Natural tracers in recent groundwaters from different ... · Esta georeferencia es particularmente...
Transcript of Natural tracers in recent groundwaters from different ... · Esta georeferencia es particularmente...
Natural tracers in recent groundwaters from different Alpine aquifers
Sybille Kilchmann · H. Niklaus Waber ·Aurele Parriaux · Michael Bensimon
Abstract Groundwater with underground residence timesbetween days and a few years have been investigated overmore than 20 years from 487 remote sites located indifferent aquifer types in the Alpine belt. Analysis of thedata reveals that groundwaters evolved in crystalline,evaporite, carbonate, molasse, and flysch aquifers can beclearly distinguished based on their major and trace ele-ment composition and degree of mineralisation. A furthersubdivision can be made even within one aquifer typebased on the trace element compositions, which arecharacteristic for the lithologic environment. Major andtrace element concentrations can be quantitatively de-scribed by interaction of the groundwater with the aqui-fer-specific mineralogy along the flow path. Because allinvestigated sites show minimal anthropogenic influ-ences, the observed concentration ranges represent thenatural background concentrations and can thus serve as a“geo-reference” for recent groundwaters from these five
aquifer types. This “geo-reference” is particularly usefulfor the identification of groundwater contamination. Itfurther shows that drinking water standards can be grosslyexceeded for critical elements by purely natural process-es.
Resumen Durnate m�s de 20 a�os se ha investigadoaguas subterr�neas con una residencia subterr�nea conuna duraci�n de d�as a varios a�os en 487 puntos remotoslocalizados en diferentes tipos de acu�feros en la cadenaalpina. El an�lisis de los datos revela que las aguas sub-terr�neas que han evolucionado dentro de acu�feros cris-talinos, evapor�ticos, carbonatos, flysch, y molasse sepueden distignuir claramente en base a la composici�n desus elementos mayores y marcadores y al grado de mi-nerlizaci�n. Asimismo es posible hacer una subdivisi�nm�s espec�fica incluso dentro de un tipo de acu�fero enbase a las composiciones de los elementos marcadores loscuales soncaracter�sticos del ambiente litol�gico. Lasconcentraciones de los elementos marcadores se puedendescribir cuantitativamente por la interacci�n de las aguassubterr�neas con la mineralog�a espec�fica del acu�fero alo largo del trayecto del flujo. Puesto qze todos los puntosinvestigados muestran m�nimas influencias antropog�ni-cas, los rangos deconcentraciones observados representanlas concentraciones delescenario natural y, por tanto,pueden servir como georeferencia para aguas subterr�neasrecientes que forman parte de estos cinco tipos de acu�-feros. Esta georeferencia es particularmente ffltil para laidentificcaci�n de contaminaci�n de aguas subterr�neas.Asimismo muestra que los est�ndares de agua potablepueden mostrar excesos en elementos cr�ticos por pro-cesos puramente naturales.
R�sum� L’eau souterraine ayant r�sid� sous la surface dusol entre quelques jours et quelques ann�es a �t� �tudi�esur une p�riode de plus de 20 ans � partir de 487 sites�loign�s situ�s dans diff�rents types d’aquif�res de laceinture alpine. L’analyse des donn�es r�v�le que l’eauqui a �volu�e dans des aquif�res cristallins, �vaporitiques,carbonat�s, molassiques et compos�s de flysch, peut Þtrefacilement diff�renci�e sur la base de sa composition en�l�ments majeurs et traces, ainsi que par son degr� demin�ralisation. Une subdivision suppl�mentaire peut Þtreapport�e � l’int�rieur mÞme d’un type d’aquif�re en sebasant sur la composition des �l�ments traces, lesquels
Received: 5 November 2002 / Accepted: 22 May 2004Published online: 25 September 2004
Springer-Verlag 2004
S. Kilchmann ())Laboratory of Engineering and Environmental Geology,Federal Institute of Technology,Lausanne, Switzerlande-mail: [email protected].: +41-31-3247691
H. N. WaberRock-Water Interaction, Institute of Geological Sciences,University of Bern,Baltzerstrasse 1, 3012 Bern, Switzerland
A. ParriauxLaboratory of Engineering and Environmental Geology,Federal Institute of Technology,1015 Lausanne, Switzerland
M. BensimonLaboratory of Engineering and Environmental Geology,Federal Institute of Technology,1015 Lausanne, Switzerland
Present address:S. Kilchmann, Water Protection and Fisheries Division,Swiss Agency for the Environment, Forests and Landscape,3003 Bern, Switzerland
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
sont caract�ristiques de l’environnement lithologique. Laconcentration en �l�ments majeurs et traces peut Þtreexpliqu�e quantitativement par l’interaction de l’eausouterraine avec la composition min�ralogique sp�cifiquede l’aquif�re le long des lignes d’�coulement. Puisque lamajorit� des sites �tudi�s ne montrent que tr�s peu d’in-fluences anthropog�niques, le registre de concentrationsobserv�es repr�sente la concentration de fond naturelle etpeut ainsi servir comme g�o-r�f�rence pour les eauxsouterraines r�centes dans ces cinq types d’aquif�res. Cesg�o-r�f�rences sont particuli�rement utiles pour l’identi-fication de la contamination des eaux souterraines. Parailleurs, cela d�montre que les standards d’eau potablepeuvent Þtre exc�d�s pour certains �l�ments en raison deprocessus purement naturels.
Keywords Natural groundwater composition · Naturaltracer elements · Rock-water interaction · Geochemicalmodelling · Groundwater observation networks
Introduction
The large increase in drinking water consumption in thepast few decades demands more rigorous groundwaterquality controls and a better understanding of naturallyoccurring groundwater mineralisation processes with re-spect to trace elements. In the past, quality control ofshallow drinking water supplies focused mainly on thecomposition of major elements and microbiological as-pects. Studies on trace element compositions of such re-sources were restricted to anthropogenic point source(e.g. industrial sites) and non-point source pollution (e.g.agriculture) investigations (Moore and Ramamoorthy1984; Salomons and Frstner 1984; Adriano 1986; Nriagu1991; Markert and Friese 2000). Systematic research onnatural background concentrations of trace elements inrecent groundwaters was limited until the increasedawareness of groundwater quality led to a more system-atic approach, and to the development of national ground-water-monitoring networks in several countries (e.g. TheNetherlands: Frapporti et al. 1996; Switzerland: BUWAL1998; Greber et al. 2002). Knowledge of natural contri-butions to trace element contents and fluxes in recentgroundwaters is necessary for groundwater quality as-sessment: For example in assessing the potential hazardsof naturally mobilised arsenic in groundwaters of theBengal delta plain (Sahu et al. 2001) or in southernSwitzerland (Pfeifer et al. 1995, 2000). It also serves as abase for drinking water regulation issues for geologicallydifferent environments and, if required, for groundwatertreatment strategies.
The projects AQUITYP and AQUISOL conductedat the Federal Institute of Technology in Lausanne,Switzerland (EPFL, Laboratory of Engineering and En-vironmental Geology), investigated trace element behav-iour in recent groundwaters and soils in different litho-logic environments in the Alpine belt. The AQUITYPproject (Parriaux et al. 1990a) focused on the chemistry of
recent groundwaters from the major aquifer types in theAlpine belt by establishing a groundwater-monitoringnetwork with over 500 springs and wells sampled over thepast 20 years. Detailed studies concentrated on recentgroundwaters in crystalline rocks (Dubois 1993), car-bonate sediments (Dematteis 1995), evaporite rocks(Mandia 1993), as well as molasse and flysch sediments(Hesske 1995; Hesske et al. 1997; Basabe 1993). Severalspecific trace elements have been identified in aquifersbased on statistical evaluation of the groundwater andaquifer lithology data combined with experimental data.These elements are referred to as natural tracers. Traceelement cycling in soil ecosystems, developed on thedifferent aquifer-rock types, was the main focus of theAQUISOL project (Atteia 1992, 1994; Dalla Piazza1996).
The present study presents a compilation and a pro-cess-orientated interpretation of the data gathered withinthese two projects. Concentration ranges for natural ma-jor and trace elements — a so-called “geo-reference” —is deduced for recent groundwaters from the differentaquifer types, and the relevant mineralisation processesand physical-chemical conditions are characterised byapplying geochemical modelling strategies.
Selection of Representative Catchments
Catchments selected for the AQUITYP project had tofulfil certain geologic, hydrogeologic and environmentalcriteria (Parriaux et al. 1990a). An attempt was made toselect aquifers that consist of a single lithologic unitwithin a simple tectonic context, to exclude the influenceof other lithologic units. In addition, springs had to belocated in remote regions with none or only minimalagricultural activity to minimise anthropogenic impacts.Waters from very superficial flow systems were excludedby choosing only perennial springs. Preference was givento spring waters emerging directly from the rock to avoidcontamination by materials used in technical installations.Variations in the natural groundwater chemistry in a givenaquifer type were traced by analysing most sampling sitestwice, with 32 representative sites being monitored overseveral years.
Methods
All chemical analyses were performed at the Laboratoryof Engineering and Environmental Geology. During dataacquisition over 20 years the analytical methods forgroundwater chemical analysis have continuously beenupdated (Basabe 1993, Dubois 1993, Mandia 1993,Dematteis 1995, Hesske 1995). A complete synopsis ofsampling techniques, sample treatment and analyticalmethods applied in the AQUITYP project is given inKilchmann (2001). On-site measurements included dis-charge (Q), electrical conductivity (EC), water tempera-ture (T), pH, and redox potential (EH). The groundwa-
644
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
ter samples were not filtered. Cation and trace elementconcentrations were performed on acidified (pH <2)samples by ion chromatography (IC), atomic absorptionspectrometry (AAS-FES) and most recently by high res-olution inductively coupled plasma source mass spec-trometry (HR-ICP-MS). Alkalinity was determined bytitration to the methyl-orange endpoint. Dissolved silicaand boron were determined by colorimetric methods. Aspecific electrode was used for fluoride analyses, while allother anions were determined by conventional IC meth-ods.
To enable a comparison of the groundwater chemicaldata gathered and analysed over such a long time span,rigorous quality control of the entire database was un-dertaken (Kilchmann 2001). The analytical data wereevaluated for overall quality and the comparability of thedifferent analytical methods was examined. The carbon-ate system and Ca-concentrations were examined for allanalyses by geochemical modelling strategies (see be-low). Filtration tests showed that only the concentrationsof dissolved Fe and Al appear to be overestimated due tothe presence of particulate or colloidal matter in non-filtered samples, while for major and other trace elementssuch a contribution lies within the analytical uncertainty.From a total of 1,824 groundwater analyses, 1,674 anal-yses passed the quality check. Out of these, 1,475 areanalyses of recent, tritiated groundwaters with residencetimes between a few days and about 20 years and wereselected for further investigation. The selected data arestored in an electronic database (AQUITYP-DataBase,Kilchmann 2001).
The hydrochemical facies (chemical water type) wasdetermined based on the relative proportions of majoranion and cation concentrations following the classifica-tion of J�ckli (1970).
