Origin and geochemical evolution of groundwater in a closed-basin clayey aquitard, Northern Mexico

Origin and geochemical evolution of groundwater in a closed-basin clayey aquitard, Northern Mexico Adria´n Ortega-Guerrero* Centro de Geociencias, Campus Juriquilla, Universidad Nacional Auto´noma de Me´xico, Apdo. Postal 1-742, Queretaro 76001, Mexico Received 19 October 2001; accepted 23 June 2003 Abstract Groundwater flow cross-sectional modeling in combination with major ion and trace element chemistry, stable isotopes of water (18O and 2H) and geochemical modeling were used to investigate the origin and evolution of groundwater in a Quaternary lacustrine clay-rich aquitard in the closed basin of the ‘La Laguna Region’, Northern Mexico. This study pertains to the Viesca Lake, one of a series of ephemeral ancient lakes that existed in this basin until the early 1900s. The Viesca aquitard overlies a regional marine Mesozoic sequence that outcrops in the surrounding mountains. Groundwater samples were collected from springs, pumping wells in a carbonate aquifer and from former industrial production wells in the aquitard. Numerical results indicate a classical gravity flow system, where the carbonate aquifer discharged about 90% of the groundwater, as springs, that fed the former Viesca Lake. Isotope content of groundwater, in the carbonate aquifer, shows that groundwater has a local meteoric origin; and its chemical evolution is mainly associated to progressive dissolution of gypsum and oxidation of pyrite along the flow paths. In contrast, groundwater in the aquitard presented enrichment of heavy isotopes 2H and 18O, indicating that this water resulted from the evaporation of the carbonate groundwater aquifer. The maximum concentrations of sulfate (77,700 mg L21), chloride (45,400 mg L21) and sodium (45,000 mg L21) in the groundwater from the aquitard were between three and four orders of magnitude greater than the fresh groundwater. Sulfate behaved as a conservative ion, contrasting with concentration of sulfate in groundwater from other aquitards in the world; which is usually reduced. Chemical patterns and enrichment of stable isotopes in groundwater from the aquitard follow general evaporation trends observed for other closed or semi-closed basins worldwide. Redox related parameters (NO3–N, NH4 þ, Mn2þ, total Fe and TOC) and redox-sensitive trace elements (As, Se, Sb, Sn, Cr, and Bi) in groundwater indicated dominant aerobic conditions in the carbonate aquifer and anaerobic conditions in the clayey aquitard. q 2003 Elsevier B.V. All rights reserved. Keywords: Aquitard; Brine-groundwater; Geochemistry; La Laguna Region; Carbonate aquifer; Sulfate 1. Introduction Quantifying groundwater origin and flow, hydro- geochemical evolution and solute transport in aquitards is important for their evaluation as sites for long term hazardous waste containment, protec- tive natural covers and geochemical influence to underlying regional fresh water aquifers. Clay-rich sediments are common around the world. In North America, most of the research on aquitards has focused on Quaternary clayey-rich glacial tills and Cretaceous clay, and there are few studies that Journal of Hydrology 284 (2003) 26–44 www.elsevier.com/locate/jhydrol 0022-1694/$ - see front matter q 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0022-1694(03)00239-7 * Corresponding author. E-mail address: [email protected] (A. Ortega- Guerrero). consider closed-basin Quaternary lacustrine clayey- rich aquitards; where, physical, chemical and biological processes in groundwater need to be evaluated under different geologic, hydrogeologic and climatic environments. Two regional aquitards exist in the La Laguna Region (LLR): the Viesca and Mayran aquitards, located at the northeastern most end of the Nazas- Aguanaval closed hydrologic basin of about 180,000 km2, where the former lakes of Viesca and Mayran developed (Fig. 1). The LLR (‘Region of perennial standing bodies of water’), is a large, alluvial plain located in the lower part of a closed hydrologic basin formed by the Nazas and Aguanaval rivers in the states of Durango and Coahuila in northern Mexico (Fig. 1). The LLR is a nearly flat plain at an average elevation of 1100 m above sea level, surrounded by folded Mesozoic marine rocks and, in less proportion, by Tertiary igneous rocks (Mayer, 1967). Fig. 1 shows the distribution of the mountainous areas and the dominant alluvial plain of the LLR. Average annual precipitation is 250 mm and potential evaporation is 2350 mm in this region. The ancient Viesca and Mayran Lakes (Fig. 1) were recognized as having highly saline water (Mayer, 1967; SARH, 1980). At the former Viesca Lake, high salinity and brine groundwater, sodium- sulfate type, was extracted for many decades, in the upper 100 m of the lacustrine aquitard by the company ‘Sulfatos de Viesca’. The origin of this high concentration of sulfate in groundwater from the Viesca aquitard is unknown. Despite the fact that much of groundwater in this region is unsuitable for domestic and agricultural use, there are not studies on the processes that control groundwater flow, geochemistry and solute trans- port in these fine-grained lacustrine sediments of the LLR. This study provides insight into the origin and salinity of groundwater in the Viesca aquitard under the influence of close-basin conditions in the former Viesca Lake, surrounded by Mesozoic sedimentary marine rocks. Specifically, this study addresses: (1) an analysis of the natural groundwater flow con- ditions between the Mesozoic mountains and the Viesca lacustrine aquitard, and (2) the environmental isotope and hydrogeochemical evolution of ground- water from the recharge zones to the aquitard final high salinity, sodium-sulfate groundwater. It makes use of existing hydraulic data in the carbonate Fig. 1. Location of the Viesca basin within the Nazas and Aguanaval rivers Basin. A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–44 27 aquifer from government agencies, and hydraulic and general chemical data from Sulfatos de Viesca, a company that extracted brine