygen isotopes The tritium (T) and chlorofluorocarbon (CFC) contents in Chlorofluorocarbons Surface water Groundwater groundwater were analyzed to determine the groundwater Sanjiang plain ages. Most groundwater were modern water with the ages \50 years. The groundwaters in mountain area and near rivers were younger than in the central plain. The oxygen Introduction isotope (d18O) was used to quantify the relationship between surface water and groundwater. The Songhua, Surface water (river, lake) and groundwater are important Heilongjiang, and Wusuli rivers were gaining rivers, but water resources for life, agriculture, and industry. Assess- the shallow groundwater recharged from rivers at the ments of the interrelation between surface water and confluence area of rivers. At the confluence of Songhua and groundwater provide information for water resource man- Heilongjiang rivers, 88 % of the shallow groundwater agement. The three basic relationships between surface water (river) and groundwater are: (1) rivers gaining water from groundwater; (2) rivers losing water to groundwater; B. Zhang Z. Wang T. Liu (3) rivers gaining in some reaches and losing in other Tianjin Key Laboratory of Water Resources and Environment, reaches (Winter et al. 1998; Woessner 2000; Banks et al. Tianjin Normal University, Tianjin 300387, China 2011). Not only the hydrogeological conditions, but also e-mail: [email protected] the human activities, such as agricultural irrigation, influ- B. Zhang X. Song (&) Y. Zhang D. Han L. Yang ence the relationship between surface water and ground- Key Laboratory of Water Cycle and Related Land Surface water (Sophocleous 2002; Anderson 2005). Processes, Institute of Geographic Sciences and Natural The application of environmental isotopes, especially Resources Research, Chinese Academy of Sciences, Beijing 100101, China the hydrogen and oxygen, was used to study the interre- e-mail: [email protected] lation between surface water and groundwater widely and efficiently (Criss and Davisson 1996; Clark and Fritz 1997; C. Tang Hunt et al. 2005; Baskaran et al. 2009; Banks et al. 2011). Departments of Environmental Science and Landscape Architecture, Faculty of Horticulture, Chiba University, The stable hydrogen and oxygen isotopes were applied to Chiba, Japan assess the relationship and water exchange between surface 123 3902 Environ Earth Sci (2014) 72:3901–3913 water and groundwater. The estimation of groundwater age dramatically reduced by agricultural development during was the key factor to identify the time scale of the water the past 60 years (Zhou and Liu 2005; Song et al. 2008; exchange. The tritium (3H or T) and chlorofluorocarbon Zhang et al. 2009; Huang et al. 2010b) Most wetland (CFCs) concentrations in groundwater were measured to wasconverted to paddy field to ensure food security. The estimate the groundwater age (Oster et al. 1996; Szabo 79 % area of the paddy field is using irrigation system. et al. 1996; Boronina et al. 2005). Furthermore, about 67 % of the irrigation area is only using Among the many chemical components in groundwater, groundwater as the irrigation water resource (Fig. 1b). The tritium concentrations allow differentiation between use of water, especially groundwater, for agricultural pro- groundwater of different mobility (Seiler and Lindner duction caused the decline of water table. The innovative 1995). Tritium is probably the most commonly employed and sustainable research and technologies of water radioisotope used to identify modern recharge. It is a short- resource and quality are required to ensure agricultural lived isotope of hydrogen with a half-life of 12.43 years. production in Sanjiang plain (Pereira et al. 2002; Wang and The atmospheric testing of nuclear devices between 1952 Tian 2003). and 1962 generated a tremendous quantity of atmospheric The agricultural development affects the interrelation tritium. The final year of megaton tests generated a huge between surface water and groundwater, especially the peak, which appeared in the spring of 1963. The curve surface water and groundwater irrigation system (Winter from the input function for tritium will have a peak cor- et al. 1998). Groundwater is the main water resource for responding to that time. Using this peak, one can get the irrigation in Sanjiang plain, but the groundwater replen- mean residence time and infiltration rate of water (Lu et al. ishment analysis and the relationship between surface 2008). As the atmospheric concentration of tritium water and groundwater were seldom studied. The envi- declined, its use for quantifying recharge by estimating the ronmental isotopes (stable hydrogen and oxygen, tritium) age of groundwater became less reliable (International in water were analyzed to assess the interrelation between Atomic Energy Agency 2006). surface water and groundwater. The purposes of this study The CFCs are used to date groundwater comparing with were (1) to characterize the field measurement indicators the tritium dating technique (Clark and Fritz 1997). CFCs (EC, pH, and water temperature), stable isotopic compo- are the unwanted contaminants in our atmosphere, sition (d18O, dD), tritium contents, and CFC concentrations including CFC-11 (CCl3F), CFC-12 (CCl2F2), and CFC- in water; (2) to assess the groundwater ages using tritium 113 (C2Cl3F3). They were widespread used commercially contents and chlorofluorocarbons concentrations, and (3) to and industrially during the second half of the twentieth describe the relationship between surface water and century (Hurtley 2011). These compounds are resistant to groundwater. The conclusions were drawn for maintaining degradation, making them a useful marker for modern agricultural development and sustainable water groundwater. CFCs are detectable in groundwater that has management. been recharged since about 1940 or in mixtures of older water with post-1940 water (Szabo et al. 1996). The use of CFC concentrations in natural waters as a potential dating Study area tool was recognized (Thompson and Hayes 1979). Mea- surement of the concentrations of a number of CFCs pro- Regional hydrogeology vides a complementary tool groundwater dating based on isotope data. The Sanjiang plain (129°1102000–135°0501000E, 43°4905500– Sanjiang plain is one of the nine grain production bases 48°2704000N) is a vast area of alluvial floodplains in in China. The cultivated lands in Sanjiang plain are well northeast Heilongjiang province, China, with a total area of known for producing large quantities of high-quality beans 108,900 km2. The Wanda Mountain separated the plain and rice in recent decades. The irrigated rice or upland into two parts. The north of Wanda Mountain is the San- crops are planted with one harvest per year after marshland jiang low plain, which is deposited by the Heilongjiang, conversion. The majority of upland crops are wheat, corn, Songhua, and Wusuli rivers. The south is the Muling- and soybean. The crop grows from May to September. The Xingkai plain, located in the north shore of the Xingkai average grain yields of rice, wheat, corn, and soybean from (Khanka) lake. There are 23 counties, 52 state-owned 1978 to 2008 were 5.23, 2.81, 4.36, and 1.82 t/ha, farms, and eight forest industry bureaus in the plain; the respectively (Fig. 1a) (Heilongjiang Land Reclamation population is 8.1 million (1990) (He 2000). The Hei- Bureau 2009). The grain production is dependent on ade- longjiang river is the international boundary between quate water supply of the usable quality and large area of Russia (northern side) and China (southern side). The fertile farmland. There is the largest area of wetlands until Wusuli river also forms the boundary between China recently in Sanjiang plain. These wetlands have been (western side) and Russia (eastern side). The Xingkai lake 123 Environ Earth Sci (2014) 72:3901–3913 3903