The geochemical computer code PHREEQC (Park-hurst and Appelo 1999) with the thermodynamic databaseWATEQ4F (Ball and Nordstrom 1991) was used to cal-culate the distribution of aqueous species and mineralsaturation indices. Mineral saturation states were onlycalculated for minerals with fast reaction kinetics, whereattainment of equilibrium between the mineral andgroundwater within the residence time of the groundwa-ters is realistic.
To delineate chemical mass transfers occurring be-tween the infiltration of rain water and the discharge ofgroundwater, mass balance calculations were performedfollowing the inverse modelling strategy by Plummer andBack (1980) and using the geochemical computer codeNETPATH (Plummer et al. 1994).
Geologic and Hydrogeologic Setting
The concentrations of trace elements in recent ground-waters depend on aquifer mineralogy and hydrogeology,and on chemical rock-water interactions. In the design ofthe groundwater observation networks of the AQUITYPproject, the hydrogeologic (Table 1) and the lithologic(Table 2) characteristics of the different aquifer typeshave been considered (Parriaux et al. 1990 a).
Springs in Crystalline RocksThe springs investigated in crystalline rocks are situatedin the Mont-Blanc and Aiguilles-Rouges massifs insouthwestern Switzerland, and in neighbouring Franceand Italy (Fig. 1). This Alpine area is characterised by avery steep topography. A large part of the catchmentsurfaces are situated at altitudes above 2,000 m above sealevel and covered by glaciers or bare rock. The studiedspring waters discharge at altitudes between 570 m and2,080 m above sea level.
Major rock types of the Mont-Blanc and Aiguilles-Rouges massifs include pre-Mesozoic granite and gneissseries (Table 2). Several episodes of brittle deformationresulted in extensive fracture networks, which act aspreferential pathways for groundwater flow. Dependingon residence time, contributions to groundwater miner-alisation by water from zones of lower permeability (e.g.partially sealed fractures, undisturbed rock matrix) haveto be considered. Hydrothermal mineralisation occurringas fracture linings is a widespread occurrence in tec-tonised zones. Long-term observations of discharge fromselected springs revealed strong seasonal fluctuations dueto snowmelt during spring and summer, while watertemperature and mineralisation showed only small vari-ations. Tracer tests and calculations based on tritium databoth suggest very short residence times of the ground-water, in the order of one to two years (Dubois 1993;Mar�chal 1998).
Springs in Carbonate RocksGroundwater evolving in carbonate rocks of variable di-agenetic and metamorphic overprint was investigated incatchments in the Jura Mountains (France and Switzer-land), the Swiss Alps, the Apennines (Italy), the Dinarids(Slovenia), and the Hellenids (Greece) (Fig. 1). Thepurpose of this large geographic distribution was to coverthe lithologic diversity of carbonate rocks of different ageand from various sedimentary and tectonic environments(details are given in Dematteis, 1995).
Table 1 Investigated aquifertypes in the Alpine region andtheir typical hydrogeologiccharacteristics
Rock type Lithology Aquifer type Predominant groundwater flow
Crystalline rocks Gneiss and granite Crystalline Fracture flowRocks mostly composedof soluble minerals
Carbonate rocks Carbonate karst Conduit and fracture flow (karst)Evaporite rocks Evaporite karst
Clastic sediments Molasse rocks Molasse Inter-granular and fracture flowFlysch rocks Flysch Fracture flow
645
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
The investigated springs discharge at altitudes between1 and 1,550 m above sea level and their locations varyfrom near the sea shore to continental. The catchmentsconsist of various carbonate rocks of Devonian to Eoceneage (Table 2). The Triassic carbonate sequences of theAlps and the Apennine carbonate rocks can be intimatelyassociated with evaporite rocks. Carbonate karst aquifersare characterised by their conduit porosity (e.g. Ford andWilliams 1989; Ford 1998). The overall permeability ofcarbonate rock aquifers is usually very high and dependson the extent of karst development. By far the largestproportion of groundwater flow occurs through conduitsand fissures, while only a small proportion is transmit-ted through the matrix porosity of the carbonate rock(Atkinson 1975). The mean residence time of karstgroundwaters is consequently short, ranging from a fewhours to some months (Dematteis 1995 and referencestherein). The mineralisation of karst waters typicallyvaries strongly in response to seasonal changes in re-charge.
Springs in Evaporite RocksCatchments in Triassic evaporite rock series were studiedin the Rhone Valley in southwestern Switzerland (Fig. 1),in various tectonic settings (Pr�alpes, Helvetic and Ul-trahelvetic units, and Penninic nappes; see Mandia 1993for details). This region is characterised by a pronouncedalpine topography and the investigated groundwatersdischarge at altitudes between 375 m and 2,200 m abovesea level.
The evaporite sequence (Table 2) is usually stronglydeformed and occurs as discontinuous shreds and lensesof variable thickness. Mineralisations containing bariteand sulphides are common. Calcium sulphate rocks ex-hibit a characteristic hydrogeology. While the high solu-bility of gypsum leads to the development of karst con-duits along near-surface fractures, the hydration of an-
Table 2 Overview of the mineralogy in the different aquifer types
Aquifertype
Crystalline rocks Carbonate rocks Evaporite rocks Molasse Flysch
Lithology Granites Poly-metamorphicortho- and para-gneisses and mica-schists, containinglenses of marble,calc-silicate rocksand amphibolites
Pure, marly anddolomitic lime-stones and dolo-stones (in partsassociated withevaporite rocks)
Gypsum and an-hydrite rocks anddolomitic lime-stones associatedwith shale andsandstone inter-layers
Conglomerates andsandstones of vari-able composition,intercalated withshale layers, and lo-cally with limestoneand coal seams
Turbidite sediments(conglomerates,breccias and sand-stones with inter-calated shale andlimestone)
Age Pre-Mesozoic Devonian to Ter-tiary
Triassic Tertiary Mid-Cretaceous toEarly Tertiary
Mainminerals
Quartz,K-feldspar,saussuriticplagioclase(mainlyoligoclase)
Quartz, plagioclase(partly saussuritic),K-feldspar, with mi-nor amounts of gar-net, white mica, andpartly chloritised bi-otite and hornblende
Calcite, dolomite;detrital silicateminerals includingquartz, feldspars,mica, and clayminerals 1
Gypsum or anhy-drite, dolomiteand calcite, withdetrital quartz,feldspars, mica,and clay minerals(illite, chlorite,smectite, corren-site)
Detrital quartz, feld-spars, micas, andclay minerals (illite,chlorite, smectite,and minor kaolin-ite), calcite, dolo-mite (cement); glau-conite; gypsum andcelestite 1
Quartz, feldspars,calcite, dolomite,clay minerals (illiteand Fe-chlorite),micas
Accessoryminerals
Chloritisedbiotite,hornblende,epidote andsericite
Epidote, diopside;carbonate mineralsin marble bands andcalc-silicate rocks
Sulphides (mainlypyrite), fossil or-ganic matter(bitumen, kerogen,graphite); barite,gypsum and celes-tite, molybdeniteand U-minerals1
Celestite, fluorite,apatite, barite,talc, chalcedony,and sulphides(pyrite, chalcopy-rite, sphalerite,galena)
Heavy minerals in-cluding epidote,biotite, spinels,pyroxenes, amphi-boles, apatite, etc.;barite in veins;molybdenite andU-minerals1
Barite in veins,heavy minerals in-cluding tourmaline,zircon, apatite andTiO2-minerals
Hydro-thermalminerals
Ubiquitous calcite, dolomite,quartz and clay minerals; fluorite,molybdenite and U- and W-miner-als (scheelite, wolframite) mainlyin granites; barite, pyrite and As-,Mo-, Pb-, Zn- and Cu-sulphidesmainly in gneiss units
Secondaryminerals
Calcite, Fe-hydroxides and clay minerals
1 depending on sedimentary environment
Fig. 1 Locations of sampling sites in Switzerland (top, Hydroge-ological sketch of Switzerland modified from Federal Office ofWater and Geology), a simplified tectonic map of Switzerlandshowing the major units used (middle, A.-R.: Aiguilles-Rouge andMont-Blanc massifs, SaM: Subalpine Molasse), and the samplingsites of carbonate groundwaters across Europe (bottom, modifiedfrom Dematteis, 1995)
646
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
647
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
hydrite and the formation of secondary gypsum may sealthe porosity farther downgradient where the water ap-proaches equilibrium with gypsum. Therefore, the per-meability of such evaporite rocks depends strongly on thegeologic-tectonic context, particularly on the thickness ofthe evaporite formation and on the extent of fracturing(e.g. Parriaux et al. 1990b). In the investigated area,catchments with groundwater recharge and discharge lo-cated within evaporite rocks occur only in the Ultrahel-vetic tectonic unit inbetween the different Helveticnappes. For the catchments in other tectonic units, re-charge occurs through various other rock types (e.g. car-bonates, gneiss, schist) before reaching the evaporitelayers. Monitoring of discharge, water mineralisation andtemperature of selected springs showed large seasonalfluctuations of discharge combined with opposite varia-tions in the degree of groundwater mineralisation. Tracertests were used to determine minimum residence timesbetween 22 hours and 4 months (Mandia 1993).
Springs in Molasse RocksThe investigated molasse catchments are situated in theTertiary Molasse Basin between Chamb�ry (France) andLinz (Austria), with a large majority of the sampling sitesbeing situated in Switzerland (Fig. 1). This area shows asmooth hilly topography and springs are located at alti-tudes between 340 m and 1,430 m above sea level.Catchments are covered by deep soils and rich vegetation.
The Molasse basin consists of marine and freshwatersediment series of variable thickness, resulting from theuplift and erosion of the Alps (Tr�mpy 1980). The in-vestigated molasse catchments are mainly situated inconglomerate and sandstone formations (Table 2). Therock fragments and heavy mineral assemblages of theconglomerates and sandstones vary on a regional scaleand are characteristic for the catchment area. This varia-tion in the mineralogical composition results in typicaltrace element compositions in molasse groundwaters fromdifferent areas. Molasse aquifers are generally character-ised by a double porosity. Groundwater can circulaterelatively fast along fractures and stratification joints(“fracture flow”), and slowly within intergranular pores(“matrix flow”). Fracture flow dominates in tectonicallydeformed areas, such as in the Subalpine molasse (Par-riaux 1981; Schoepfer 1989). Matrix flow can be domi-nant in weakly deformed or undeformed areas. All in-vestigated molasse groundwaters contain tritium andresidence times range from a few days to several years(Hesske 1995). The observed negative correlation be-tween groundwater discharge and water mineralisationconfirms the influence of the residence time on ground-water mineralisation.
Springs in Flysch RocksSprings in flysch rocks were studied in the Swiss Prealps(Fig. 1). The investigated springs are located between650 m to 1,950 m above sea level. The lower catchments
are covered with meadows and forest, while no soil coveris present in the higher regions.