from the aquitard for industrial purposes. The work presented in this paper represents a first stage of a long-term study of the origin of solutes and mechanisms for the transport and the hydraulic aquitard-aquifer interaction in the LLR. 2. The hydrogeology of the study area The alluvial sedimentation of the Nazas and Aguanaval rivers formed an important unconfined granular aquifer towards the middle of the plain and two aquitards in the more distant and lower parts of the LLR, which corresponds to the ancient lakes of Viesca and Mayran (Fig. 1). La Laguna Region is surrounded by folded Jurassic and Cretaceous marine rocks described in Table 1 (Mayer, 1967). The Cretaceous formations of Cupido and Aurora are the main carbonate aquifer in the LLR; they extend from the mountains towards the bottom of the granular aquifer at variable depths (Fig. 1) (SARH, 1980; CNA, 1993). This study focuses on the Viesca aquitard, which increases gradually in thickness from the periphery of the former lake to a maximum of 350 m in the central area at site DEW (Sulfatos de Viesca, 1971) (Fig. 2a and b). Local stratigraphy of the fine-grained lacustrine sequence, in the upper 60 m, is presented in Fig. 2c. The regional carbonate aquifer, in the Cupido and Aurora Formations, beneath and beyond the Viesca Plain is the most recent of the major aquifers in the LLR to be exploited. In the periphery of the Viesca plain, the carbonate aquifer provides fresh water for drinking and for the development of small-scale agriculture (Fig. 2a). Important groundwater dis- charge areas, as springs, were located in the transition between the carbonate aquifer and the lacustrine aquitard (Fig. 2a) and disappeared about 50 years ago as a consequence of extensive groundwater exploita- tion in the LLR (SARH, 1980). The northern area of the Viesca Plain contains soils rich in salts and almost total absence of vegetation. In contrast, in the southern area, salt-tolerant vegetation grows. Fig. 2c shows the results of sediment chemistry of the lacustrine sediments at site IW1 (Sulfatos de Viesca, 1971), which show the persistence of high concentrations of sodium and sulfate with depth. Natural manifestations of the groundwater conditions prior to aquifer exploitation, in addition to soil salinity and vegetation are considered for a quantitative interpretation of the groundwater flow system as suggested by To´th (1966, 1999). 3. Methods and materials 3.1. Groundwater flow modeling The numerical model CROSSFLO (Mc Laren, 1988) was used for the regional flow modeling. This model directly solves for potentials and stream functions based on the theory of Frind and Matanga (1985); Frind et al. (1985). For the purpose of this work and the scale of analysis of groundwater flow, the influence of density flow in the aquitard is not considered. A hydrogeological cross-section oriented along the assumed principal direction of hydraulic Table 1 Mesozoic lithologic formations in the La Laguna Region Formation Composition Thickness (m) Period Zuloaga Limestone 500 Jurassic La Gloria Sandstone and limestone 500 Jurassic La Casita Sandstone, limestone and gypsum 60–70 Jurassic Taraises Sandstone, mudstone and siltstone 250 Cretaceous Las Vigas Sandstone and shale 150 Cretaceous Cupido Limestone 550 Cretaceous La Pen˜a Mudstone and shale 150 Cretaceous Aurora Limestone 550 Cretaceous Indidura Shale and mudstone 70 Cretaceous A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–4428 Fig. 2. (a) Hydrogeologic units and location of groundwater sampling points. (b) Hydrogeologic cross-section. (c) Distribution of NaCl and Na2SO4 (percent per weight) based on chemical determinations of the lacustrine sediments, and local stratigraphy at the IW1 site (After Sulfatos de Viesca, 1971). A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–44 29 permeability was constructed, based on the regional geological information supplied by Mayer (1967). Hydraulic conductivity values, for the numerical analysis, were collected from available tests in the industrial area of Sulfatos de Viesca and from production wells in the carbonate aquifer obtained from government agencies SARH (1980) and CNA (1993). 3.2. Groundwater sampling Water samples for isotope and major ion analysis were collected in February 1998. From a total of fifteen samples, six of them were obtained from brine production wells in the aquitard (IW1–IW6), using a 2-liter polyethylene bailer in the upper 20 m, with water tables at depths of between 35 and 10 m Fig. 2 (continued ) A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–4430 (Fig. 2a). About 5–10 m of the water column in the boreholes were extracted and additional two to three episodes of borehole purging were conducted, before sample collection. Water samples were also collected from one spring (S1) and from six water production wells drilled in the upper 200–300 m depths of the limestone aquifer (W1–W6), at the edges of the lacustrine plain with the surrounding mountains; two shallow wells, about 30 m depth, constructed in an alluvial fan overlying the lacustrine aquitard (AF1 and AF2), were also sampled (Fig. 2a). During sampling, pH, temperature, alkalinity and electrical conductance measurements were made in the field. All water samples for chemical analysis were filtered through 0.45 mm pore-size filters. Water samples were separated in three aliquots. One was acidified with concentrated HNO3 to a pH , 2 for cation determinations, while the second one was left unacidified for anion analysis. The third aliquot was used for stable isotope analysis. 