(a) 1.0x107 (b) 250 Rice 6

8.0x10 Wheat ) 2 200

Corn m Jowar 8 6.0x106 Millet Soybean 150 Cultivation area Irrigation area 4.0x106 Surface water irrigation 100 Groundwater irrigation Sprinkler irrigation Grain output (Million t) 6 2.0x10 50 Cultivation and Irrigation area (10 0.0 0

1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 1976 1980 1984 1988 1992 1996 2000 2004 2008 Year Year

Fig. 1 The increase of grain production from 1950 to 2008 (a) and irrigation area from 1990 to 2008 (b). Data from the Statistical yearbook of Heilongjiang reclamation area (Heilongjiang Land Reclamation Bureau 2009) is also the international lake; the northern part belongs to sandstone, and gravel. The thickness of the unconfined China, and the southern part belongs to Russia (Fig. 2a). aquifer is about 100–200 m in the Sanjiang low plain, The lengths of the Heilongjiang, Songhua, and Wusuli while it is about 40–80 m in the Muling-Xingkai plain rivers in Sanjiang plain are 406, 357, and 478 km, (Fig. 3). respectively (Song et al. 2010). The depth to water table is about 3–5 m in Sanjiang Climate/land use plain. The hydraulic gradient of the groundwater is about 1/1,000 in the hilly region, 1/5,000–1/10,000 in the central The Sanjiang plain is temperate humid and sub-humid plain, and 1/500–1/150 near the river. The groundwater continental monsoon climate. The mean annual tempera- discharges into the river and lake (Song et al. 2010). The ture ranges from 1.4 to 4.3 °C, with the maximum of 21 to flow direction of groundwater is from southwest to north- 22 °C in July, and the minimum of -18 °C in January. The east in the northern areas of the plain. However, the annual precipitation is 500–650 mm, and 80 % of rainfall groundwater flows from northwest to southeast in the south occurs during May–September. The frost-free period is of the Wanda Mountain. 120–140 days (Huang et al. 2010a). The temperature in the There are several hills consisting of volcanic rock in the south and plain area is higher than the north and mountain middle of the depression zone. The large-scale depression area, respectively. The annual C10 °C active accumulate occurred from Neocene to early period of Quaternary. The temperature is 2,300–2,500 °C. The annual potential place raised little after early period of Quaternary, then evaporation is 550–840 mm. depressed slowly, forming Quaternary deposit extensively. The most common landscape types are wetland and The Sanjiang low plain belongs to intermountain basin. cultivated land. The area of wetland is 99105 ha, The Xingkai lake basin formed in the Neocene, mainly by accounting for 27 % of the provincial wetland. The culti- the Muling river alluviation and the Xingkai lake deposi- vated land is 49106 ha, accounting for 30 % of cultivated tion. The Quaternary terrane is enough weathered by the land in the Heilongjiang province (Zhang et al. 2009). The uplifted edge zone of the plain and the surround mountains. main soil types are albic soil, meadow soil, and marsh soil, Thus, the depth of the cracked weathering zones is about and the natural vegetation is mainly of marsh vegetation, 10 m. The deep and extensive loosen sediment was with woodland meadow scattered on relatively high alti- deposited by the long time and slow depression of the tudes. The water and soil in marshes are completely frozen plain. In the eastern part of Sanjiang plain, the sandy clay from late October to following April and begin to melt in depth is about 4–20 m near the Qixing river (Heilongjiang late April (Pan et al. 2010). The boreal climate conditions geological survey hydrogeology engineering geology bat- and low slope grade have made the largest area of fresh- talion 1959). The unconfined aquifer consists of sand, water wetlands in China.

123 3904 Environ Earth Sci (2014) 72:3901–3913

(a)

Qiqihar

Harbin

GNIP station

(b)

Fig. 2 The Sanjiang plain location (a) and the distribution of water samples (b)

123 Environ Earth Sci (2014) 72:3901–3913 3905

Fig. 3 The geology schematic diagram of cross sections I–I0 (a) and II–II0 (b)(m.a.s.l means meters above sea level)

Table 1 The stable hydrogen Years Months d18O(%) dD(%) Years Months d18O(%) dD(%) and oxygen isotopes in precipitation at the Sanjiang 2005 1 -28.2 -207.3 2006 7 – – field station 2005 2 -19.9 -158.7 2006 8 – – 2005 3 -13.4 -100.9 2006 9 – – 2005 4 -15.0 -102.0 2006 10 -7.3 -54.2 2005 5 -13.1 -92.4 2006 11 -8.3 -75.2 2005 6 -5.9 -53.7 2006 12 – – 2005 7 -8.1 -61.2 2007 1 -21.2 -163.2 2005 8 -10.3 -78.6 2007 2 -9.3 -100.3 2005 9 -8.8 -67.6 2007 3 -16.9 -140.0 2005 10 -8.3 -58.5 2007 4 – – 2005 11 -14.8 -125.1 2007 5 -11.6 -93.1 2005 12 -24.1 -181.7 2007 6 – – 2006 1 -7.6 -54.5 2007 7 -6.2 -57.4 2006 2 – – 2007 8 – – 2006 3 -7.6 -58.1 2007 9 – – 2006 4 -18.8 -141.4 2007 10 – – 2006 5 -19.2 -148.5 2007 11 – – ‘‘–’’ indicates no data or no 2006 6 -4.7 -38.3 2007 12 – – precipitation during the month

Methods 2005 to 2007 (Table 1). In the Sanjiang field station (one station of the Chinese Network Isotopes in Precipitation, Water sampling CHNIP), a set composite of a polyethylene bottle and a funnel was placed outside as a rain collector. A ping pong The monthly composite precipitation sample collection has ball was put at the funnel to prevent evaporation. The snow been carried out over the observation period years from samples were collected using a pail installed on the ground.