The flysch rocks consist of clastic turbidite sedimentsdeposited in deep marine environments. These sedimentscontain detrital components of various alpine rock types,such as carbonate rocks, shale, sandstone, schist, gneiss,quartzite, amphibolite, and various igneous rocks. Con-glomerates and sandstones are partly cemented with car-bonate minerals or contain a clay matrix. The abundantintercalated shale and fine-grained siltstone beds limitregional groundwater circulation, and catchments arecommonly small, with 80% covering an area of less thanabout 1 km2. The alpine tectonic overprint resulted inintense fracturing, internal folding, and over-thrusting ofthe flysch rocks, giving rise to open fracture networks andrapid groundwater flow. In local limestone bands, micro-karst features have developed. Seasonal variations indischarge of recent flysch groundwaters are usually high,while the variations in temperature and mineralisation arelimited. Tracer tests revealed groundwater residencetimes in the order of days to months (Basabe 1993).
Hydrochemistry of Recent Groundwaters
The chemical composition of recent groundwater ischaracteristic for each aquifer type, i.e. the aquifer li-thology in which the groundwaters evolve. While thereexists a certain overlap in major element composition dueto the limited variability of kinetically fast reactingminerals, as shown in Figs. 2 and 3, Fig. 4 and Table 3illustrate that the trace element concentrations reveal cleartrends for the different aquifer types. Common features ofall sampled groundwaters are their low water tempera-tures (between 0.2º and 18.3 C), the occurrence ofabundant dissolved oxygen, and pH values between 5.9and 8.5 (Table 3).
Crystalline-Rock AquifersGroundwaters from the crystalline rocks (granite andgneiss; Fig. 1) of the Mont-Blanc and Aiguilles-RougesMassifs (Switzerland and France) are very dilute [22 to158 mg/L total dissolved solids (TDS), Table 3] andgenerally of the Ca-HCO3-SO4 type. In certain cases,Na+, Mg2+, and F- occur as additional major cations andanions, respectively. Groundwaters evolving in gneiss andgranite aquifers have different chemical compositions.Gneiss groundwaters tend to have a higher total miner-alisation and pH value than granite groundwaters due tohigher concentrations of Mg2+, Ca2+, Sr2+ and associatedhigher alkalinity in gneiss groundwaters (Table 3). Incontrast, granite groundwaters typically contain higherconcentrations of Na+, K+, and F-. In both types, a posi-tive trend between water temperature and dissolved silicaconcentration is developed (Fig. 5). Despite their low totalmineralisation, granite- and gneiss-derived groundwatershave high concentrations of characteristic trace elements.For the granite groundwaters these are Mo, U, W, and As
648
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
(Table 3, Fig. 4). Gneiss groundwaters generally containhigher Ba and As concentrations than groundwaters fromgranite.
Within the analytical uncertainty, groundwaters fromthe granite and gneiss aquifers are undersaturated withrespect to calcite, dolomite, gypsum, fluorite, celestite,strontianite, barite and all slowly reacting Al-silicate
minerals (Table 4). Only equilibrium with respect tochalcedony is approached, indicating that dissolved silicais subjected to solubility control by a pure SiO2 mineralphase. In the granite groundwaters, an evolutionary trendtowards solubility control of F� by fluorite is observedand, in a few samples, fluorite equilibrium is almostreached. The gneiss groundwaters, in contrast, tend to becloser to saturation with respect to calcite, dolomite,gypsum, celestite, strontianite and barite than the granitegroundwaters, due to their higher activities of CO3
2�
(related to the higher pH), Ca2+, Mg2+, Sr2+, and Ba2+
(Table 4). Calculated PCO2 values range from 10�1.5 bar to10�3.9 bar with a median of 10�2.6 bar. At comparabletotal inorganic carbon (TIC) contents, granite groundwa-ters tend to have higher PCO2 values than the gneissgroundwaters due to their lower pH value.
Carbonate-Rock AquifersCarbonate karst groundwaters are characterised by a totalmineralisation between 161 to 547 mg/L TDS (Table 3).Groundwaters that discharge at low altitude springs aremore highly mineralised than those at high altitudesprings. Depending on aquifer lithology, three major hy-drochemical facies can be discerned. Groundwaters frompure limestone aquifers are dominantly of the Ca-HCO3type. Groundwaters from dolomitic limestone or dolo-stone aquifers are of the Ca-Mg-HCO3 type, while thosefrom aquifers composed of carbonate series containingevaporite beds are of the Ca-Mg-HCO3-SO4 type. Thecontents of Na+ and Cl- are generally low with the ex-ception of springs close to the coast in Slovenia andGreece, where the groundwater chemistry is influencedby sea spray. The Na+/Cl� ratio remains constant over theentire concentration range and is similar to that of rainand seawater, respectively. There is a positive correlationbetween dissolved silica concentration and water tem-perature (Fig. 5). In dolomitic limestone and dolostoneaquifers, Sr shows a positive trend with Ca, while incarbonate rocks with evaporite beds Sr is positively cor-
Fig. 2 Piper diagram (Piper 1953) depicting the distribution ofmajor element compositions in groundwaters from the differentaquifer types (milliequivalents normalised to 100%)
Fig. 3 Box plots showing natural F-, Sr2+, and Si concentrationranges in groundwaters from the different aquifer types and inprecipitation (CARB=groundwaters from carbonate rocks;CRY=groundwaters from crystalline rocks, EVAP=groundwatersfrom evaporite rocks, FLY=groundwaters from flysch rocks,MOL=groundwaters from molasse rocks, PREC=precipitation wa-ter). The boxes show the inter quartile range (25th percentile to 75th
percentile) and median values (the line through the middle of thebox). The “whiskers” above and below the boxes extend from the10th percentile to the 90th percentile and the black point displays thearithmetic mean. Outliers (lower than 10th percentile or higher thanthe 90th percentile) are not plotted. D.L.=detection limit, valuesbelow are qualitative
649
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
related with SO4. Trace element concentrations in car-bonate karst groundwaters are generally very low, butsome of them appear to represent natural tracers: Iodine isoften present at elevated concentrations in carbonate karstgroundwaters, independent of the sedimentary environ-ment of the aquifer (Table 3, Fig. 4). Groundwatersevolving in deep-sea limestone are commonly enrichedwith Ba (Apennine, NW Greece, and Malm of the Pr�-alpes). Vanadium is enriched in groundwaters from theMalm limestone of the Swiss Jura Mountains and fromthe Trieste region. Mo and U are often enriched ingroundwaters from dolomite-bearing aquifers.
Calcite equilibrium is only attained in the warmer,more mineralised karst groundwaters that discharge atlow altitude springs, while the cold waters from mountainsprings are undersaturated with respect to calcite (Ta-ble 4). Calculated PCO2 values are distinctly higher thanthat of the atmosphere and range from 10�1.1 to 10�2.8 bar.Most of the carbonate karst groundwaters are undersatu-rated with respect to dolomite, and all of them are un-dersaturated with respect to gypsum, fluorite, strontianite,and celestite. Saturation with chalcedony is only reachedin some of the warmer and more mineralised karstgroundwaters.
Evaporite-Rock AquifersEvaporite groundwaters have the highest mineralisationof all investigated groundwaters (760 to 2,788 mg/L TDS,Table 3) and show only small compositional variation.They are typically of the Ca-Mg-SO4-HCO3 chemicaltype with Ca2+, Mg2+, and SO4
2� making up 80 to 90% ofthe total mineralisation. Strontium is usually present athigher concentrations than Na+, K+, F- and Cl� and showsa positive correlation with SO4
2-. Characteristic trace el-
ements in evaporite-derived groundwaters include Li, Rb,Ni, Cu, Cd, and Mn (Table 3, Fig. 4). All of these traceelements have also been detected in the leachate solutionsof gypsum and/or anhydrite-bearing evaporites, confirm-ing their natural origin (Mandia 1993). Elevated U con-tents occur in the evaporite groundwaters of the Helveticand Penninic tectonic units.
Most of the recent evaporite-derived groundwaters areclose to or in equilibrium with calcite, dolomite, gypsum,barite, and chalcedony, and they are undersaturated withrespect to fluorite (Table 4). In some groundwaters,equilibrium is attained with celestite, while strontianite isundersaturated in all groundwaters, due to the high Sr2+
and SO42�, but low CO3
2� activities. It appears that theconcentrations of dissolved Ca2+, Mg2+, SO4
2�, Sr2+,CO3
�2, and silica are limited by the mineral solubility ofcalcite, gypsum, dolomite, chalcedony, and celestite.
Molasse-Rock AquifersThe majority of recent groundwaters evolving in molasseaquifers is of the Ca-Mg-HCO3 type with strongly vari-able total mineralisations between about 48 mg/L and714 mg/L and pronounced concentrations of dissolvedsilica (Table 3, Fig. 2). Several molasse formations havespecific mineralogical compositions that result in distinctgroundwater compositions (Hesske 1995, Hesske et al.1997). Groundwaters of the Ca-(Mg)-SO4-HCO3 chemi-cal type, a total mineralisation that exceeds 1,000 mg/L,and elevated contents of Sr, Li, Mo, U, Br, and B (Ta-ble 3) are observed in the “Gypsum-bearing molasse”(Lower Freshwater Molasse). This formation occurs alongthe Jura Mountains in western Switzerland and consists ofgypsum-bearing sandstones and marls with intercalateddolomitic limestone layers. Elevated SO4
2� contents are
Fig. 4 Box plots showing natu-ral concentration ranges of se-lected trace elements ingroundwaters from the differentaquifer types and in precipita-tion. See Fig. 3 for abbrevia-tions and Table 3 for completeconcentration ranges
650
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
Tab
le3
Ave
rage
chem
ical
com
posi
tion
sof
prec
ipit
atio
nw
ater
and
rece
ntgr
ound
wat
ers
from
diff
eren
taq
uife
rty
pes
Par
am-
eter
Uni
tP
reci
pita
tion
Gro
undw
ater
sfr
omgn
eiss
aqui
fers
Gro
undw
ater
sfr
omgr
anit
eaq
uife
rsG
roun
dwat
ers
from
carb
onat
eaq
uife
rsG
roun
dwat
ers
from
evap
orit
eaq
uife
rs
min
med
max
nm
inm
edm
axn
min
med
max
nm
inm
edm
axn
min
med
max
n
Alt
itud
em
570
1953
2,08
064
1,10
01,
491
2,01
450
151
01,
550
8337
51,
020
2,22
091
Dis
char
geL
/min
0.1
123,
467
625
112
2,45
054
290
060
,000
442
712,
111
35T
C1.
44.
78.
663
0.2
8.9
18.3
463.
39.
814
.186
4.2
8.7
13.8
91pH
pH-u
nits
3.7
5.6
7.1
426.
67.
68.
517
5.9
6.5
8.0
356.
57.
27.
985
7.0
7.0
7.6
20E
Hm
V28
443
951
16
254
402
428
620
744
051
972
E.C
.at
20 C
�S/c
m21
8319
063
2279
159
5514
734
760
979
750
1,69
32,
347
91
Tot
.ha
rdne
ss F
0.98
4.1
8.8
642.
03.
16.
847
9.5
20.0
47.8
8649
.513
0.3
196.
391
TD
Sm
g/L
22.2
575
.215
7.7
6428
.769
.114
5.8
5416
0.5
345.