3.3. Isotope and chemical analysis Groundwater samples for chemical determinations were analyzed at MSD Environment Services Limited in Toronto, Canada, within a period of one week following their collection. Alkalinity was measured by volumetric titration. Samples were analyzed by ion chromatography and automated colorimetry US EPA methods No. 300, 350.1, 354.1, 353.1, 365.1, 375.4. Metals were analyzed by ICP-MS, ICP-AES, US EPA No. 200.7 and 200.8 (modified). Ammonium was determined by colorimetry in continuous liquid flow. Carbonate content was calculated based on alkalinity, pH and bicarbonate content US EPA No. 150.1, Standard Methods No. 232B and 4500 CO2B. All water compositions have an ionic balance within 5%, except samples AF2, IW2, IW3, IW6 that are below 8% error (Table 2). Stable isotope determinations for d18O and d2H, were carried out at the Environmental Isotope Laboratory at the University of Waterloo. Isotopic data is reported with respect to SMOW (Standard Mean Ocean Water), and the analytical reproduci- bility was better than 0.2‰ and 2‰ for d18O and d2H, respectively. 3.4. Geochemical modeling Saturation indices for various minerals that may be associated with the groundwater flow in the main carbonate aquifer and in the lacustrine aquitard were calculated using the computer code PHRQPITZ (Plummer et al., 1988) based on PHREEQE (Parhurst et al., 1980). 4. Results and discussion 4.1. Groundwater flow modeling 4.1.1. Modeling strategy The distribution of the hydrogeological units in the study area is shown in Fig. 2a and b. A cross-section in the Viesca basin from the mountain ranges into the valley was selected for the 2-D modeling analysis (Fig. 2b). This section present the following charac- teristics: (1) it is perpendicular to regional contours of hydraulic head in the carbonate aquifer (CNA, 1993); (2) it passes through areas that have the highest soil salinity, which are qualitatively considered the main groundwater discharge zone of the flow system, based on field evidences as suggested by To´th (1999); and (3) it passes through points where the hydrogeological information is best known (Fig. 2a). The difference in topography of Sierra Parras (that includes Sierras Las Buras and Las Pen˜as), to the north, and Sierra La Cadena (that incudes Sierras Alamos, Presitas, Tres Flores and Paredones) to the south, required that these sections included both sides of the mountain massifs. Both Sierras are symmetric massifs with a well- defined height that is assumed to be the groundwater divide. It is also assumed that the groundwater flow system is steady state, that the carbonate aquifer and aquitard are heterogeneous and anisotropic, and that the fluid in the aquifer is considered isothermal, dilute and incompressible. Density driven flow, due to high salinity in groundwater from the aquitard, is not considered in this regional scale analysis. The unsaturated zone is assumed to be in dynamic equilibrium with the water table. The carbonate aquifer in the folded and fractured Cretaceous rocks is represented as an equivalent porous medium. The finite element grid and the boundary conditions for potentials and stream functions are presented in A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–44 31 Table 2 Results of the chemical and isotopic analysis in groundwater samples: From the spring (S1), the carbonate aquifer (W1–W6) and from the clayey aquitard (IW1–IW6). The log of the Saturation Indices for calcite, dolomite, gypsum, halite and miralite obtained by the geochemical model PHRQPITZ are also shown Sample ID T (8C) pH field EC pH lab Na (mg L21) K (mg L21) Ca (mg L21) Mg (mg L21) T Hardness (mg L21) ALK CaCO3 (mg L21) HCO3 (mg L21) CO3 (mg L21) SO4 (mg L21) CI (mg L21) SiO2 (mg L21) Phosphate (mg L21) NO2 2 N (mg L21) NO3–N (mg L21) NH4 (mg L21) S1 26.9 7.74 4.52 7.9 19 1.5 77 27 303 208 206 2 112 18 14.7 0.06 0.02 0.24 0.05 W1 26.5 7.49 1090 7.92 110 5.3 100 34 390 182 181 1 293 71 30.5 0.44 0.02 3.79 0.09 W2 27.3 7.41 1680 7.68 128 4.8 217 52.8 759 24 24 0 628 186 4.48 0.01 0.02 0.13 0.05 W3 24.5 7.51 1820 7.79 340 12 210 56 755 174 173 1 786 433 27.8 0.1 0.03 4 0.05 W4 32.3 7.31 4840 7.64 750 26 390 88 1340 170 169 1 1850 914 26.9 0.13 0.02 1.13 0.05 W5 24.5 7.31 4920 7.61 880 16 240 92 978 147 146 1 1240 1160 1.27 0.3 0.02 0.1 0.17 W6 27.5 7.04 5510 7.83 776 9.4 647 184 2370 108 107 1 2320 761 21.7 0.1 0.02 0.1 0.05 AF2 23.8 7.11 12,300 7.92 2300 73 660 180 2390 150 149 1 3700 3230 30.4 0.34 0.02 0.16 0.05 AF1 21.9 7.53 6830 7.91 1500 62 600 130 2030 148 147 1 3910 1260 78.3 0.32 0.02 0.37 0.07 IW1 23.6 7.23 65,600 7.78 27,000 500 520 2300 10,800 1150 1140 6 43,500 23,000 7.55 3.1 0.03 0.1 19.2 IW2 19.2 7.12 74,900 7 78 39,000 1000 460 3000 13,500 1060 1050 6 77,700 24,300 78.5 3.8 0.02 0.1 2.85 IW3 21.1 6.87 76,700 7.73 36,500 1200 440 2600 11,800 990 985 5 60,500 31,300 66.4 3.5 0.03 0.1 1.59 IW4 22 7.4 77,800 7.76 37,000 1300 510 3200 14,500 982 977 5 51,700 27,300 55.1 3.8 0.02 0.1 1 28 IW5 22.1 8.48 92,800 8.24 43,000 530 480 2200 10,300 286 281 5 41,500 45,000 74.2 2.2 0.02 0.1 0.84 IW6 23.1 9.07 93,000 8.63 45,000 520 170 1600 7010 340 319 20 55,500 45,400 2.43 1.4 0.02 0.1 1.57 Sample ID NO3–N (mg L21) NH4 (mg L21) TOC (mg L21) FeT (mg L21) Mn (mg L21) Cu (mg L21) Zn (mg L21) Al (mg L21) Sb (mg L21) As T (mg L21) Ba (mg L21) Be (mg L21) Bi (mg L21) B (mg L21) Cd (mg L21) Cr T (mg L21) Co (mg L21) Pb (mg L21) Mo (mg L21) S1 0.24 0.05 0.5 0.02 0.01 0.01 0.11 0.01 0.002 0.002 0.055 0.005 0002 0.05 0.0001 0.002 0.001 0.0001 0.011 W1 3.79 0.05 0.5 0.029 0.01 0.01 0.029 0.01 0.002 0.049 0.027 0.005 0.002 0.36 0.0001 0.002 0.001 0.0002 0.012 W2 0.13 0.05 4.9 0.01 0.01 0.001 0.001 0.001 0.0002 0.0002 0.007 0.0005 0.0002 0.064 0 0.0002 0.0001 0 0.005 W3 4 0.05 0.5 0.063 0.01 0.01 0.015 0.01 0.002 0.097 0.023 0.005 0.002 0.43 0.0001 0.002 0.001 0.0005 0.015 W4 1.13 0.05 1.1 0.025 0.01 0.01 0.034 0.011 0.002 0.23 0.022 0.022 0.005 0.66 0.0001 0.002 0.001 0.0001 0.026 W5 0.1 0.17 4.2 0.02 0.18 0.01 3 0.01 0.002 0.004 0.028 0.005 0.002 0.66 0.0001 0.002 0.001 0.0006 0.018 W6 0.1 0.05 0.5 0 0 0 0 0.001 0.0002 0.25 0.01 0.0005 0.0002 0.4 0 0.0002 0.0001 0 0.01 AF2 0.16 0.05 25 0.035 0.22 0.022 4.6 0.025 0.005 0.11 0.016 0.005 0.005 2 0.0003 0.005 0.001 0.0018 0.15 AF1 0.37 0.07 5.3 0.026 0.044 0.017 0.029 0.025 0.005 0.21 0.017 0.005 0.002 0.98 0.0003 0.005 0.001 0.0007 0.053 IW1 0.1 19.2 38.5 0.17 0.31 0.01 0.02 0.1 0.02 0.095 0.02 0.02 0.02 7.2 0.001 0.02 0.003 0.0032 0.025 IW2 0.1 2.85 25 0.027 0.1 0.01 0.02 0.1 0.02 1.7 0.05 0.02 0.02 13 0.001 0.02 0.003 0.0034 1.3 IW3 0.1 1.59 25 0.33 0.04 0.01 0.035 0.01 0.02 3.1 0.02 0.02 0.02 14 0.001 0.02 