123 3906 Environ Earth Sci (2014) 72:3901–3913

After each snowfall event, the snow samples melt at room equation: d = dD - 8 d18O. On a global basis, d averages temperature. These two kinds of samples were transferred about 10 %, but regionally it varies due to humidity, wind into a 50-ml polyethylene bottle as monthly samples. speed, and sea surface temperature during primary evapo- Surface water and groundwater were sampled along the ration (Clark and Fritz 1997). Songhua river, Heilongjiang river, Wusuli river and Xingkai lake for isotope and CFC analysis during 10–18, Tritium content analysis September 2009 (Fig. 2b). The wetland water was col- lected in the Sanjiang plain marsh ecological experiment The tritium concentration in the groundwater is a func- station. The groundwater samples were collected from tion of its residence time in the recharge environment. shallow (sampling depth\60 m) and deep (sampling depth The exponential piston flow model (EPM) is appropriate [60 m) wells. The surface water samples were collected at to real situations (Małoszewski and Zuber 1982, 1991; 30 cm under water surface. The groundwater was sampled McGuire et al. 2002; Morgenstern et al. 2010). The EPM after pumping the resident water in the well. The field of tritium concentration is predicted by the following measurement indicators, including electrical conductivity equations: (EC), water temperature, and pH, were measured in situ via Z1 an EC/pH meter (WM22EP, Toadkk, Japan), which was 0 0 0 0 CoutðtÞ¼ Cinðt t Þ expðkt Þgðt Þdt ; ð2Þ previously calibrated. 0 One 50 ml polyethylene bottle with watertight caps ðT=gÞ1 expðgt=T þ g 1Þ for t Tð1 g1Þ was used to store filtered (0.45 lm Millipore membrane gðt0Þ¼ ; ð3Þ 0 for t \T 1 g1 filter) water for stable hydrogen and oxygen isotopic ð Þ composition analysis. One litre filtered water in the k ¼ ln2=T1=2 ¼ 0:693=T1=2; ð4Þ polyethylene bottle with watertight caps was sampled for where C (t) and C (t) are the output and input concen- tritium content analysis. Two 100 ml brown glass bottles out in trations, respectively. t0 is the integration variable which capped with a special foil-lined cap were used to store physically represents the exit age of the tracer. T is the groundwater for chlorofluorocarbon (CFC) analysis. The 1/2 half-life of tritium, T is the mean residence time. The refrigeration grade copper tubing was required. The bot- parameter g is the ratio of the total volume to the volume tles and caps were thoroughly rinsed with the ground- with the exponential distribution of transit times. For this water. The bottles were filled underwater in a steel beaker area, the exponential fraction of the EPM is 80 %, and the and capped under water. All water samples in the bottles parameter g = 1.25. The tritium content of water samples for isotope and CFC analysis were without air and tightly was analyzed by Liquid Scintillation Spectrometer (1220 sealed to prevent evaporation. After bottling, all samples Quantulus, PerkinElmer, USA) in the Institute of Hydro- were stored at 4 °C. geology and Engineering Geology, Chinese Academy of Geological Sciences, Zhengding, China. The limit of Analytical methods detection is 1 TU, and the measurement precision is 0.6 TU. Stable isotope analysis Chlorofluorocarbon (CFC) concentration analysis The precipitation, surface water, and groundwater samples were analyzed in the Environmental Isotopes Laboratory of The CFC concentration of shallow groundwater was Institute of Geographic Sciences and Natural Resources determined by gas chromatography with electron capture Research, Chinese Academy of Sciences. A Finnigan MAT detector (ECD) system by cry focusing method (GC-14B, 253 mass spectrometer (Finnigan, USA) and the TC/EA Shimadzu, Japan) in Chiba University. The limit of method were used to carry out the isotopic composition detection is 0.01 pmol/kg, and the precision of the mea- measurements The results were expressed conventionally surement is 1 %. as d values, representing deviation in per mil (%) from the isotopic composition of a specified standard (Vienna standard mean ocean water, V-SMOW), 18 Results d OðdDÞ¼½ðRsample=RstandardÞ11; 000 & ð1Þ where R refers to the 2H/1H(18O/16O) ratio in sample and Characteristics of field measurement indicators standard. The measurement accuracy was consis- tently ± 1 % for dD and ±0.3 % for d18O, respectively. The characteristics of EC, pH, and water temperature The value of deuterium excess d is calculated by the measured in situ are shown in the Table 2. The EC mean 123 nio at c 21)72:3901–3913 (2014) Sci Earth Environ Table 2 The location, electrical conductivity, pH, isotopic composition and ages of water samples Sample no. Sample Latitude N Longitude E Elevationb Well EC (lS/cm) pH Temp.c dD(%) d18O(%) Tritium GW ages by CFCs typea (m) depth (°C) (TU) tritium (years) ages (m) (years)