454
7.2
8776
0.3
1782
.22,
787.
591
Alk
alin
ity
mg/
L9.
1539
.177
.564
15.0
29.9
54.1
5578
.021
3.5
372.
187
85.0
237.
959
3.7
91F-
mg/
L<
0.2
<0.
21.
464
<0.
21.
14.
055
<0.
2<
0.2
0.4
87<
0.2
<0.
20.
983
Cl-
mg/
L<
0.2
0.4
2.9
70<
1<
1<
164
<1
<1
17.1
55<
12.
752
.987
<1
2.9
78.1
91N
O3-
mg/
L<
0.2
1.3
16.6
70<
1<
12.
564
<1
<1
5.0
55<
13.
030
.187
<1
<1
41.7
91S
O4
2-
mg/
L<
0.2
1.2
17.1
694.
6012
.146
.064
3.0
12.5
38.7
55<
18.
614
8.9
8716
5.0
1,04
3.3
1,62
0.0
91M
g2+
mg/
L<
0.2
<0.
20.
873
<0.
20.
76.
164
<0.
20.
31.
455
<0.
24.
119
.887
16.6
57.7
227.
791
Ca2
+m
g/L
<0.
20.
53.
773
3.45
14.2
33.1
647.
011
.825
.655
28.7
72.3
126.
687
114.
342
6.0
573.
691
Sr2
+m
g/L
<0.
001
0.00
10.
009
48<
0.01
0.04
0.85
63<
0.01
0.04
0.45
550.
010.
151.
7587
1.00
8.22
17.6
091
Na+
mg/
L<
0.2
<0.
21.
466
0.60
1.5
5.4
640.
53.
423
.755
<0.
21.
645
.287
1.0
4.8
30.6
91K
+m
g/L
<0.
2<
0.2
2.9
73<
0.2
0.7
3.3
64<
0.2
1.3
3.2
55<
0.2
0.8
6.5
870.
41.
617
.791
Si
tota
lm
g/L
<0.
5<
0.5
1.3
731.
242.
54.
663
1.6
3.2
9.8
46<
0.5
3.0
8.2
871.
63.
09.
485
Li
tota
l�g
/L<
5<
543
64<
5<
521
055
<1
<1
687
<5
2811
491
Rb
tota
l�g
/L<
0.2
<0.
22.
272
<0.
20.
77.
658
<0.
21.
47.
242
<0.
20.
67.
587
1.5
17.4
53.3
78B
ato
tal
�g/L
<0.
21.
944
.073
0.20
2.9
40.1
57<
0.2
0.6
10.4
41<
0.2
11.1
220.
087
1.3
9.6
38.3
78A
lto
tal
�g/L
<0.
2<
0.2
<0.
247
<0.
23.
013
7.0
87<
0.2
14.1
93.8
73V
tota
l�g
/L<
0.2
<0.
23.
073
<0.
20.
41.
487
<0.
20.
41.
170
Cr
tota
l�g
/L<
0.2
<0.
25.
071
<0.
20.
42.
887
<0.
2<
0.2
1.1
73M
nto
tal
�g/L
<0.
24.
030
.072
<0.
21.
136
.987
<0.
212
.731
.273
Fe
tota
l�g
/L<
2<
271
.046
<2
<2
1352
.987
Co
tota
l�g
/L<
0.2
<0.
20.
587
Ni
tota
l�g
/L<
0.2
<0.
22.
271
<0.
20.
54.
078
<0.
23.
135
.778
Cu
tota
l�g
/L<
0.2
1.7
14.1
71<
0.2
0.3
50.8
87<
0.2
2.9
14.4
78Z
nto
tal
�g/L
1.3
18.5
85.9
68<
0.2
0.9
86.6
87<
0.2
2.1
156.
278
Yto
tal
�g/L
<0.
20.
62.
472
Mo
tota
l�g
/L<
0.2
3.1
21.6
580.
266
.914
0.4
42<
0.2
0.3
10.7
87C
dto
tal
�g/L
<0.
2<
0.2
0.2
18<
0.2
0.2
0.9
73W
tota
l�g
/L<
0.2
0.2
32.4
40<
0.2
3.1
14.0
38P
bto
tal
�g/L
<0.
21.
044
.071
<0.
20.
46.
087
Uto
tal
�g/L
<0.
25.
150
.558
<0.
213
5.2
2,09
2.7
42<
0.2
0.3
3.1
87<
0.2
5.2
67.0
78B
rto
tal
�g/L
<1
396
87<
18
5873
Ito
tal
�g/L
<1
1838
19B
tota
l�g
/L<
13.
094
.072
<25
<25
4745
<25
2656
38<
18
4987
<25
<25
9580
As
tota
l�g
/L<
0.5
13.8
225.
258
<0.
55.
876
.442
<0.
5<
0.5
>1.
487
<0.
5<
0.5
4.3
72
651
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
Par
am-
eter
Uni
tG
roun
dwat
ers
from
mol
asse
aqui
fers
Gro
undw
ater
sfr
om“G
ypsu
m-b
eari
ngm
olas
se”
aqui
fers
Gro
undw
ater
sfr
om“G
limm
ersa
nd”
mol
asse
aqui
fers
Gro
undw
ater
sfr
omfl
ysch
aqui
fers
min
med
max
nm
inm
edm
axn
min
med
max
nm
inm
edm
axn
Alt
itud
em
340
650
1,43
096
435
473
484
444
056
461
88
915
1502
1950
53D
isch
arge
L/m
in1
221,
260
950
460
42
2312
78
111
129
6750
T C
6.5
8.7
12.9
969.
811
.416
.34
9.1
9.8
11.9
81.
25.
79.
250
pHpH
-uni
ts6.
57.
28.
194
7.4
7.5
7.8
47.
57.
78.
18
5.9
7.0
7.7
46E
Hm
V28
839
148
494
333
417
434
433
740
144
08
449
454
477
5E
.C.
at20
C�S
/cm
2242
080
696
1,00
81,
141
1,32
64
561
642
698
815
927
047
453
Tot
.ha
rdne
ss F
1.2
25.0
43.3
9655
.273
.386
.04
16.7
37.9
39.4
89.
716
.327
.953
TD
Sm
g/L
48.4
418.
371
4.4
9698
4.4
1,14
6.9
1,34
6.0
455
6.9
638.
067
4.4
815
9.7
267.
145
8.8
53A
lkal
init
ym
g/L
7.3
281.
543
2.9
9637
5.8
403.
946
3.5
435
7.7
394.
343
0.1
811
3.7
186.
931
4.2
53F
-m
g/L
<0.
2<
0.2
0.2
960.
20.
20.
24
<0.
2<
0.2
<0.
28
<0.
2<
0.2
<0.
253
Cl-
mg/
L<
13.
032
.096
28.0
34.5
50.0
42.
08.
327
.78
<1
<1
3.5
51N
O3
-m
g/L
<1
8.1
157.
996
28.0
33.5
40.0
4<
13.
532
.38
<1
2.4
8.1
53S
O4
2-
mg/
L<
112
.160
.696
255.
038
6.5
445.
04
32.9
56.3
92.6
82.
58.
236
.253
Mg2
+m
g/L
0.7
13.5
36.1
9635
.439
.543
.24
25.3
26.1
30.8
81.
33.
223
.553
Ca2
+m
g/L
6.5
82.2
117.
496
168.
523
4.3
289.
24
94.7
117.
312
3.7
824
.056
.297
.353
Sr2
+m
g/L
0.01
00.
215
0.99
696
4.03
04.
693
4.79
74
0.16
00.
265
0.68
18
0.11
10.
368
3.11
453
Na+
mg/
L0.
32.
436
.496
7.6
9.6
29.5
42.
42.
68.
58
0.3
0.8
10.2
51K
+m
g/L
0.2
0.8
4.2
962.
03.
98.
14
0.6
0.8
0.9
8<
0.2
0.4
3.0
53S
ito
tal
mg/
L2.
05.
919
.096
3.7
4.9
7.4
42.
87.
08.
68
<0.
51.
54.
653
Li
tota
l�g
/L<
12
2996
1720
294
58
138
<1
329
53R
bto
tal
�g/L
<0.
20.
53.
096
0.9
1.1
1.8
40.
60.
81.
98
0.3
0.6
5.8
53B
ato
tal
�g/L
2.4
26.8
333.
796
20.7
32.9
43.0
410
.131
.165
.38
5.1
26.5
128.
053
Al
tota
l�g
/L<
0.2
8.2
80.5
96<
0.2
3.6
6.3
41.
11.
421
2.0
8<
0.2
0.9
23.2
53V
tota
l�g
/L<
0.2
0.5
1.1
960.
40.
60.
74
0.2
0.3
1.0
8<
0.2
0.2
0.5
53C
rto
tal
�g/L
<0.
20.
86.
996
0.6
0.7
0.8
40.
20.
21.
18
<0.
20.
21.
153
Mn
tota
l�g
/L<
0.2
0.8
374.
196
<0.
20.
50.
64
<0.
25.
545
.78
<0.
20.
54.
953
Fe
tota
l�g
/L<
2<
215
5.0
96<
2<
211
.44
<2
<2
161.
88
<2
10.7
57.9
44C
oto
tal
�g/L
<0.
2<
0.2
0.8
960.
40.
40.
64
<0.
2<
0.2
0.2
8<
0.2
<0.
20.
653
Ni
tota
l�g
/L<
0.2
0.5
3.2
960.
71.
31.
74
0.3
1.0
3.9
8<
0.2
0.5
4.0
49C
uto
tal
�g/L
<0.
2<
0.2
15.2
962.
32.
949
.54
<0.
20.
32.
88
<0.
20.
53.
748
Zn
tota
l�g
/L<
0.2
0.9
155.
096
5.6
9.1
11.0
4<
0.2
0.4
4.9
8<
0.2
1.1
75.0
46Y
tota
l�g
/LM
oto
tal
�g/L
<0.
20.
31.
996
1.3
2.6
2.7
32.
63.
89.
27
Cd
tota
l�g
/L<
0.2
<0.
20.
396
<0.
2<
0.2
<0.
24
<0.
2<
0.2
<0.
28
Wto
tal
�g/L
Pb
tota
l�g
/L<
0.2
<0.
28.
796
0.4
0.5
2.1
4<
0.2
<0.
20.
58
<0.
20.
36.
353
Uto
tal
�g/L
<0.
20.
53.
396
2.4
3.0
3.6
42.
03.
19.
08
<0.
20.
20.
953
Br
tota
l�g
/L<
19
126
9628
3745
42
2237
8<
13
853
Ito
tal
�g/L
<1
28
42
315
44
1B
tota
l�g
/L<
13
9996
1226
524
23
78
27
7353
As
tota
l�g
/L<
0.5
<0.
51.
296
<0.
5<
0.5
0.5
4<
0.5
<0.
52.