0.003 0.0039 0.49 IW4 0 1 1.28 25 0.025 0.33 0.01 0.02 0.1 0.02 0.29 0.05 0.02 0.02 13 0.001 0.02 0.003 0.0026 0.32 IW5 0.1 0.84 25 0.1 1.2 0.025 0.04 0.2 0.04 1 0.1 0.04 0.04 11 0.002 0.04 0.006 0.0006 1.9 IW6 0.1 1.57 25 0.1 0.68 0.036 0.04 0.2 0.04 0.38 0.1 0.04 0.04 7 0.002 0.04 0.006 0.0042 2.8 Sample ID Ni (mg L21) Se (mg L21) Ag (mg L21) Sr (mg L21) Tl (mg L21) Sn (mg L21) Ti (mg L21) U (mg L21) V (mg L21) Deuterium (‰) Oxygen-l8 (‰) log SI CALCITE log SI DOLOMIT log SI GYPSUM log SI HALITE log at MIRABILIT Ionic St S1 0.002 0.002 0.00005 1.3 0.0001 0.002 0.002 0.0033 0.002 267.5 210.22 0.39 0.84 21.53 28.04 28.317 0.011 W1 0.002 0.005 0.00005 2.7 0.001 0.002 0.002 0.0051 0.005 262 28.31 0.11 0.27 21.11 26.7 26.424 0.018 W2 0.002 0.002 0 5.9 0 0.0002 0.006 0 2.0002 20.6 21.27 20.62 26.25 26.111 0.03 W3 0.002 0.002 0.00005 5 0.0006 0.002 0.002 0.0063 0.004 262.5 28.49 0.16 0.25 20.61 25.47 25.089 0.04 W4 0.002 0.002 0.00027 8.5 0.0004 0.002 0.006 0.0043 0.002 266.5 29.02 0.27 0.49 20.21 24.87 24.535 0.09 W5 0.005 0.002 0.00005 10 0.0011 0.002 0.004 0.001 0.002 267 29.4 20.05 20.011 20.55 24.67 24.226 0.08 W6 0.0002 0.002 0 11 0 0.0002 0.021 0.004 0.0002 20.07 20.14 0.01 24.93 24.279 0.13 AF2 0.005 0.005 0.0002 16 0.0015 0.005 0.009 0.0099 0.006 259 28.53 0.34 0.52 0.1 24.58 23.546 0.17 A . O rteg a -G u errero / Jo u rn a l o f H yd ro lo g y 2 8 4 (2 0 0 3 ) 2 6 – 4 4 3 2 Fig. 3a. The boundary condition along the water table (upper boundary) is a flux boundary and the position of the water table in the mountains adjusts to the input flow, because the position is unknown a priori. For potentials, the specified values are used directly as a head boundary (first-type or Dirichlet boundary condition); while for stream function, gradient in the normal direction is considered, in the form of a second-type Neumann boundary. The reference for datum elevation for potentials is sea level. A constant head boundary of 1080 m—the approximate elevation of the ancient lake surface—was specified for the upper boundary at the lacustrine plain. 4.1.2. Hydraulic conductivity The hydraulic conductivity obtained from about ten industrial production wells in the aquitard (Sulfatos de Viesca, 1971), have a mean value of 1.0 £ 1028 m/s in the upper 50 m. of the lacustrine sequence, indicating the presence of heterogeneities within the lacustrine sequence and perhaps the formation of micro-fractures induced by aquitard pumping. These range of values are consistent with values of hydraulic conductivity influence by dis- continuities (micro-fractures or interbbeding of thin sandy layers) reported in other aquitards in the world (Van der Kamp, 2001). The hydraulic conductivity of the limestone aquifer from wells W1, W3, W4 and W6 (SARH, 1980 and CNA, 1993) is 15.0 £ 1026 m/s on average. A sensitivity analysis was performed for the recharge from precipitation and for the bulk hydraulic conductivities of the hydrogeological units. Two different scenarios are presented in the modeling analysis results: one in which the relative aquifer/ aquitard hydraulic conductivity ratio is 1/1.0 £ 102 as compared with the field data; while the second scenario is 1/1.0 £ 104, considering that below the 50 m depth, the hydraulic conductivity must decrease with depth, reaching a more realistic value for a non- fractured clayey aquitard of 1 £ 10210 m/s (Van der Kamp, 2001). 4.1.3. Flow nets With respect to the cross-section (Fig. 3b–e), the flow nets and the distribution of the velocity vectors indicate areas of downward components of groundwater in the mountains and areas of upward groundwater flow into the lacustrine plain that areAp 1 0 .0 0 5 0 .0 0 5 0 .0 0 0 2 1 2 0 .0 0 2 5 0 .0 0 5 0 .0 0 8 0 .0 0 8 1 0 .0 1 4 2 6 3 .5 2 9 .2 2 0 .0 1 2 0 .0 7 0 .0 6 2 3 .9 3 2 3 .2 3 2 0 .2 1 IW 1 0 .0 2 0 .0 2 0 .0 0 0 5 1 5 0 .0 0 1 0 .0 2 0 .0 2 0 .0 0 5 6 0 .0 2 2 4 3 2 3 .9 0 .4 3 2 .1 0 .1 1 2 2 .1 4 2 0 .8 8 3 2 .1 4 IW 2 0 .0 2 0 .0 2 0 .0 0 0 5 1 2 0 .0 0 1 0 .0 2 0 .0 2 0 .0 1 2 0 .0 2 2 4 1 2 2 .7 2 0 .0 9 1 .5 9 0 .1 3 2 2 .0 6 2 0 .4 7 1 3 .2 2 IW 3 0 .0 2 0 .0 2 0 .0 0 0 5 1 2 0 .0 0 1 6 0 .0 2 0 .0 2 0 .0 0 1 0 .0 2 2 4 4 2 2 .5 2 2 0 .0 8 1 .2 1 0 .0 9 2 1 .8 9 2 0 .5 7 8 2 .9 2 IW 4 0 .0 2 0 .0 2 0 .0 0 0 5 1 2 0 .0 0 1 0 .0 2 0 .0 2 0 .0 0 1 9 0 .0 2 2 4 3 2 2 .9 4 0 .4 5 2 .2 9 0 .0 3 2 1 .8 4 - 2 0 .5 0 6 2 .8 6 IW 5 0 .0 4 0 .0 4 0 .0 0 1 1 0 0 .0 0 2 0 .0 4 0 .0 4 0 .0 0 2 0 .0 4 2 3 2 0 .6 9 0 .9 6 3 .1 9 0 .0 2 2 1 .5 1 2 0 .5 0 9 2 .9 7 IW 6 0 .0 4 0 .0 4 0 .0 0 1 3 .5 0 .0 0 2 0 .0 4 0 .0 4 0 .0 1 0 .0 4 2 3 5 0 .6 6 5 .8 8 3 .3 6 2 0 .3 2 2 1 .5 5 2 0 .4 9 4 3 .2 3 A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–44 33 compatible with the observed discharge zones of former springs. The more realistic scenario of recharge is when a 10% of average precipitation in the Sierra La Cadena and Parras is considered. The model produces a classic gravity groundwater flow system that follows general trends developed for other groundwater basins in the world. (To´th, 1966; 1997; Ortega and Farvolden, 1989). The transition between the downward and upward flow occurs in the lateral geologic contact between Fig. 3. (a) Finite element grid and boundary conditions for the modelling analysis. (b) Flow nets from modelling results (potential (F) and stream function (C)) with a hydraulic conductivity contrast between the aquifer and the aquitard of two orders of magnitude, and (c) the groundwater velocity field distribution for this scenario. (d) Flow nets from modelling results (potential (F) and stream function (C)) with a hydraulic conductivity contrast between the aquifer and the aquitard of four orders of magnitude, and (e) the groundwater velocity field distribution for this scenario. A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–4434 the Cretaceous rocks and the lacustrine aquitard, which contrast in hydraulic conductivity, where shallow and deep groundwater flow lines converge, causing that more than 90% of the groundwater discharge occurred as springs in the past. This indicates that most of the former Viesca lake was feed by groundwater from the