SHJ27 LW 45°20047.300 132°22031.500 69 – 257 8.7 17.8 -57.6 -6.0 17 SHJ28 LW 45°18028.500 132°36047.900 69 – 213 8.68 19.1 -55.2 -5.9 17.5 Mean LW – 235.0 ± 31.1 8.7 18.5 ± 0.9 -56.4 ± 1.7 -6.0 ± 0.1 17.3 ± 0.4 SHJ01 RW 46°49024.900 130°21026.300 89 – 180 7.83 17.5 -80.8 -11.4 – SHJ02 RW 47°02023.000 130°42049.500 80 – 155.3 7.85 18.4 -84.8 -11.1 30.8 SHJ04 RW 47°14032.000 131°56004.800 59 – 156 8.22 18 -85.7 -10.7 23.9 SHJ06 RW 47°42018.300 132°31040.600 54 – 140 8.08 16.4 -83.4 -10.9 30.2 SHJ07 RW 47°4307.300 132°31009.400 54 – 64.4 7.73 14.7 -103.0 -13.9 28.4 SHJ08 RW 47°42036.800 132°31058.700 53 – 86.8 7.93 15.2 -99.1 -13.3 31.2 SHJ11 RW 48°15045.800 134°40052.800 40 – 103 7.7 15.5 -85.9 -11.4 22.1 SHJ13 RW 48°22028.400 134°17038.300 44 – 112 8 16 -90.3 -10.6 21.2 SHJ14 RW 48°05029.800 133°17055.900 36 – 120 8.14 17.4 -87.6 -11.7 19.5 SHJ17 RW 47°03034.400 133°14025.000 49 – 141.4 7.81 18.1 -73.4 -9.1 21.3 SHJ20 RW 46°48012.500 134°01044.600 43 – 80 8.4 18.2 -80.3 -11.0 15.4 SHJ22 RW 46°5100.400 133°20051.900 77 – 84.7 8.47 13.5 -91.3 -12.4 – SHJ23 RW 45°58059.000 133°40043.900 42 – 161.3 8.4 18.4 -74.9 -9.9 15.8 SHJ25 RW 45°32017.400 131°58016.200 112 – 366 8.13 15.3 -76.6 -9.9 16 SHJ29 RW 46°2005.800 132°15010.300 86 – 193 8.6 13.9 -84.0 -11.2 19.8 Mean RW – 142.9 ± 72.7 8.1 ± 0.3 16.4 ± 1.7 -85.4 ± 8.2 -11.2 ± 1.3 22.7 ± 5.8 SHJ03 SGW 46°59023.200 130°45013.300 87 20 468 7.12 8.5 -75.2 -9.4 19.8 41 ± 4 39–41 SHJ05 SGW 47°13039.500 131°56056.100 67 26 1,042 7.55 8.5 -81.0 -10.4 20.4 41 ± 4 [50 SHJ09 SGW 47°40048.800 132°35013.800 58 8 242 7.98 8.1 -85.3 -11.2 7.2 48 ± 4 37–38 SHJ15 SGW 48°0004.000 133°16027.400 57 30 222 7.83 6.5 -87.1 -11.9 3 50 ± 4 43–44 SHJ16 SGW 47°34052.000 133°07029.100 49 27 273 7.24 6.3 -85.8 -11.0 1.7 51 ± 4 [50 SHJ18 SGW 47°03024.700 133°15028.700 65 20 466 7.8 7.3 -85.6 -11.6 2.2 50 ± 4 [50 SHJ21 SGW 46°47024.100 134°01026.800 50 19 310 7.87 7.5 -88.5 -11.4 1.8 51 ± 4 [50 SHJ24 SGW 46°0009.000 133°38004.200 64 23 454 7.4 8.5 -77.1 -11.1 61.2 39 ± 10 40–42 SHJ26 SGW 45°32015.900 131°58018.900 114 3 328 8.15 12.2 -74.0 -8.0 19 42 ± 4 41–42 SHJ31 SGW 46°20007.100 132°15006.500 86 30 201 7.85 7.9 -86.9 -11.4 6.4 49 ± 4 41–43 SHJ33 SGW 46°45011.700 131°0800.700 97 30 371 8.45 6.6 -87.4 -11.0 6.6 49 ± 8 47–49 Mean SGW 21 ± 9 397.9 ± 234.6 7.7 ± 0.4 8.0 ± 1.6 -83.1 ± 5.3 -10.8 ± 1.1 13.6 ± 17.4 46 ± 4

123 SHJ03D DGW 46°59029.900 130°45017.400 76 120 236 7.21 8.2 -87.8 -11.5 – – SHJ05D DGW 47°13038.500 131°57006.000 67 80 387 7.75 6.8 -84.8 -10.8 – – 3907 SHJ12 DGW 48°15045.800 134°40052.800 40 100 260 7.26 9.4 -110.7 -15.1 11.5 45 ± 4 3908 Environ Earth Sci (2014) 72:3901–3913

values of water samples are sorted as shallow groundwater (397.9 lS/cm) [ deep groundwater (320.7 lS/cm) [ CFCs ages (years) spring (315 lS/cm) [ lake water (235.0 lS/cm) [ river water (142.9 lS/cm) [ wetland water (58.4 lS/cm). The pH mean value of lake water is 8.70, which is the most 4 4 4 ± ± ± basic, while the wetland water is neutral with the pH value 50 GW ages by tritium (years)

[ 7.44. The surface water is more basic and warmer than groundwater. The lake water is the warmest while the shallow groundwater is the coldest according to the water 5.8 47

± temperature. 1.0

(TU) The variations of electrical conductivity, pH, and water \ temperature along the flow direction of rivers are shown in 1.5 8.3 ) Tritium the Fig. 4. The values of electrical conductivity and the ± % water temperature decreased along the flow direction of O( 11.4 12.211.7 1.6 –9.610.2 51 – – – – – 10.8 11.8 45 11.9

18 Songhua river. The values of EC and water temperature ------d decreased sharply at the confluence of Songhua and Hei-

9.4 longjiang rivers, but increased along the flow direction of ) ± Heilongjiang river. The value of pH was constant without % 87.5 90.2 82.2 73.3 79.4 88.4 90.2 significant variation along the flow directions of Songhua D( d ------river and Heilongjiang river. The value of electrical conductivity decreased along the 1.4 c flow direction of Naoli river and Muling river to Wusuli ± C)

° river. The EC value of water sample SHJ22 was the least ( among the surface water along Naoli river, because the

0.4 8.8 water sample was collected in a creek at the mountain area ± after a precipitation. The values of pH decreased along the flow direction with the flow flux increasing. The value of water temperature increased along the flow direction of