58
Tab
le3
cont
inue
d
652
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
also recorded for groundwaters evolving in the “Glimm-ersand Formation” (Upper Freshwater Molasse) that oc-curs along the Jura Mountains in northeastern Switzer-land and in the Schw�bische and Fr�nkische Alb re-gion (southeastern Germany). This formation consists ofgranite detritus and is developed as well-sorted, poorly-cemented mica-rich calcareous sandstones. Here, the el-evated SO4
2- contents are related to sulphide mineraloxidation and characteristic trace elements include Mo,U, and Li (Table 3). Groundwater from the glauconite-bearing sandstones of the Upper Marine Molasse inwestern Switzerland is characterised by naturally elevatedCr concentrations (1.9 to 6.9 �g/L). Dissolved Ba inconcentrations between 50 and 334 �g/L is a naturaltracer for groundwater that evolved in the conglomeratesediments of the Lower Freshwater and Upper MarineMolasse in certain regions of western and centralSwitzerland (Gibloux fan, Rigi-Rossberg fan, and Gr�s dela Cornalle).
Most of the investigated molasse groundwaters are inequilibrium with calcite and chalcedony and are at orclose to equilibrium with dolomite and barite (Table 4),which is consistent with the relatively long average resi-dence time. Calculated PCO2 values ranging from 10�1.3 to10�3.0 bar indicate that the uptake of soil CO2 plays animportant role in the mineralisation of molasse ground-waters. Molasse groundwaters are generally undersatu-rated with respect to gypsum, celestite, strontianite andfluorite. Similar to the evaporite groundwaters, ground-waters from the “Gypsum-bearing molasse” are close tosaturation with respect to gypsum and celestite.
Flysch-Rock AquifersMost of the flysch groundwaters are of the Ca-HCO3 or Ca-(Mg)-HCO3 chemical type with total mineralisations be-tween 160 mg/L and 459 mg/L (Table 3). Flysch springwaters generally show low concentrations of dissolved Na+,K+, Cl�, SO4
2�, Sr2+, and silica. Typically, they have verylow trace element concentrations, with the exception of Ba.
As a consequence of the commonly very short resi-dence time of flysch groundwaters, only a few of themhave reached equilibrium with calcite, while the majorityof these waters are undersaturated with respect to all in-vestigated mineral phases (Table 4). Dissolved CO2 in theflysch groundwaters is mainly soil derived, and PCO2values vary between 10�0.9 and 10�2.8 bar.
Geochemical Evolution of Recent Groundwaters
The geochemical evolution of recent groundwater is de-termined by the abundance of minerals with fast reactionkinetics and by their accessibility to water. While theabundance of such minerals is the result of the geologicalhistory of the aquifer rock, their probability to dissolvedepends on the groundwater flow characteristics of theaquifer. The chemical typology combined with the totalmineralisation of the 1,475 recent groundwaters fromdifferent aquifer types indicates that the carbonate and thesulphate/sulphide systems play dominant roles in theevolution of such groundwaters. Depending on aquifermineralogy and type of groundwater flow, however,groundwater mineralisation results from different pro-cesses, which are of special importance to the minor andtrace element mineralisation. Fast reacting carbonate,sulphate and/or sulphide minerals are present in all aquiferrock types. The dissolution of calcite, and in some casesdolomite, gypsum, anhydrite and/or pyrite (and other lessabundant metal sulphides) therefore largely determine themajor element composition of these groundwaters.
Most recent groundwaters from crystalline, carbonatekarst, molasse, and flysch aquifers have molar HCO3/Caratios of 2:1, as illustrated in Fig. 6. Within the pH rangeobserved for these groundwaters, this is characteristic ofcalcite dissolution promoted by CO2 uptake, according to
CaCO3ðsÞþH2Oþ CO2 , Ca2þþ2HCO�3 ð1ÞThe CO2 in this reaction is derived from the atmo-
sphere and, where present, from the soil cover. Deviationsfrom this general trend are mainly due to additionalsources or sinks of Ca (see below). In aquifer rocks thatcontain sulphate minerals, Ca and SO4 are mainly derivedfrom the dissolution of gypsum (reaction 2) and/or an-hydrite (reaction 3) as exemplified by the observed molarCa/SO4 ratio of close to 1:1 in recent groundwaters fromevaporite aquifers (Fig. 7) :
CaSO4 � 2H2OðsÞ , Ca2þþSO2�4 þ2H2O ð2Þ
CaSO4ðsÞ , Ca2þþSO2�4 ð3Þ
In aquifer rocks that contain no sulphate but sulphideminerals, dissolved SO4 is derived in oxygenated recentgroundwaters by sulphide (mainly pyrite) oxidation, ac-cording to
FeS2þ15=4O2þ7=2H2O, FeIIIðOHÞ3ðsÞþ2SO2�4 þ4Hþ
ð4Þ
Fig. 5 Si concentrations compared to groundwater temperature inthe different aquifer types
653
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
Tab
le4
Che
mic
alco
mpo
siti
ons
and
satu
rati
onst
ates
ofse
lect
edre
cent
grou
ndw
ater
sfr
omdi
ffer
ent
aqui
fer
type
s(a
vera
ges
ofm
ulti
ple
sam
ples
from
the
sam
esi
te)
The
“EM
OE
/84”
,“S
A2”
,“F
EY
”,“6
9b”,
and
“RN
”sp
ring
sdi
scha
rge
the
leas
tm
iner
alis
ed
grou
ndw
ater
sof
the
diff
eren
taq
uife
rty
pes,
resp
ecti
vely
.T
heot
her
spri
ngs
disc
harg
eth
em
ost
high
lym
iner
alis
edgr
ound
wat
ers
ofth
eco
rres
pond
ing
aqui
fer
type
s.
Cry
stal
line
Car
bona
teka
rst
Eva
pori
tes
Mol
asse
Fly
sch
Sit
eE
MO
E84
EM
OW
72E
MO
W66
NA
NT
PR
EM
OE
107
EM
OE
94S
A2
TIL
J8G
R1
FE
YQ
UE
69b
86b
109b
148c
RN
TF
Geo
logi
cal
unit
Gra
nite
Gne
iss
Gne
iss
Gne
iss
Gra
nite
Gra
nite
Tri
assi
cT
rias
sic
US
MO
MM
OS
M,
Gli
mm
er-
sand
e
US
M,
Gyp
sum
�M
ol.
Wat
erty
peC
a�N
a�H
CO
3\
mS
O4�
F
Ca�
Na�
SO
4�
HC
O3
Ca�
Mg�
HC
O3�
SO
4
Ca�
HC
O3�
SO
4
Ca�
HC
O3�
SO
4
Ca�
Na�
HC
O3�
SO
4�
F
Ca�
Mg�
HC
O3
Ca�
HC
O3
Ca�
Mg�
HC
O3
Ca�
Mg�
HC
O3�
SO
4
Ca�
Mg�
HC
O3�
SO
4
Ca�
Mg�
SO
4�
(HC
O3)
Ca�
Mg�
HC
O3�
(SO
4)
Ca�
Mg�
HC
O3
Ca�
Mg�
HC
O3�
(SO
4)
Ca�
(Mg)
�S
O4�
HC
O3
Ca�
(Mg)
�H
CO
3
Ca�
(Mg)
�H
CO
3
TD
S(m
g/L
)41
.413
415
115
7.7
111.
710
3.6
160.
5554
7.17
443.
6949
7.73
760.
324
07.8
114.
962
3.7
674.
413
4616
9.76
458.
75
T(
C)
4.2
4.8
7.3
6.3
5.5
11.8
4.6
9.1
12.3
7.9
8.4
9.3
7.9
8.6
9.3
16.3
76.
9pH (p
H�
unit
s)6.
27.
88
7.4
6.2
6.4
77.
37.
47.
57
7.5
6.8
7.3
7.6
7.8
6.0
7.6
Na+
(�m
ol/L
)78
184
154
8610
122
3n.
d.17
461
122
285
370
5718
710
440
012
84
K+
(�m
ol/L
)14
1013
846
535
4318
1982
528
2318
207
726
Mg2
+
(�m
ol/L
)8
8325
251
3532
144
253
782
416
2089
4001
140
1440
1267
1605
122
459
Ca2
+
(�m
ol/L
)21
567
161
682
663
749
480
131
5918
9428
1428
5313
344
477
2495
3061
7216
894
2427
Sr2
+
(�m
ol/L
)0.
126.
729.
720.
153.
830.
610.
111.
466.
7319
.97
11.5
791
.17
0.46
2.85
3.42
54.7
52
14
Alk
alin
ity
asH
CO
3�
(�m
ol/L
)
356
751
1270
1165
887
645
1899
5687
5298
3426
6464
4221
1080
6973
7049
7596
1921
5149
F� (�m
ol/L
)76
511
1121
150
21
194
33
511
511
n.d.
n.d.
Cl�
(�m
ol/L
)4
n.d.
n.d.
2411
1328
408
113
4734
212
931
138
195
959
1125
NO
3�
(�m
ol/L
)6
n.d.
n.d.
8n.
d.n.
d.52
485
5039
n.d.
n.d.
113
106
n.d.
645
24n.
d.
SO
42�
(�m
ol/L
)41
470
325
412
249
290
4511
998
1550
1735
1545
972
279
643
4633
9027
6
Si
tota
l(�
mol
/L)
8314
610
596
112
167
492
9812
118
516
416
231
830
821
341
75
Ba
tota
l(�
mol
/L)
0.00
50.
207
0.18
90.
050.
069
0.00
4n.
d.0.
102
0.10
30.
577
0.14
60.
076
0.10
60.
328
0.23
20.
313
0.13
60.
304
log
P CO
2�
2.0
�3.
3�
3.2
�2.
7�
1.6
�1.
9�
2.0
�1.
9�
2.0
�2.
3�
1.5
�2.
25�
2.07
�1.
79�
2.08
�2.
25�
1.03
�2.
22S
Ica
lcit
e�
3.2
�0.
8�
0.4
�0.
9�
2.3
�2.
3�
1.1
0.2
0.1
0.1
�0.
20.
58�
1.71
0.18
0.56
1.12
�2.
030.
34
SI
dolo
mit
e�
8.0
�2.
7�
1.3
�3.
2�
6.1
�5.
8�
3.2
�0.
8�
0.2
�0.
9�
0.6
0.55
�4.
09�
0.01
0.63
1.61
�5.
08�
0.2
SI
fluo
rite
�1.
1�
3.1
�2.
5�
2.3
�1.
8�
0.3
�4.
0�
4.2
�1.
7�
2.8
�3.
2�
2.82
�3.
18�
2.08
�2.
62�
1.92
��
SI
gyps
um�
3.6
�2.
1�
2.4
�2.
1�
2.4
�2.
5�
3.1
�2.
3�
2.6
�1.
2�
1.3
�0.
05�
3.07
�2.
05�
1.63
�0.
64�
2.75
�2
SI
bari
te�
2.5
0.1
�0.
2�
0.7
�0.
7�
2.0
��
1.1
�1.
20.
70.
10.
17�
1.03
�0.
26�
0.11
0.43
�0.
87�
0.2
SI
cele
stit
e�
4.8
�2.
1�
2.1
�3.
8�
2.6
�3.
3�
4.9
�3.
6�
3.0
�1.
3�
1.6
�0.
17�
4.00
�2.
94�
2.54
�0.