carbonate aquifer. The main groundwater velocity occurs in the unconfined part of the carbonate aquifer. This groundwater flow conditions would be important on the interpretation of isotope and concentration of salts from deeper formations of the Mesozoic sequence. A contrast of four orders of magnitude in hydraulic conductivity between the carbonate aqui- fer and the aquitard, allows groundwater from Sierra La Cadena to flow underneath the aquitard and discharge near the limits of the Sierra de Parras (Fig. 3d–e). This former case may explain a contrast in salinization between the northern and southern areas of the aquitard that is described below. Present groundwater conditions in the Viesca area indicate piezometric level in the aquitard between 3 and 5 m; whereas in the carbonate aquifer, they are between 50 and 100 m based on the regional trends in the last two decades. Present groundwater interaction between the carbonate aquifer and aquitard is unknown due to the absence of detailed instrumenta- tion with depth in the aquitard to measure ground- water flow direction and its transient interaction with the aquifer. 4.2. Environmental isotopes The isotope composition of groundwater in the carbonate aquifer plots along the Global Meteoric Water Line (GMWL), with equation d2H ¼ 8 d18O þ 10 (Craig, 1961), indicating its meteoric origin (Fig. 4). This isotope composition also shows a narrow range, between 28‰ and 210‰ for d18O and 260‰ and 265‰ for d2H (Table 2 and Fig. 4). This narrow range indicates that the carbonate aquifer in the Viesca area should have the same general Fig. 3 (continued ) A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–44 35 recharge altitude, which is consistent with the similar elevation of the water table in the mountains surrounding the Viesca plain, as obtained with the flow model. The isotopic content of groundwater in the carbonate aquifer shows no influence of evaporites from the deeper formations in the Mesozoic sequence. In contrast, the isotope composition of ground- water in the aquitard, plotted in the d18O versus d2H diagram, shows a different pattern than the isotope composition in the carbonate aquifer (Fig. 4). These data define a line with a slope lower than the GMWL, with a regression equation of d2H ¼ 3 d18O 2 35. This line intersects the GMWL at the mean isotopic composition of the carbonate aquifer water (29.02‰ for d18O and 264‰ for d2H), indicating that the water in the carbonate aquifer represents the average isotopic composition of the water source of the former Viesca Lake. Therefore, this line can be interpreted as evaporation line instead a mixing line between fresh water and a saline. These changes in isotopic pattern are typical for waters that have been affected by evaporation in closed-basin lakes (Kharaka and Carothers, 1986) and in fossil salt lakes where isotope composition remains in the pore water of the low permeability sediments (Ortega et al., 1997). This last situation seems to prevail in the Viesca aquitard. The isotopic behavior of the groundwater in the Viesca aquitard sediments indicates that this water was originally meteoric water that reached the former lake as groundwater, primarily in the form of springs. This evidence is consistent with the results of the groundwater flow modeling analysis presented above, where about 90% of groundwater discharged as springs. Once in the former lake this water undergone evaporation, perhaps as sedimentation of fine grain particles occurred, remaining as groundwater. Present hydraulic heads in the aquitard and carbonate aquifer would permit groundwater flow from the aquitard into Fig. 4. Plot of oxygen-18 versus deuterium in groundwater from the carbonate aquifer and clayey aquitard. A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–4436 the aquifer, process that is not yet evident based on the isotopic content in the samples collected from the carbonate aquifer. 4.3. Geochemistry Groundwater chemical compositions are given in Table 2 and sample locations in Fig. 2a. Groundwaters were analyzed with emphasis on concentrations of chloride with respect to major ions and trace elements. Chloride was emphasized because it is the least reactive of the major ions and is not precipitated except at very high salinity (Jones et al., 1977). The electrical conductance of groundwater, as an indication of the total dissolved ionic constituents of the water, is the lowest in the sample S1 is 452 microsiemens (mS) that corresponds to the spring water. In the carbonate aquifer, water electrical conductance ranges between 1000 and 5000 mS with an extreme value of 5510 mS at site W6; whereas, in the aquitard electrical conductance concentrations were obtained in the range of 74,000 and 93,000 mS at sites IW1 to IW6. Electrical conductance in the thin alluvial fan is between 6840 and 12,300 mS, showing the influ- ence of the aquitard below (sites AF1 and AF2). The alluvial fan sites represent local hydrogeologic conditions that are not considered in the rest of the geochemical analysis. The chloride concentrations in the carbonate aquifer range between 70 and 3500 mg L21 and in the aquitard between 24,000 and 45,500 mg L21, whereas the concentrations of Naþ vary from 110 to 2000 mg L21 in the carbonate aquifer and from 24,000 to 43,000 mg L21 in the aquitard. Chloride and Naþ in groundwater from the carbonate aquifer and the lacustrine aquitard follow a 1:1 molar ratio line, over a concentration range of four orders of magnitude with almost no scatter (Fig. 5a). The concentrations of Kþ range between 12 and 520 mg L21, more than a 40-fold variation (Fig. 5b). Most of the Cl2 and Kþ concentration points lie below the 1:1 molar ratio line, where the Kþ/Cl2 ratio is near 0.025. The concentrations of Mg2þ (Fig. 5d) for the more diluted samples plot near the 1:1 molar ratio line, and as the samples get enriched in Cl2 the concentrations of Mg2þ progressively reach a Mg2þ/Cl2 ratio near 0.1. The plots of Cl2 versus Ca2þ (Fig. 5c) and HCO3 2 (Fig. 5f) show a much different relation than that of the previous ions; where concentration of Ca2þ and HCO3 