122.3 7.6 Muling river and Naoli river. The Muling river and Naoli ± S/cm) pH Temp.

l river are the tributaries of Wusuli river. The flow quantities of the tributaries are less than Wusuli river. The water was EC ( standard deviation cold at stem stream of Wusuli river, since the water sample ±

19 320.7 SHJ11 was collected in the main stream. ± The pH values of shallow and deep groundwater are not Well depth (m) significantly different. The mean water temperature of deep b groundwater is 0.8 °C, which is warmer than that of shal- low groundwater; however, the EC value of shallow 575966 60 90 90 34853 32381 155 – 7.15 – 8.08 58.4 7.62 7.3 315 9.5 9.8 7.44 7.9 18.1 9.7

(m) 103 80 536 7.88 10.4 groundwater is larger than that of the deep groundwater

00 00 00 00 00 00 (Table 2). The water was slightly acidic with the well 0.8

0 depth increasing both in Songhua–Heilongjiang and Wus- 32.4 31.8 01.7 31.1 07.5 0 0 0 0 0 08 °

07 15 38 30 15 uli river basins (Fig. 5). The electrical conductivity of ° ° ° ° °

131 groundwater decrease with the well depth. The range and 133 133 133 133 132 standard deviation of shallow groundwater were larger than 00 00 00 00 00 00 deep groundwater. 15.2 27.9 01.7 16.9 19.7 13.5 0 0 0 0 0 0 0 ° 35 45 20 03 35 ° ° ° ° ° Stable isotopic composition Latitude N Longitude E Elevation

a The value of stable isotopic composition is sorted as lake water (d18 % [ d18 % type O =-6.0 ) wetland water ( O =-9.6 ) [ spring (d18O =-10.2 %) [ shallow groundwater (d18O continued =-10.8 %) [ river water (d18O =-11.2 %) [ deep groundwater (d18O =-11.9 %) (Table 2). The d18O value Elevation is the meter above sea level LW, RW, SGW, DGW, SP and WW stand forTemp. lake water, stands for river water water, temperature, shallow the groundwater, mean deep value groundwater, is spring shown and as wetland mean water, respectively 18 SHJ16D DGW 47 SHJ32 DGWSHJ30 46 SP 46 SHJ18D DGW 47 SHJ24D DGW 46 Mean DGW 89 SHJ10 WW 47 ‘‘–’’ indicates no data or below the limit of detection Table 2 Sample no. Sample a b c of surface water ranges from 13.9 to -5.9 %. The d O 123 Environ Earth Sci (2014) 72:3901–3913 3909

(a) 400 Flow direction 20 (b) 400 Flow direction 20 350 18 350 18 300 16 300 16

250 14 C) o C) 250 o 14 200 12 200 12 EC(µS/cm) 150 10 EC(µS/cm) 150 pH,Temp.( 10 pH,Temp.( 100 100 8 8 50 50 6 6 EC pH Temp. EC pH Temp. SHJ01 SHJ02 SHJ04 SHJ06,07,08 SHJ14 SHJ13 SHJ27 SHJ28 SHJ25 SHJ23 SHJ20 SHJ11 SHJ29 SHJ22 SHJ17 SHJ11 Songhua R. Heilongjiang R. Xingkai Lake Muling R. Naoli R. Sample No. Sample No.

Fig. 4 The variations of EC, pH and water temperature of surface water along flow direction

EC(µS/cm) EC(µS/cm) 200 400 600 800 1000 1200 200 400 600 800 1000 1200 (a) 0 0 (b) 0 0

20 20 20 20

40 40 40 40

60 60 60 60

80 80 80 80 Well depth (m) Well depth (m) 100 100 100 100 EC EC pH pH 120 120 120 120 Temp. Temp. 6 7 8 9 10111213 7 8 9 10 11 12 13 o pH,Temp.(oC) pH,Temp.( C)

Fig. 5 The EC, pH and water temperature values of groundwater versus well depth. a Songhua–Heilongjiang basin, b Wusuli river basin values of shallow groundwater and deep groundwater range -12.4 %) was the most depleted, which was sampled in a from -11.9 to -8.0 and -15.1 to -10.8 %, respectively. creek in the mountain area. The values of dD and d18O The d18O values of most groundwater were about -12 %. decreased along Muling river to Wusuli river. However, The variations of stable isotopic composition along flow the stable isotopes in the surface water were enriched along direction are shown in the Fig. 6. The stable oxygen iso- Naoli river caused by the water evaporation, but depleted tope enriched along the flow direction of Songhua river, after confluence with Wusuli river. and the range of values is from -11.4 to -10.7 %. The relationship between dD and d18O is shown in the However, the values of dD and deuterium excess decreased Fig. 7. The local meteoric water line (LMWL) was estab- along the flow direction, and the range of values is from lished by the monthly isotopic compositions in precipitation -84.8 to -85.7 %. The values of d18O in Heilongjiang at Sanjiang station. The LMWL is dD = 7.3 d18O–6.7. The river water increased after the confluence with Songhua slope and deuterium excess of LMWL is \8 and 10 of the river (the value ranges from -11.7 to -10.6 %), and the global meteoric water line (GMWL), respectively, indicat- value of dD increased along flow direction because of ing the large local humidity and evaporation in the Sanjiang evaporation (the value ranges from -90.3 to -87.6 %). plain. The linear fit of stable hydrogen and oxygen isotopes The isotopic composition in lake water was the most in surface water is dD = 5.7 d18O - 21.9. enriched (dD =-56.4 %, d18O =-6.0 %), while the The lake water and one shallow groundwater (SHJ26, value of d was the least. The stable isotopic composition in well depth was 3 m) fall on the upper right region of the the water sample SHJ22 (dD =-91.3 %, d18O = figure below the LMWL. The most depleted water (SHJ12,