72�
3.28
�2.
21
SI
stro
ntia
nite
�5.
5�
1.9
�1.
3�
3.7
�3.
6�
4.3
�4.
1�
2.2
�1.
5�
1.2
�1.
6�
0.69
�3.
80�
1.87
�1.
51�
0.15
�3.
7�
1.01
SI
chal
cedo
ny�
0.3
�0.
0�
0.2
�0.
2�
0.2
�0.
1�
1.6
�0.
3�
0.3
�0.
20.
0�
0.04
�0.
030.
260.
23�
0.02
�0.
62�
0.35
n.d.
=no
tde
tect
ed
654
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
In this case, there is no correlation between majorcations and dissolved SO4, as is shown, for example, bythe crystalline recent groundwaters (Fig. 7). Because theinvestigated groundwaters have an oxidising redox po-tential, the FeII released during pyrite oxidation is oxi-dised and precipitates as Fe-hydroxides. Other tracemetals may be released during sulphide oxidation, but arenot necessarily subjected to rigorous solubility control bya hydroxide mineral phase as in the case of iron. Conse-quently, sulphide oxidation is an important process for themobilisation of aquifer lithology-specific trace elementsin recent groundwaters.
Reactions (1) to (4) describe the most important pro-cesses for the mineralisation of recent groundwaters fromdifferent aquifer types. Further processes are aquifer type-specific. The comparison of major element compositionand total mineralisation of the investigated groundwatersshows that the overall geochemical evolution is similarfor waters from carbonate rocks (carbonate karst) andcarbonaceous clastic sediments (molasse and flysch), butdifferent for evaporite and crystalline aquifer environ-ments (Fig. 6).
Recent Groundwaters from Crystalline CatchmentsIn crystalline aquifers, recent groundwaters with a resi-dence time of up to two years circulate predominantlyalong tectonic discontinuities (fracture flow) and theminerals present along fracture and fissure planes controltheir hydrochemistry. The fracture mineral assemblage ismainly of hydrothermal origin and differs substantiallyfrom the bulk rock composition. It generally includesmore reactive minerals, such as sulphides, carbonates, andfluorite, compared to the kinetically sluggish Al-silicatesof the rock matrix.
The general evolution of the granite and gneissgroundwaters can be explained by the dissolution ofcalcite, fluorite, a pure silica phase, sulphides (pyrite),and by Na-Ca ion exchange on clay minerals as a sourceof Na+. Mass balance calculations show that mass transferalong the flow paths from infiltrating rainwater togroundwater discharge is small, in the order of a fewtenths of a millimole per litre of solution (Table 5). Thecalculations suggest that dissolved Ca2+ is mainly derivedby the dissolution of small amounts of hydrothermalcarbonate minerals and not from weathering of Ca-bear-ing Al-silicate minerals such as plagioclase, which is inagreement with other weathering studies in crystallineenvironments (White et al. 1999; Blum et al. 1998; Harriset al. 1998). In catchments with a soil cover, calcite dis-solution is promoted by soil CO2 according to reaction(1). Calcite dissolution in the crystalline-rock groundwa-ters is further promoted by the acidity produced duringpyrite oxidation, according to
CaCO3þHþ , Ca2þþHCO�3 ð5ÞA comparison of reactions (4) and (5) shows that four
moles of calcite are needed per mol pyrite dissolved tobuffer the acidity (note that the crystalline-rock ground-waters have pH values between 6 and 8). As the molarHCO3/Ca ratio is 1:1 in reaction (5), this will result in alower molar HCO3/Ca ratio compared to calcite dissolu-tion promoted by CO2. As shown in Fig. 6, this is ob-served for many of the crystalline-rock groundwaters.Calcite dissolution caused by pyrite oxidation has alsobeen proposed by Lebdioui et al. (1990) based on stableisotope signatures (34S, 18O and 13C) in groundwatersfrom the Mont-Blanc road tunnel.
A similar shift of the HCO3/Ca ratio of 2:1 towardslower values is induced by the dissolution of fluorite
Fig. 6 Molar Ca2+ concentrations compared to alkalinity ingroundwaters from the different aquifer types. Three groups can bedistinguished: 1: groundwaters from crystalline rocks, 2: ground-waters from carbonate rocks and from calcareous clastic sediments(molasse and flysch) and 3: groundwaters from evaporite rocks. Seetext for an explanation of the 1:2 ratio
Fig. 7 Molar Ca2+ concentrations compared to SO42- in ground-
waters from the different aquifer types. See text for an explanationof the 1:1 ratio
655
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
(reaction 6). This shift, however, is much smaller due tothe smaller amount of fluorite dissolution required (Ta-ble 5).
CaF2 , Ca2þþ2F� ð6ÞFluorite is a typical hydrothermal mineral in the Mont-
Blanc granite and because of its fast dissolution kinetics itis likely to be the major source of dissolved F�. Theconcentration of dissolved F� is limited only by the sol-ubility of fluorite, as illustrated in Fig. 8. Because Ca2+
activities are low in the crystalline-rock groundwaters,solubility control by fluorite only sets an upper limit tohigh F- activities corresponding to F- concentrations ofapproximately 4.8 mg/L. In comparison, saturation withfluorite in evaporite aquifers is attained at significant-ly lower F- concentrations (around 1.1 mg/L). Naturalbackground concentrations of F- in crystalline-rock (es-pecially granite) groundwaters might therefore surpassWorld Health Organisation (WHO) drinking water rec-ommendations (1.5 mg/L, WHO 1993, 1998).
Readily dissolvable Na-bearing minerals occur neitherin the granite nor in the gneiss. Therefore, the most likelysource of dissolved Na+ in the crystalline-rock aquifers isCa-Na ion exchange on clay minerals, which liberates 2moles of Na+ per mole of adsorbed Ca2+:
2Na� Xþ Ca2þ , 2NaþþCa� X ð7ÞClay minerals that can serve as exchangers (X), occur
as abundant fracture linings in the crystalline rocks. Inaddition, ion exchange reactions proceed quickly com-pared to mineral dissolution or precipitation reactions,and equilibrium is almost always established (Appelo andPostma 1996).
The distinct occurrence of natural trace elements suchas U, Mo, As, W, and Ba in crystalline-rock groundwatersmight be used as evidence for hydraulic conditionsdominated by fracture flow: These trace elements occur inthe abundant hydrothermal fracture linings in the Mont-
Tab
le5
Res
ults
ofin
vers
em
odel
calc
ulat
ions
:A
mou
nts
ofm
iner
als
wea
ther
edfr
omin
filt
rati
onto
disc
harg
ein
mm
olpe
rkg
H2O
.Pos
itiv
eva
lues
indi
cate
diss
olut
ion,
nega
tive
valu
esin
dica
tepr
ecip
itat
ion
ofth
eco
rres
pond
ing
min
eral
phas
e.S
eeT
able
4fo
rco
mpo
siti
onof
sele
cted
grou
ndw
ater
s.
Pha
ses
For
mul
aG
neis
sG
rani
teE
vapo
rite
rock
s
Sam
ple
EM
OW
/72
EM
OW
/66
NA
NT
PR
EM
OE
/84
EM
OE
/107
EM
OE
/103
EM
OE
/94
FE
YQ
UE
Wat
erty
peC
a�N
a�S
O4�
HC
O3
Ca�
Mg�
HC
O3�
SO
4
Ca�
HC
O3�
SO
4
Ca�
Na�
HC
O3�
SO
4�
FC
a�H
CO
3�
SO
4
Ca�
Na�
HC
O3�
SO
4
Ca�
Na�
HC
O3�
SO
4�
FC
a�M
g�H
CO
3�
SO
4
Ca�
Mg�
SO
4�
(HC
O3)
Pha
ses
For
mul
am
mol
/kg
H2O
mm
ol/k
gH
2O
mm
ol/k
gH
2O
mm
ol/k
gH
2O
mm
ol/k
gH
2O
mm
ol/k
gH
2O
mm
ol/k
gH
2O
mm
ol/k
gH
2O
mm
ol/k
gH
2O
CO
2ga
sC
O2
0.01
20.
598
0.42
50.
859
1.91
60.
294
0.78
34.
762
2.31
9C
alci
teC
aCO
30.
752
0.67
90.
855
0.20
80.
669
0.48
20.
522
�0.
842
�5.
878
Dol
omit
eC
aMg(
CO
3) 2
2.09
04.
011
Gyp
sum
CaS
O4. 2H
2O
1.71
415
.395
Pyr
ite
FeS
20.
230
0.15
80.
201
0.01
60.
120
0.08
90.
140
Cel
esti
teS
rSO
40.
012
0.09
1F
luor
ite
CaF
20.
003
0.00
60.
006
0.03
80.
011
0.02
40.
075
0.00
160.
0015
SiO
2S
iO2
0.13
90.
098
0.08
90.
076
0.10
50.
102
0.16
00.
175
0.15
5Io
nex
chan
geC
a2+,
Na+
0.09
00.
075
0.04
10.
037
0.04
80.
069
0.10
90.
139
0.18
2Fig. 8 Comparison of the fluorite solubility in Ca2+-poor ground-waters from crystalline rocks and in Ca2+-rich groundwaters fromevaporite rocks (values plotted are activities)
656
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
Blanc and Aiguilles-Rouges Massifs, but not (or only verysubordinately) in the rock matrix. The natural trace ele-ment concentrations observed in the groundwaters arederived from (partial) dissolution of hydrothermal frac-ture mineral phases such as molybdenite (MoS2), variousU-minerals, scheelite (CaWO4), arsenopyrite (FeAsS),and barite (BaSO4). Aqueous speciation calculations re-vealed that under the prevailing pH and redox (oxidising)conditions, U, As, Mo and W in solution will be present inanionic forms such as UVI-carbonate complexes, andAsV�-, MoV, VI-, and WVI-oxy-anions, respectively. Un-der present conditions, there is no mineral solubilitycontrol for these trace metals, and therefore dissolvedconcentrations can be limited only via sorption, e.g. onsecondary Fe-hydroxides.
Recent Groundwaters from Carbonateand Calcareous Clastic-Sediment CatchmentsGroundwater flow in conduits is rapid, sometimes eventurbulent, leading to very short residence times of karstgroundwaters. In contrast, diffuse flow through fissuresand pores in clastic sediments (molasse and flysch) ismuch slower, and results in residence times of water of upto a few years. The hydrochemical evolution of thegroundwaters from carbonate sediments and calcareousmolasse and flysch rocks is dominated by calcite disso-lution. Calcite dissolution in these soil-covered catch-ments is enhanced by the uptake of soil CO2 (reaction 1),as indicated by molar HCO3/Ca ratios of around 2:1(Fig. 6). Some of the carbonate karst, molasse, and flyschgroundwaters are still undersaturated with respect tocalcite. This can be related to the complex calcite disso-lution kinetics which depend on factors such as flow ve-locity, residence time, and temperature of the ground-water (Appelo and Postma 1996; Dreybrodt 1998).