2 remain within a range of concentration of 1 £ 102 mg L21 and 1 £ 103 mg L21 independently of the enrichment of Cl2. Sulfate concentrations range from 293 to 55,500 mg L21; a three orders of magnitude variation (Fig. 5e). The Cl2 versus SO4 22 plot shows that most of the concentration points lie on the 1:1 molar ratio line with almost no scatter, indicating that sulfate behaves, in the Viesca aquitard as a conservative ion, contrasting with concentration of sulfate in ground- water from aquitards in the world that is usually reduced; fact that is analyzed in the discussion section. Further evidence that supports the role of evapor- ation as the main process responsible for the high salinity of groundwater in the aquitard is presented in Fig. 5g and h that show a clear correspondence between chloride and sulfate with the d18O. Sulfate is the dominant anion in both the aquifer in Cretaceous rocks and in the lacustrine aquitard. The log of the saturation indices (log SI) obtained from the equilibrium speciation model PHRQPITZ (Table 2), for gypsum (CaSO4 2H2O) in groundwater from the carbonate aquifer, range from 21.53 to 20.21, indicating that gypsum is undersaturated in the groundwater, except for sample W6 that is super- saturated (þ0.01). Whereas the log SI in groundwater from the aquitard range between þ0.02 and þ0.13, indicating that gypsum in the groundwater from the aquitard is supersaturated, except for sample IW6 that is undersaturated (log SI ¼ 20.06), and that has a lower concentration of Caþþ compared with the other samples in the aquitard. These results are consistent with the presence of thin layers of gypsum in the lacustrine sequence, reported during drilling (Sulfatos de Viesca, 1971). The log SI for mirabilite (Na2SO4 10H2O) is undersaturated in the groundwater; it varies from 28.317 in the spring water to 2 4.711 in the aquitard. Halite (NaCl) is always undersaturated (Table 2). Variation in redox related parameters are presented in Fig. 6. Nitrate (NO3–N) concentrations are in the range of 0.0–4 mg L21 (Fig. 6a) in the carbonate aquifer, and low total Fe, indicating aerobic con- ditions below a concentration of about 1000 mg L21 A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–44 37 Fig. 5. Plots of Cl concentration and other major ions: (a) Na, (b) K, (c) Ca, (d) Mg, (e) SO4, (f) HCO3, and relationship of d 18O with the more abundant anions (g) Cl and (h) SO4. A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–4438 Fig. 6. Plots of Cl against redox related parameters: (a) NO3, (b) NH4, (c) Fe T, (d) Mn, and other trace elements: (e) Sr, (f) B, (g) Ba and (h) Pb. A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–44 39 Fig. 7. Plots of Cl against redox-sensitive trace elements: (a) As T, (b) Se, (c) Sb, (d) Sn, (e) Cr T, (f) U, in addition to (g) TOC and (h) Bi. A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–4440 for Cl2. In contrast, above that Cl2 concentration, anaerobic conditions dominate as shown by high NH4 þ and Mn2þ concentrations (Fig. 6b and d) and higher total Fe (Fig. 6c), except for wells IW1 and IW2 that present low Fe. A similar behaviour is observed for the concentrations of TOC that are below 10 mg L21 in the carbonate aquifer and above 10 mg L21 in the aquitard (Fig. 7g). A redox boundary is, therefore, defined between the carbonate aquifer and the aquitard. Redox sensitive trace elements: As (Fig. 7a), Se (Fig. 7b), Sb (Fig. 7c), Sn (Fig. 7d), Cr (Fig. 7e), and Bi (Fig. 7h), are also very low in the carbonate aquifer and high in the groundwater from the aquitard, indicating the dominant aerobic and anaerobic conditions, respectively. Uranium (Fig. 7f) and Sr (Fig. 6e) show a different trend, where they progressively increase as Cl concentration increases in the groundwater from the carbonate aquifer and progressively decrease concentration in the ground- water from the aquitard, perhaps precipitating to a solid phase under extreme brine concentration. Concentration of B (Fig. 6f) and Pb (Fig. 6h) behave conservatively in the carbonate aquifer and the aquitard where progressively increase as Cl2 concentration increases too. High concentration of As in groundwater at the study area is of particular interest because it represents endemic poisoning in most of the LLR (Chavez et al., 1964; Cebrian et al., 1983). Concentration of As in the groundwater in most of the sampled wells is above 10 mg/l, and in the aquitard is 2.5 orders of magnitude higher. The analysis of the origin and behavior of As in groundwater in other parts of the LLR is the focus of present research. 5. Discussion Groundwater modeling results indicate that under natural groundwater flow conditions, about 90% of groundwater from the carbonate aquifer discharged in the boundary between the carbonate aquifer and the lacustrine aquitard mainly as springs, the other 10% would discharge through the lacustrine aquitard. This indicates that most of the former Viesca lake was fed by groundwater from the carbonate aquifer mainly as springs. Additionally, isotopic composition of groundwater in the carbonate aquifer provides evidence that it is derived from local rainfall. Redox related parameters and redox sensitive trace elements in groundwater indicate aerobic conditions in the carbonate aquifer, consistent with local recharge. In contrast, salt concentrations in groundwater in the Viesca aquitard are between three and four orders of magnitude higher than in the more diluted water from the spring and from the carbonate aquifer. The concentrations of salts in the groundwater of the aquitard follow evaporation concepts developed for other closed or semi-closed basins in the world. The chemical trends are also consistent with the enrich- ment of oxygen-18 and deuterium results. Therefore, the isotope and chemical patterns in the Viesca aquitard are mainly a reflection of evaporation processes that has affected the paleo-lake water. Persistence of this evaporated water in the aquitard indicates contemporaneous sedimentation of the clayey sediments in this arid basin during geologic time. The other processes that should influence