123 3910 Environ Earth Sci (2014) 72:3901–3913

-2 20 (a) 18 δO δD d -4 -50

0 d(‰) -6 -20 -60 -8 δ D=5.7δ 18O-21.9 -10 -40 -70 -12 O(‰) -60 -80 18 δ

-14 δ D(‰) -80 -90 -16 LMWL -100 Lake water -18 SHJ06 D (‰)VSMOW -100 River water SHJ01 SHJ02 SHJ04 SHJ08 SHJ14 SHJ13 δ -20 SHJ07 -120 Shallow groundwater 400 300 200 100 0 -110 Deep groundwater Songhua River Flow direction Heilongjiang Spring Distance to confluence of Heilongjiang and Wusuli River (km) -120 Wetland water δ δ 18 20 LMWL: D=7.3 O-6.7 linear fit of surface (b) -2 δO18 δD d water

-4 0 d(‰) -16 -14 -12 -10 -8 -6 -4 δ18 -6 -20 O (‰) VSMOW -8 -40 18 -10 Fig. 7 The relationship between dD and d O in surface water and O(‰)

18 groundwater

δ -12 -60 δ

-14 D(‰) -80 Heilongjiang river (T content is 23.9 TU) (Fig. 8a). How- -16 SHJ20 SHJ11 -100 ever, the tritium concentration in the river water was not -18 SHJ27 SHJ28 SHJ23 SHJ17 Naoli-Wusuli SHJ25 SHJ29 SHJ22 SHJ19 Muling-Wusuli significantly variable along Wusuli river (the value ranges -20 -120 400 350 300 250 200 150 100 50 0 from 15.4 to 22.1 TU) (Fig. 8b). The tritium content in the Flow direction Distance to confluence of Heilongjiang and Wusuli river(km) shallow groundwater decreased progressively along Song- hua–Heilongjiang river (the value ranges from 20.4 to 18 Fig. 6 The variations of d O, dD and deuterium excess (d)of 1.7 TU) and Wusuli river flow direction (the value ranges surface water along flow direction from 19 to 1.8 TU), except the water sample SHJ33 (6.6 TU). The tritium concentrations in the deep ground- well depth was 100 m) falls at the left bottom of the figure, water samples were equal to the shallow groundwater indicating the stable isotopes in the recharge source water nearby. to this well were depleted. The most depleted river waters (SHJ07, SHJ08) were sampled in Heilongjiang river before CFC concentration in the shallow groundwater and after the confluence with Songhua river, respectively. The wetland water falls on the LMWL, indicating that the The CFC content in the shallow groundwater were deter- source of wetland water was precipitation. The spring mined. However, the concentration of CFC-11 was useless, water sample falls on the LMWL, surrounding by river which may be affected by sampling or local, near-surface water and shallow groundwater. sources. The average concentrations of CFC-12 and CFC- 113 were 0.53 and 0.20 pmol/kg water, respectively. The Tritium contents CFC-12 content in the water sample SHJ05 was the largest (1.25 pmol/kg), followed by the water sample SHJ09 The mean values of tritium concentrations in river water, (1.17 pmol/kg). The CFC-113 content in the water sample lake water, shallow groundwater, and deep groundwater SHJ03 was the largest with the value 0.71 pmol/kg, while were 22.7, 17.3, 13.6, and 8.3 TU, respectively. The tritium the second largest was water sample SHJ09 with the value content of shallow groundwater sample SHJ24 was the 0.44 pmol/kg. largest (61.2 TU). The tritium contents in deep ground- water samples SHJ32 (11.8 TU) were larger than those of the shallow groundwater SHJ33 (6.6 TU). However, the Discussion tritium values of the deep groundwater SHJ16D (1.0 TU) and SHJ18D (1.6 TU) were the least. Groundwater dating The tritium content in river water decreased along the flow direction of Songhua–Heilongjiang river (the value The groundwater dating by tritium concentration was cal- ranges from 30.8 to 19.5 TU), except the river waters culated by the Eqs. (2), (3) and (4). The weighted annual sampled at the confluence of Songhua river and average tritium concentration in precipitation was 123 Environ Earth Sci (2014) 72:3901–3913 3911

(a) 70 were modern waters (age \50 years) and were recharged Flow direction 60 River water since 1950 or in mixture of older water with post 1950 Shallow ground water 50 Deep ground water water (Table 2). Tritium concentrations in groundwater 40 06 08 30 02 were not typically affected by local sources or sampling 14 20 03 04 07 13 equipment. However, the CFC contents were sometimes Tritium (TU) 10 32 05 16 33 09 16D affected by those factors (Szabo et al. 1996). Consequently, 0 15 400 350 300 250 200 150 100 50 0 the ranges of CFCs’ ages (37–50 years) were larger than that of tritium ages of shallow groundwater (41–50 years). (b)70 Flow direction Lake water The mean value of deep groundwater dating by tritium 60 24 River water 50 Shallow ground water concentration was 47 years. The deep groundwater age was 40 Deep ground water larger than that of the shallow groundwater (46 years) 30 29 17 26 27 28 23 20 11 (Table 2). The deep groundwater samples SHJ32 Tritium (TU) 20 25 10 18 12 (45 years) and SHJ12 (45 years) were the youngest among 31 21 0 18D deep groundwater samples. The water sample SHJ32 400 300 200 100 0 Distance to the confluence of Helongjiang and Wusuli River (km) located in the mountain area. The hydraulic gradient of the groundwater is about 1/1,000 in the mountain area. How- Fig. 8 The variations of tritium concentration in water along flow ever, the groundwater hydraulic gradient is about direction of Songhua–Heilongjiang River (a) and Wusuli River (b) 1/5,000–1/10,000 in the central plain (Song et al. 2010). Consequently, the groundwater flow rate was higher in the 2500 mountain area than in the alluvial area. The deep ground- water samples SHJ16D ([50 years) and SHJ18D