In some carbonate karst, molasse, and flysch aquifers,dolomite constitutes an important mineral phase. Given along enough residence time, dolomite dissolution pro-moted by soil CO2 becomes important:
CaMgðCO3Þ2ðsÞþ2H2Oþ 2CO2 , Ca2þþMg2þþ4HCO�3
ð8ÞDolomite dissolution is responsible for the higher
concentrations of HCO3� and Mg2+ in the molasse
groundwaters and in some more evolved carbonate karstand flysch groundwaters, and for the observed shift in theHCO3/Ca ratio towards values greater than 2:1 (Fig. 6). Inthe special case of the “Gypsum-bearing molasse” andcertain carbonate karst aquifers containing evaporite beds,the dissolution of gypsum (reaction 2) causes a shift in theHCO3/Ca ratio in the opposite direction, towards higherCa2+ concentrations.
Carbonate karst groundwaters generally contain verylow trace element contents due to their very short resi-dence times and because calcite contains only very re-stricted amounts of trace elements. Organic matter (in-cluding fossil organic substances) and a few accessory
minerals are the most important sources for the observedtrace element contents in carbonate karst groundwaters.Observations from recent marine carbonate sediment en-vironments show that I- is normally associated with or-ganic material and is released into solution during de-composition of this material under oxidising conditions(Martin et al 1993). The observed elevated I� contents incarbonate karst groundwaters can therefore be attribut-ed to the decomposition of fossil organic matter. Ingroundwaters from deep sea limestone, the characteristi-cally elevated Ba contents can be connected to the dis-solution of dispersed micro-crystalline barite, which oc-curs in these rocks in quantities up to 10 wt% (Church1979). Uranium and molybdenum can be present at ele-vated concentrations in groundwaters from certain dolo-mite-bearing aquifers containing minor sulphide miner-alisations including U and Mo, which are dissolved by theoxygenated groundwaters. Small amounts of U and Mocan also be incorporated into dolomite and are releasedduring its dissolution (Wedepohl 1978). Groundwatersfrom carbonate karst aquifers containing evaporite layersare characterised by elevated Sr2+ and Li+ concentrationsdue to the dissolution of gypsum and celestite (Holser1979a, 1979b). Malm limestones in the western SwissJura (particularly the Portlandian Formation (Malm);Dematteis 1995) and the bituminous Cretaceous lime-stones near Trieste have elevated V concentrations as dotheir associated recent groundwaters. In these catchments,V is released from the rock during oxidation of organicmaterial and is present in the groundwater as highlymobile anionic complexes such as H2VO4
- and HVO42�.
The particular mineralogy of certain molasse forma-tions is reflected in specific natural trace element com-positions of associated groundwaters. Elevated Cr con-centrations are recorded for the Upper Marine Molassesandstones in western Switzerland. These sandstonescontain abundant ophiolite detritus from the westernPrealps with several Cr-bearing mineral phases (spinel,pyroxene, etc., Allen et al. 1985). Weathering of theseminerals releases Cr into the groundwater, where it ispresent as highly soluble and mobile chromate anions(CrVIO4
2�). Because of the oxidising and neutral to al-kaline pH conditions, reduction of CrVI to CrIII does notoccur, and therefore no solubility control by poorly sol-uble CrIII-hydroxide occurs. In the Subalpine and foldedmolasse units, barite fracture mineralisation is the sourceof elevated Ba concentrations in associated groundwaters.In the “Glimmersand-Formation”, granite detritus con-taining sulphides, U minerals and abundant mica leads toelevated Mo, U, and Li contents in the groundwater. Fi-nally, the dissolution of evaporite minerals (Li, Sr), sul-phides (Mo), and U minerals (U) lead to the special traceelement content in groundwaters from the “Gypsum-bearing molasse”.
The short residence time and consequently poorchemical evolution of recent flysch groundwaters is onereason for their low trace element contents. In addition,the flysch rocks lack particular fracture mineralisationsthat could mark the groundwater circulating in these
657
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
fractures. The only exception is locally occurring barite,which results in elevated Ba contents in some of the re-cent flysch groundwaters. Lead concentrations of up to6.3 �g/L occur in some recent carbonate karst and flyschgroundwaters. These maximum concentrations lie in thesame range as in precipitation (Table 3, Fig. 4). In addi-tion, leaching experiments with carbonate and flyschaquifer rock material did not reveal any detectable Pb inthe leachate solution, indicating that there is no readilysoluble source of Pb in these aquifer types (note that Pbhas not been analysed in crystalline and evaporite aquifergroundwaters). This suggests that in these fast circulatinggroundwaters most of the Pb is derived from the atmo-sphere.
In this group of recent groundwaters, Ba is a wide-spread trace element. The comparison of the calculatedbarite solubility to the Ba2+ and SO4
2� activities present inthe groundwaters shows that high barium concentrationscan only occur in SO4
2�-poor groundwaters, as illustratedin Fig. 9. In SO4
2�- rich groundwater, only small amountsof barite can be dissolved before saturation is attained.
Recent Groundwatersfrom Evaporite Rock CatchmentsIn spite of their short residence time, evaporite-rockgroundwaters are by far the most mineralised waters of allthe investigated groundwaters. This is explained by thehigh solubility of gypsum and anhydrite (reactions 2 and3). Calculations showed that in pure water at 20 C,gypsum is about 225 times more soluble than calcite and167 times more soluble than dolomite. Evaporite rockscommonly contain carbonate minerals and gypsum/anhydrite dissolution is accompanied by calcite anddolomite dissolution according to reactions (1) and (8).These four reactions are linked to each other in the fol-lowing way : The dissolution of gypsum/anhydrite re-leases abundant Ca2+ into the groundwater, so that equi-
librium with calcite is rapidly reached, while gypsum/anhydrite are still undersaturated. Ongoing gypsum/an-hydrite dissolution will thus induce calcite precipitation.Calcite precipitation decreases the CO3
2� activity in thegroundwater, thereby provoking further dissolution ofdolomite and a resulting increase in Mg2+ concentration,according to
1:8CaSO4ðsÞþ0:8CaMgðCO3Þ2ðsÞ , 1:6CaCO3ðsÞþCa2þ
þ0:8Mg2þþ1:8SO2�4 ð9Þ
This process is known as de-dolomitisation and was de-scribed for various dolomite-bearing evaporite aquifersand gypsum-bearing carbonate aquifers (Back and Han-shaw 1970; Wigley 1973; Back et al. 1984; Plummer etal. 1990). Mass balance calculations confirm that de-dolomitisation takes place in the present evaporite-rockgroundwaters (Table 5).
The concentrations of Na+ and K+ are generally low inthe evaporite-rock groundwaters (Table 3). Only in a fewcases the Cl- contents are significantly higher than in in-filtrating precipitation water and Na+ and K+ contentsappear to be derived from Cl-salts (e.g. halite, silvite,carnallite). In most evaporite-rock waters, however, Clcontents are only slightly higher than in infiltrating pre-cipitation and possible sources for Na and K include ionexchange with clay minerals or dissolution of sulphatesalts (e.g. glauberite, polyhalite, bloedite, langbeinite)locally present in such rocks. However, this cannot beresolved based on the groundwater composition alone andmass balance calculations show that the small masstransfer (i.e. ion exchange in Table 5) required to obtainthe observed concentrations would not alter the generalreaction pattern.
During evaporation of sea or fresh water a large varietyof trace element bearing accessory minerals form (e.g.chloride and sulphate salts, carbonates, phosphates), mostof which are highly soluble. In addition, trace elementsare incorporated as solid solutions into major mineralphases. While the concentrations of dissolved Sr2+, Ba2+,F- and silica can be associated mainly with the dissolutionof mineral phases such as celestite, barite, fluorite, andquartz, other elements present at characteristically ele-vated concentrations, including Mn, Ni, Cu, Li, Rb, andCd, can originate from multiple sources. Mn, Ni, Cd, andCu can be derived from dissolution of carbonates andoxidation of sulphides (present in argillaceous layers), Cdfrom phosphates, and Li and Rb from brine fluid inclu-sions. Desorption processes may also be involved for allof these elements. Complexation of metal2+ ions such asMn2+, Ni2+ and Cu2+ with the major anions SO4
2� and/orCO3
2� decreases their activity and thus leads to an in-creased undersaturation with the corresponding oxide orhydroxide phase. The solubility of these elements is thusenhanced by complexation. Lithium and rubidium areboth highly soluble and chemically conservative ele-ments, and there seems to be no mechanism that removessignificant amounts of Li or Rb from solution.
Fig. 9 Comparison of the calculated barite solubility in SO42--poor
groundwaters and in SO42-- rich groundwaters (values plotted are
activities)
658
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
Conclusions
Groundwaters from different aquifer types with residencetimes between days and a few years (so-called recentgroundwaters) show characteristic signatures in majorelement composition and in the degree of mineralisation.Based on these signatures, recent groundwaters fromcrystalline-rock aquifers, carbonate-rock and calcareousclastic-sediment aquifers and evaporite-rock aquifers canbe clearly distinguished. Recent groundwaters from car-bonate karst and calcareous clastic-sediment aquifers(Molasse and Flysch units), however, are all of a similarCa-(Mg)-HCO3 chemical type with similar total miner-alisation. To distinguish such groundwaters derived fromdifferent lithologic environments, trace elements prove tobe useful. It is shown that certain natural trace elementsare present in concentration ranges that are characteristicfor groundwaters from different aquifer rock formationson a local and on a regional scale. In Table 6 a summaryof these natural tracers is given.
The chemical evolution of groundwaters from fivedifferent aquifer types is quantitatively described bymodelling the interaction of the groundwater with min-erals present along the flow path. The observed con-centration ranges of major and trace elements thereforerepresent natural background concentrations and canserve as a “geo-reference” for such groundwaters fromhydrogeologically different environments. This geo-ref-erence can be particularly useful for the identification
of groundwater contamination. It is shown that WHOdrinking water standards can be grossly exceeded forcritical elements (e.g. F, As) by purely natural process-es. This clearly supports the conclusion that effectivegroundwater management and quality assessment cannotbe made without a comprehensive knowledge of thegroundwater chemistry, geology and hydrology of acatchment. In addition, with its characteristic chemicalinformation about the recharge environment of infil-trating groundwater, the geo-reference can be of poten-tial use to flow path identification and thus to watermanagement in underground works (e.g. tunnel con-struction). The dependence of the major and traceelement composition on reaction kinetics (and thusgroundwater residence time) and on mineral occurrenceallows one to clearly distinguish between systemsdominated by fracture flow, such as the aquifers in thecrystalline, and systems with combined matrix andfracture flow, such as the molasse aquifers, where thespecific trace element concentrations partly result fromthe dissolution of minerals disseminated in the matrix.The chemical composition of recent groundwaters fromdifferent aquifer types therefore also contains certaininformation about the governing groundwater flow re-gime in a catchment.