the isotopic composition of groundwaters during their residence in sediments of low hydraulic conductivity are diffusion (Desaulnier et al., 1981; Hendry and Wassenaar, 2000), or, in the case of aquitards influenced by regional groundwater flow systems such as the aquitard in Mexico City, upward advective flux against downward diffusion (Ortega et al., 1997). These controls on solute transport in the lacustrine aquitard will be the focus of further research. Insight the geochemical evolution of groundwater, a 1:1 molar ratio Cl–Na line is typical of present day closed saline lakes that undergone evaporation (Eugster, 1970; Jones et al., 1977; Eugster and Hardie, 1978; Gueddari et al., 1983; Garrels and Mackenzie, 1967), indicating that the groundwater in the Viesca aquitard was mainly influenced by evaporation of groundwater from the carbonate aquifer that dis- charged as springs in the past. Deviation from the 1:1 molar ratio should be related to other geochemical process such as ion exchange and long-term solute transport. Sulfate concentrations in groundwater at the Viesca aquitard range between 41,500 and 78,000 mg L21. These values contrast with that reported in glacial tills in North America where mean SO4 22 concentration in groundwater is 1177 mg L21 in non-weathered till, and 2646 mg L21 in weathered till (Hendry et al., A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–44 41 1986). According to the authors, the origin of sulfate in the weathered till is associated to the oxidation of organic sulfur which was incorporated into the till from the bedrock during glaciation and by lateral variability of unleached sulfate settings adjacent to leached bicarbonate settings (Keller et al., 1991). For the Viesca area the sources of sulfate in the carbonate aquifer are the dissolution of gypsum present in the Mesozoic sedimentary sequence and the oxidation of arsenopyrite present in the Mesozoic shales that additionally explains the high arsenic content in the groundwater. Insight into the chemical evolution of the pore water is provided by studies of closed saline lakes elsewhere that have been reported to show a wide range of chemical compositions (Eugster, 1970; Jones et al., 1977; Gueddari et al., 1983; Garrels and Mackenzie, 1967). The composition of the final brine is determined by that of the diluted water from which the brine was derived in present lake environments and is inherited from the earliest stages of evolution, such as the mineral weathering reactions occurring in the watershed (Eugster and Hardie, 1978). Under these conditions, the most probable chemical path for explaining the final alkali– magnesium–sulfate–chloride composition of the pore water in the aquitard in the Viesca Plain is derived from the Garrels and Mackenzie (1967) and Hardie and Eugster (1970) models for the chem- istry of waters undergoing evaporation. Starting with Na–Ca–SO4–Cl water as the dominant inflow composition, at the early stages of evaporation this causes the precipitation of calcite, which continues to precipitate until most of the calcium is removed from the solution. Water collected in the carbonate aquifer was saturated with calcite or near saturation as shown in Table 2. At this point, the ratio of carbonate alkalinity and the concentration of Ca determine a chemical divide, with 2mCa 2þ . alkalinity that, according to the Hardie and Eugster model, would proceed to a natural sulfate or chloride brine. When the concentration of sulfate is greater than that of the remaining calcium, the resulting brine will have chloride and sulfate as major anions, and sodium and magnesium as major cations (Hardie and Eugster, 1970). Concentration of sulfate is always greater than that of calcium, and therefore, reproduces the final composition of the groundwater in the Viesca aquitard. The origin of high concentration of Naþ in the aquitard should be related to the original concen- tration of sodium in groundwater from springs that progressively concentrated during evaporation or by cation exchange on clays. However, both hypothesis and other scientific questions will be assessed after detailed instrumentation of the aquitard in further stages of research. 6. Summary of conclusions Groundwater flow in the Viesca Basin has been studied using a finite cross-sectional flow modeling of hydrogeological conditions from field observations and historical accounts. A classical gravity flow system is produced by the model, which clearly demonstrates the main features of the flow system and proves to be completely compatible with the results of isotopic and hydrogeochemical investigations. The isotopic composition of the carbonate aquifer water fits closely with the global meteoric water line, indicating that this water was not affected by evaporation during its infiltration in the recharge areas. The aquifer water isotopic compositions within the Viesca Basin are between 28‰ and 210‰ for d18O and 260‰ and 265‰ for d2H. These narrow ranges also suggest that recharge areas for the regional Cretaceous aquifer had a similar general altitude. Groundwater from the aquitard plots along an evaporation line d2H ¼ 3 d18O 2 35 indicating that pore water in the aquitard resulted from the evapor- ation of groundwater fed towards springs located in the margins of the Mesozoic marine rocks and the aquitard, which is consistent with both the manifes- tations of groundwater in the past and the flow analysis. Salt concentrations in groundwater in the Viesca aquitard are between three and four orders of magnitude higher than in the diluted fresh water from springs and groundwater from the carbonate aquifer. The concentrations of salts in the pore water of the aquitard follow evaporation concepts developed for other closed or semi-closed basins in the world. The chemical trends are also consistent with the enrichment of oxygen-18 and deuterium results. A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–4442 Variation in redox related parameters such are nitrate, ammonium, manganese, total iron ad TOC indicate that aerobic conditions prevail in the carbonate