2000 ([50 years) were sampled in the central area in the plain. The deep groundwater was old in the center of the plain because the groundwater flow rate in the mountain area 1500 was about 5–10 times than that in the center area. Considering the water infiltration as the piston flow, the 1000 well depths and shallow groundwater ages were used to calculate the vertical water flow rate. The range of infil- Tritium concentration (TU) 500 tration rate was 0.07–0.63 m/year. The vertical flow rate of the water sample SHJ26 was the least with the well depth 3 m and the groundwater age 42 years. The vertical infil- 0 tration rate of the water sample SHJ05 was the largest with 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 Year the well depth of 26 m and the groundwater age of 41 years. Fig. 9 The reconstructed tritium concentration in precipitation at Harbin Relationship between surface water and groundwater reconstructed using the data of global network of isotopes Two components approach was applied to calculate the in precipitation (GNIP) stations (Harbin: from 1986 to 1998 and Qiqihar: from 1988 to 1992, China). If the GNIP contributions of surface water and groundwater to the data was not available, the linear regression to Ottawa mixture water, respectively (Clark and Fritz 1997). (Canada) was used to calculate the tritium content in pre- Qt ¼ Qgw þ Qr ð5Þ cipitation (Wang 1998; Boronina et al. 2005). The recon- Qt dt ¼ Qgw dgw þ Qr dr; ð6Þ structed tritium concentration in precipitation at Harbin is shown in the Fig. 9. As the atmospheric concentration of where Q is the discharge component, d is the oxygen iso- tritium declined, its use for quantifying recharge by esti- tope d18O contents, and the subscripts represent total or mating the age of groundwater became less reliable. The mixture flow (t), groundwater (gw), and river flow (r), CFCs’ apparent age was analyzed based on its concentra- respectively. tion in water samples and air (International Atomic Energy The surface water level and groundwater table were Agency 2006). measured before water sampling. Along the flow direction Most shallow groundwater ages based on tritium con- of Songhua river, the altitude of the river water sample tents are the same as the CFCs concentrations. The SHJ02 (80 m) was lower than that of the water table of the groundwater ages indicated that the shallow groundwaters shallow groundwater sample SHJ03 (80.2 m). This

123 3912 Environ Earth Sci (2014) 72:3901–3913 indicates that the groundwater may flow into the river. The of tritium content and CFC concentrations was an effective river flow of upstream contributed 85 %, and the flow of method for groundwater dating. Most shallow groundwater shallow groundwater contributed 15 % to the river water ages based on tritium content were the same as the CFCs sample SHJ02. The water table of the shallow groundwater concentration. Most groundwaters were modern waters with SHJ05 (65.4 m) was higher than the river surface (59 m). the age \50 years. The groundwater ages in the mountain The shallow groundwater may flow into the river. The area were younger than in the central plain. Two-components shallow groundwater discharge contributed 43 % of the approach based on stable isotopes was applied to calculate the river discharge quantity, while the contribution of upstream exchange contribution between surface water and ground- river flow was 57 %. The Songhua river was a gaining river water. The Songhua, Heilongjiang, and Wusuli rivers were according to the analysis and the exchange contribution gaining rivers, but the shallow groundwater recharged from between surface water and groundwater. rivers at the confluence area of rivers. The combination At the confluence of Songhua river and Heilongjiang method of stable isotopes, tritium and CFCs could be applied river, the river water before and after joint were both in other watershed to study the groundwater ages and inter- sampled. The surface water level and groundwater table relation between surface water and groundwater. Practically, were also measured in situ. The altitudes of the river water to maintain the agricultural development and sustainable sample SHJ06, 07 and 08 were 54, 54 and 53 m, respec- water management, the farmlands near the river and under tively. The water table of the shallow groundwater SHJ09 foot of the mountain could be cultivated, but the farmlands in was 52 m. The shallow groundwater table was lower than the central plain should be controlled. river water level, indicating that the Songhua and Hei- longjiang rivers maybe seeps to shallow groundwater. The Acknowledgments This research was supported by the Main shallow groundwater was recharged about 12 % from Direction Program of Knowledge Innovation of Chinese Academy of Sciences: study on water cycle, water resource carrying capacity, and Heilongjiang river, and 88 % from Songhua river. The optimization configuration (No. KZCX2-YW-Q06-1); the Key Pro- water table of the shallow groundwater SHJ15 (47 m) was gram of National Natural Science Foundation of China: Hydrological higher than river water surface (45 m). The shallow cycle observing compare and contrast at different scales catchments: groundwater maybe flows into the river. The shallow A case study of Baiyangdian River basin (No. 40830636) and the National Natural Science Foundation of China (21307090). The groundwater discharged into Heilongjiang river in down- authors sincerely thank the editor and anonymous reviewers for stream; however, the river water seeped to shallow reviewing the manuscript and providing critical comments to improve groundwater at the confluence of Songhua and Heilongji- the paper. ang rivers. Along the flow direction of Wusuli river, the altitudes of the river water samples SHJ23 (42 m) and SHJ20 (43 m) References were lower than the shallow groundwater table of the water samples SHJ24 (49 m) and SHJ21 (48 m), respectively. The Anderson EI (2005) Modeling groundwater–surface water interac- comparison of surface water level and groundwater table tions using the Dupuit approximation. Adv Water Resour 28(4):315–327 indicates that the groundwater may discharge to the river. Banks EW, Simmons CT, Love AJ, Shand P (2011) Assessing spatial The contribution of the shallow groundwater (SHJ21) to the and temporal connectivity between surface water and ground- river flow (SHJ20) was 73 %, while the upstream river flow water in a regional catchment: implications for regional scale contributed 27 %. The river waters (SHJ11, 13) and the water quantity and quality. J Hydrol 404:30–49. doi:10.1016/j. jhydrol.2011.04.017 groundwater (SHJ12) were sampled at the confluence of Baskaran S, Ransley T, Brodie RS, Baker P (2009) Investigating Heilongjiang and Wusuli rivers. The altitudes of Heilongji- groundwater–river interactions using environmental tracers. ang river (SHJ13) and Wusuli river (SHJ11) surface were J Geol Soc Aust 56(1):13–19 44 m and 40 m, respectively, while the altitude of the Boronina A, Renard P, Balderer W, Stichler W (2005) Application of tritium in precipitation and in groundwater of the Kouris groundwater table of the water sample SHJ12 was 32.5 m. catchment (Cyprus) for description of the regional groundwater The comparison of river water level and groundwater table flow. Appl Geochem 20(7):1292–1308. doi:10.1016/j.apgeo indicated the rivers may lose water to groundwater. chem.2005.03.007 Clark ID, Fritz P (1997) Environmental isotopes in hydrology. Lewis publishers, New York Criss RE, Davisson ML (1996) Isotopic imaging of surface water Conclusions groundwater interactions, Sacramento Valley, California. J Hydrol 178:205–222 To study the relationship between surface water and He L (2000) China Sanjiang plain (in Chinese). Heilongjiang Science and Technology Press, Haerbin groundwater, the stable isotopes, tritium and CFC concen- Heilongjiang Geological Survey Hydrogeology Engineering Geology trations in water were analyzed in Sanjiang plain. The local Battalion (1959) The groundwater in Northeast Chian Plain (in meteoric water line is dD = 7.3d18O–6.7. The combination Chinese). Geological Publishing House, Beijing 123 Environ Earth Sci (2014) 72:3901–3913 3913

Heilongjiang Land Reclamation Bureau (2009) Statistical yearbook of Plain, Northeast China. J Environ Sci 22(7):968–974. doi:10. Heilongjiang reclamation area (in Chinese). China Statistics 1016/s1001-0742(09)60206-0 Press, Beijing Pereira LS, Oweis T, Zairi A (2002) Irrigation management under Huang N, Wang Z, Liu D, Niu Z (2010a) Selecting sites for water scarcity. Agric Water Manag 57(3):175–206 converting farmlands to wetlands in the Sanjiang plain, North- Seiler K-P, Lindner W (1995) Near-surface and deep groundwaters. east China, based on remote sensing and GIS. Environ Manag J Hydrol 165:33–44 46(790–800). doi:10.1007/s00267-010-9547-6 Song K, Liu D, Wang Z, Zhang B, Jin C, Li F, Liu H (2008) Land use Huang Y, Sun W, Zhang W, Yu Y, Su Y, Song C (2010b) Marshland change in Sanjiang plain and its driving forces analysis since conversion to cropland in northeast China from 1950 to 2000 1954 (in Chinese). Acta Geographica Sinica 63(1):93–104 reduced the greenhouse effect. Glob Change Biol Song X, Zhang B, Zhang Y, Han D (2010) The harmonious ideal 16(2):680–695. doi:10.1111/j.1365-2486.2009.01976.x land-natural environment and water resources development in Hunt RJ, Coplen TB, Haas NL, Saad DA, Borchardt MA (2005) Sanjiang plain, Northeast China(in Chinses). In: ICCS Journal of Investigating surface water–well interaction using stable isotope Modern Chinese Studies, vol 1. Aichi University, p 263–275 ratios of water. J Hydrol 302(1–4):154–172 Sophocleous M (2002) Interactions between groundwater and surface Hurtley S (2011) Ozone and rainfall. Science 332(6032):891. doi:10. water: the state of the science. Hydrogeol J 10(1):52–67 1126/science.332.6032.891-g Szabo Z, Rice DE, Plummer LN, Busenberg E, Drenkard S, Schlosser International Atomic Energy Agency (2006) Use of chlorofluorocar- P (1996) Age dating of shallow groundwater with chlorofluoro- bons in hydrology: a guidebook. International Atomic Energy carbons, tritium/helium: 3, and flow path analysis, southern New Agency, Vienna Jersey coastal plain. Water Resour Res 32(4):1023–1038. doi:10. Lu Y, Tang C, Chen J, Song X, Li F, Sakura Y (2008) Spatial 1029/96WR00068 characteristics of water quality, stable isotopes and tritium Thompson GM, Hayes JM (1979) Trichloromethane in groundwater: associated with groundwater flow in the Hutuo River alluvial fan a possible tracer and indicator of groundwater age. Water Resour plain of the North China Plain. Hydrogeol J 16:1003–1015. Res 15:546–554. doi:10.1029/WR015i003p00546 doi:10.1007/s10040-008-0292-3 Wang F (1998) The regional recovery model of tritium concentration Małoszewski P, Zuber A (1982) Determining the turnover time of for meteoric water in Jilin province (in Chinese). Jilin Geol groundwater systems with the aid of environmental tracers: 1. 17(3):75–81 Models and their applicability. J Hydrol 57(3–4):207–231 Wang S, Tian Y (2003) Preliminary research on groundwater table Maloszewski P, Zuber A (1991) Influence of matrix diffusion and change and causes in Sanjiang plain (in Chinese). J Irrig Drain exchange reactions on radiocarbon ages in fissured carbonate 22(2):61–64 aquifers. Water Resour Res 27(8):1937–1945 Winter TC, Harvey JW, Franke OL, Alley WM (1998) Ground water McGuire KJ, DeWalle DR, Gburek WJ (2002) Evaluation of mean and surface water: a single resource. Diane Pub Co, Denver residence time in subsurface waters using oxygen-18 fluctuations Woessner WW (2000) Stream and fluvial plain ground water during drought conditions in the mid-Appalachians. J Hydrol interactions: rescaling hydrogeologic thought. Ground Water 261(1–4):132–149 38(3):423–429 Morgenstern U, Stewart MK, Stenger R (2010) Dating of streamwater Zhang J, Ma K, Fu B (2009) Wetland loss under the impact of using tritium in a post nuclear bomb pulse world: continuous agricultural development in the Sanjiang Plain, NE China. variation of mean transit time with streamflow. Hydrol Earth Environ Monit Assess 166(1–4):139–148. doi:10.1007/s10661- Syst Sci 14(11):2289–2301 009-0990-x Oster H, Sonntag C, Mu¨nnich KO (1996) Groundwater age dating Zhou ZQ, Liu T (2005) The current status, threats and protection way with chlorofluorocarbons. Water Resour Res 32(10):2989–3001 of Sanjiang Plain wetland, Northeast China. J For Res Pan X, Yan B, Yoh M, Wang L, Liu X (2010) Temporal variability of 16(2):148–152 iron concentrations and fractions in wetland waters in Sanjiang

123