aquifer; whereas, anaerobic conditions dominate in the aquitard. Redox sensitive trace elements, such as As, Se, Sb, Sn, total Cr, are very low in the carbonate aquifer and high in the groundwater from the aquitard, indicating the dominant aerobic and anaerobic conditions, respectively. Boron and Pb behave conservatively in both the carbonate aquifer and in the aquitard, increasing their concentrations as Cl2 concentrations increase too. The isotope and chemical patterns in groundwater from the aquitard are mainly a reflection of evapor- ation processes that affected the paleo-lake water fed by groundwater from the carbonate aquifer. Persist- ence of this paleo-lake saline water in the aquitard indicates that advective flow was not sufficient large to displace this groundwater. Acknowledgements Financial support was provided by Mexican taxpayers and CONACYT Grant 3454P-T. The author wishes to thank the company ‘Sulfatos de Viesca’ to permit the access at the well field areas and its field personnel for assistance. The comments and sugges- tions of three anonymous reviewers have contributed significantly to the improvement of the manuscript. References Cebria´n, M.E., Albores, A., Aguilar, M., Blakely, E., 1983. Chronic arsenic poisoning in the north of Mexico. Hum. Toxicol. (2), 121–133. Cha´vez, A., Bogrand, R., Ramirez, J., Perez, H.C., Herna´ndez, M., Rojas, A., 1964. I. Estudios en una comunidad con arsenicismo cro´nico ende´mico. Salud Pu´blica, Me´xico. Epoca V 6 (3), 421–433. Comisio´n Nacional del Agua (CNA), 1993. Actualizacio´n del estudio geohidrolo´gico de la Regio´n Lagunera, Coah-Dgo (unpublished). Craig, H., 1961. Isotope variations in meteoric waters. Science 133, 1702–1703. Desaulniers, D.E., Cherry, J.A., Fritz, et al., 1981. Origin, age and movement of pore water in argillaceous Quaternary deposits at four sites in southwestern Ontario. J. Hydrol. 50, 231–257. Eugster, H.P., 1970. Chemistry and origin of the brines of Lake Magadi, Kenya. Mineral. Soc. Amer. Special Paper No. 3, 215–235. Eugster, H.P., Hardie, L.A., 1978. Saline lakes. In: Lerman, A., (Ed.), Lakes-Chemistry, Geology, Physics, Springer, New York, pp. 237–293. Frind, E.O., Matanga, G.B., 1985. The dual formulation of flow for contaminant transport modelling.: 1 Theory. Water Resour. Res. 21 (2), 159–169. Frind, E.O., Matanga, G.B., Cherry, J.A., 1985. The dual formulation of flow for contaminant transport modelling.: 2. The Borden aquifer. Water Resour. Res. 21 (2), 170–182. Garrels, R.M., Mackenzie, F.T., 1967. Origin of the chemical compositions of some springs and lakes. In Equilibrium Concepts in Natural Water Systems. Am. Chem. Soc., Adv. Chem. (67), 222–242. Gueddari, M., Monnin, C., Perret, D., Fritz, B., Tardy, Y., 1983. Geochemistry of brines of the Chott El Jerid in southern Tunesia, Application of Pitzer’s equations. Chem. Geol. 39, 165–178. Hardie, L.A., Eugster, H.P., 1970. The evolution of closed basin brines. Mineral Soc. Am. Special Publication 3, 273–290. Hendry, M.J., Wassenaar, L.I., 2000. Controls on the distribution of major ions in groundwaters of a thick surficial aquitard. Water Resour. Res. 36 (2), 503–513. Hendry, M.J., Cherry, J.A., Wallick, E.I., 1986. Origin and distribution of sulfate in a fractured till in Southern Alberta. Canada. Water Resour. Res. 22 (1), 45–61. Jones, B.F., Eugster, H.P., Rettig, S.L., 1977. Hydrochemistry of Lake Magadi Basin, Kenya. Geochim. Cosmochim. Acta 41, 53–72. Keller, K.C., van der Kamp, G., Cherry, J.A., 1991. Hydrogeo- chemistry of a clayey aquitard till. I. Spatial variability. Water Resour. Res. 27 (10), 2543–2554. Kharaka, Y.K., Carothers, W.W., 1986. Oxygen and hydrogen isotope geochemistry of deep basin brines. In: Fritz, J.C.F., (Ed.), Handbook of Environmental Isotope Geochemistry, 2. The Terrestrial Environment B, Elsevier, Amsterdam, pp. 305–353. Mayer, P.F., 1967. Resumen de la geologı´a de la hoja Viesca 13R- 1(5). Instituto de Geologı´a, Univ. Nac. Aut. Mex. Mc Laren, R.G., 1988. CROSFLO, a 2-D steady state flow in cross- section. Waterloo, Cent. Groundwater Res. Univ. of Waterloo (unpublished). Ortega-Guerrero, M.A., Farvolden, R.N., 1989. Computer ana´lisis of regional groundwater flow and boundary conditions in the Basin of Me´xico. J. Hydrol. 110, 271–294. Ortega, G.A., Cherry, J.A., Aravena, R., 1997. Origin of pore water and salinity in the lacustrine Aquitard overlying the regional aquifer of Mexico City. J. Hydrol. 197, 47–69. Parhkurst, D.L., Thorstenson, D.C., Plummer, L.N., 1980. PHREEQE, a computer program for geochemical calculations. US Geological Survey Water Resources Investigations Report, 80-96, 210 p. A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–44 43 Plummer, L.N., Parkhurst, D.L., Fleming, G.W., Dunkle, S.A., 1988. A computer program incorporating Pitzer’s equations for calculation of geochemical reactions in brines. US Geological Survey Water Resources Investigations Report, 88-4153, 310 p. Secretarı´a de Agricultura y Recursos Hidra´ulicos (SARH), 1980. Nota informativa acerca de la actualizacio´n del estudio geohidrolo´gico de la Comarca Lagunera, Coah., Dgo. Sulfatos de Viesca, 1971. Drilling surveys in the Viesca brine fields. Internal report. To´th, J., 1966. Mapping and interpretation of field phenomena for groundwater reconnaissance in a praire environment, Alberta Canada. Int. Assoc. Sci. Hydrol. Bull. 16 (2), 20–68. To´th, J., 1999. Groundwater as a geologic agent: an overview of the causes, processes and manifestations. Hydrogeol. J. 7 (1), 1–14. Van der Kamp, G., 2001. Methods for determining the in situ hydraulic conductivity of shallow aquitards—an overview. Hydrogeol. J. 9, 5–16. A. Ortega-Guerrero / Journal of Hydrology 284 (2003) 26–4444 Origin and geochemical evolution of groundwater in a closed-basin clayey aquitard, Northern Mexico Introduction The hydrogeology of the study area Methods and materials Groundwater flow modeling Groundwater sampling Isotope and chemical analysis Geochemical modeling Results and discussion Groundwater flow modeling Environmental isotopes Geochemistry Discussion Summary of conclusions Acknowledgements References