Acknowledgements The linguistic review of the manuscript byMonique Y. Hobbs (Rock-Water Interaction, Institute of Geologi-cal Sciences, University of Bern) is greatly acknowledged. Reviewcomments made by Dr. S. Dogramaci and Prof. Y. Travi helped to
Table 6 Summary of natural tracers in recent groundwaters from different aquifer types. The concentrations of these elements can serve asgeoreference (natural background concentrations) for such groundwaters
Chemical characteristics /natural tracer
Crystalline Carbonate Molasse Flysch Evaporite
TDS low medium medium medium highSO4 0 2 (Triassic) 2 (Gypsum�bearing and
Glimmersand molasse)0 1
F 1 0 0 0 0I � 1 0 � �
Li 0 0 2 (Gypsum�bearing andGlimmersand molasse)
0 1
Rb 0 0 0 0 1Sr 0 2 (Triassic) 2 (Gypsum�bearing molasse) 0 1Ba 2 (Gneiss) 2 (deep sea limestone) 2 (Conglomerates of certain
formations in USM and OMM)1 0
Mn � 0 0 0 1Ni � 0 0 0 1Cu � 0 0 0 1Cd � 0 0 � 1
As 1 0 0 � 0Cr � 0 2 (OMM in W�CH) 0 0Mo 1 2 (dolomite bearing
aquifers)2 (Gypsum�bearing andGlimmersand molasse)
� �
U 1 2 (Gypsum�bearing andGlimmersand molasse)
0 2 (Helveticand Pennine)
V � 2 (Malm CH Jura,Trieste region)
0 0 0
0: not indicative as tracer; 1: highly indicative; 2: indicative for specific rock types / formations within an aquifer type; -: no data; USM:Lower Freshwater Molasse; OMM: Upper Marine Molasse; W-CH: western Switzerland
659
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
improve the clarity of the manuscript and are highly appreciated.This study was financially supported by the Swiss National ScienceFoundation (Grant Nº20–45730.95).
References
Adriano DC (1986) Trace elements in the terrestrial environment.Springer, Heidelberg Berlin New York
Appelo CAJ, Postma D (1996) Geochemistry, Groundwater andPollution. Balkema, Rotterdam, Brookfield
Allen PA, Mange-Rajetzky M, Matter A, Homewood P (1985)Dynamic palaeogeography of the open Burdigalien seaway,Swiss Molasse Basin. Eclogae Geol Helv 78:351–381
Atkinson TC (1975) Diffuse flow and conduit flow in limestoneterrain in the Mendip Hills, Sommerset, England. In: IAHCongress, Huntville, Alabama
Atteia O (1992) R�le du sol dans le transfert des elements traces ensolution — Application � l’�tude de quelques �cosystemsd’altitude. (The role of the soil in the transport of dissolvedtrace elements—Application to different Alpine ecosystems)PhD, Federal Institute of Technology Lausanne, Switzerland
Atteia O (1994) Transport of major and trace elements in soils andaquifers of different ecosystems of Switzerland. Eclogae GeolHelv 87:409–428
Back W, Hanshaw BB (1970) Comparison of chemical hydroge-ology of the carbonate peninsula of Florida and Yucatan.Journal of Hydrology 10:330–368
Back W, Hanshaw BB, Plummer, LN, Rahn PH, Rightmire CT,Rubin M (1984) Process and rate of dedolomitisation: masstransfer and C14 dating in a regional carbonate aquifer. GeolSoc Am Bull 95:1,415–1,429
Ball JW, Nordstrom DK (1991) WATEQ4F — User’s manual withrevised thermodynamic data base and test cases for calculatingspeciation of major, trace and redox elements in natural waters.US Geological Survey Open-File Report 90–129
Basabe PP (1993) Typologie des eaux souterraines du flysch de lanappe tectonique du Niesen (Pr�alpes Suisses). [Typology ofgroundwaters derived from the flysch of the Niesen tectonicnappe (Swiss Prealps)] PhD, Federal Institute of TechnologyLausanne, Switzerland
Blum JD, Gazis CA, Jacobson AD, Chamberlain CP (1998) Car-bonate versus silicate weathering in the Raikhot watershedwithin the high Himalayan crystalline series. Geology 26:411–414
Bundesamt f�r Umwelt, Wald und Landschaft (BUWAL) (1998)Nationale Grundwasserqualit�tsbeobachtung (NAQUA). [Na-tional groundwater quality monitoring (NAQUA)] Bern,Switzerland
Church TM (1979) Marine barite. In: Burns RG (ed) MarineMinerals. Min Soc Am Reviews in Mineralogy 6:175–209
Dalla Piazza R (1996) G�ochimie des alt�rations dans trois�cosyst�mes de sol temp�r�s — Application � l’acquisition descaract�ristiques chimiques des solut�s. (Geochemistry of min-eral weathering in three temperate soil ecosystems —Charac-terisation of dissolved constituents.) PhD, Federal Institute ofTechnology Lausanne, Switzerland
Dematteis A (1995) Typologie g�ochimique des eaux des aquif�rescarbonat�s des cha�nes alpines d’Europe centrale et m�rid-ionale. (Geochemical typology of groundwaters derived fromAlpine carbonate-rock aquifers of central and southern Europe.)PhD, Federal Institute of Technology Lausanne, Switzerland
Dreybrodt W (1998) Limestone dissolution rates in karst environ-ments. Bulletin d’Hydrog�ologie 16:167–183
Dubois J-D (1993) Typologie des aquif�res du cristallin: Exempledes massifs des Aiguilles Rouges et du Mont Blanc (France,Italie et Suisse). [Typology of crystalline-rock aquifers: Ex-ample of the Aiguilles Rouges and Mont Blanc massifs (France,Italy and Switzerland).] PhD, Federal Institute of TechnologyLausanne, Switzerland
Ford DC (1998) Perspectives in karst hydrogeology and caverngenesis. Bulletin d’Hydrog�ologie 16:9–29
Ford DC, Williams P (1989) Karst Geomorphology and Hydrology,Unwin Hyman, London
Frapporti G, Vriend SP, Van Gaans PFM (1996) Trace elements inthe shallow groundwater of the Netherlands. A geochemicaland statistical interpretation of the National Monitoring Net-work data. Aquatic Geochemistry 2:51–80
Greber E, Baumann A, Cornaz S, Herold T, Kozel R, Muralt M,Zobrist J (2002) Grundwasserqualit�t der Schweiz. (Ground-water quality in Switzerland) GWA, 3:191–201(in German)
Harris N, Bickle M, Chapman H, Fairchild I, Bunbury J (1998) Thesignificance of Himalayan rivers for silicate weathering rates:evidence from the Bhote Kosi tributary. Chem Geol 144:205–220
Hesske S (1995) Typologie des eaux souterraines de la Molasseentre Chamb�ry et Linz (France, Suisse, Allemagne, Autriche).[Typology of groundwaters derived from molasse aquifers be-tween Chamb�ry and Linz (France, Switzerland, Germany,Austria)] PhD, Federal Institute of Technology Lausanne,Switzerland
Hesske S, Parriaux A, Bensimon B (1997) Geochemistry ofspringwaters in Molasse aquifers: Typical mineral trace ele-ments. Eclogae Geol Hel 90:151–171
Holser WT (1979a) Mineralogy of evaporites. In: Marine Minerals.Min Soc Am Reviews in Mineralogy 6:211–294
Holser WT (1979b) Trace elements and isotopes in evaporites. In:Marine Minerals. Min Soc Am Reviews in Mineralogy 6:295–346
J�ckli H (1970) Kriterien zur Klassifikation von Grundwasser-vorkommen. (Criteria for groundwater classification) EclogaeGeol Helv 63:389–434
Kilchmann S (2001) Typology of Recent Groundwaters from Dif-ferent Aquifer environments Based on Geogenic Tracer Ele-ments. PhD, Federal Institute of Technology Lausanne,Switzerland
Lebdioui S, Michelot J-L, Fontes J-CH (1990) Origine de lamin�ralisation sulphat�e et carbonat�e des eaux du tunnel sousle Mont-Blanc. (Origin of the sulphate and carbonate mineral-isation of groundwaters discharging in the Mont-Blanc roadtunnel.) In: A. Parriaux (ed) Proceedings of the 22nd Congressof IAH: Water Resources in Mountainous Regions 1:496–503
Mandia Y (1993) Typologie des aquif�res �vaporitiques du Triasdans le Bassin l�manique du Rh�ne (Alpes occidentales).[Typology of Triassic evaporite-rock aquifers in the SwissRh�ne Valley (western Alps)] PhD, Federal Institute of Tech-nology Lausanne, Switzerland
Mar�chal JC (1998) Les circulations d’eau dans les massifs cris-tallins alpins et leurs relations avec les ouvrages souterrains.(Groundwater flow in Alpine crystalline-rock massifs and itsimplications for underground construction.) PhD, Federal In-stitute of Technology Lausanne, Switzerland
Markert B, Friese K (eds) (2000) Trace elements — Their distri-bution and effects in the environment. Elsevier, Amsterdam.
Martin JB, Gieskes JM, Torres M, Kastner M (1993) Bromine andiodine in Peru margin sediments and pore fluids: Implicationsfor fluid origin. Geochim Cosmochim Acta 57:4,377–4,389
Moore WS, Ramamoorthy J (1984) Heavy metals in natural waters:applied monitoring and impact assessment. Springer, Heidel-berg Berlin New York
Nriagu JO (1991) Trace metals in the environment, Elsevier,Amsterdam
Parkhurst DL, Appelo CAJ (1999) User’s guide to PHREEQC(version 2) — a computer program for speciation, batch reac-tion, one-dimensional transport, and inverse geochemical cal-culations. US Geol Survey Water-Resources InvestigationsReport 99–4259
Parriaux A (1981) Contribution � l’�tude des ressources en eau dubassin de la Broye. (Contribution to the study of water re-sources in the Broye basin.) PhD, Federal Institute of Tech-nology Lausanne, Switzerland
660
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9
Parriaux A, Dubois J-D, Mandia Y, Basabe P, Bensimon M (1990a)The AQUITYP Project: Towards an aquifer typology in thealpine orogen. In: Parriaux A (ed.) Proceedings of the 22nd
Congress of IAH: Water Resources in Mountainous Regions1:254–262
Parriaux A, Mayoraz R, Mandia Y (1990b) Impact assessment ofdeep underground works on a mineral water resource in analpine evaporitic context. In: Parriaux A (ed) Proceedings ofthe 22nd Congress of IAH: Water Resources in MountainousRegions 2:1,249–1,258
Pfeifer H-R, Hansen J, Hunziker J, Rey D, Schafer M, Serneels V(1995) Arsenic in Swiss soils and waters and their relation torock composition and mining activities. In: 3rd Int Conf on theBiogeochemistry of Trace Elements, Paris
Pfeifer H-R, Derron M-H, Rey D, Schlegel C, Atteia O, DallaPiazza R, Dubois J-D, Mandia Y (2000) Natural trace elementinput to the soil-sediment-water-plant system: examples ofbackground and contaminated situations in Switzerland, East-ern France and Northern Italy. In: Markert B, Friese K (eds)Trace elements — Their distribution and effects in the envi-ronment. Elsevier, Amsterdam.
Piper AM, Garret AA (1953) Native and contaminated groundwa-ters in the Long Beach-Santa Ana area, California. US GeolSurvey Water Supply Paper 1,136
Plummer LN, Back W (1980) The mass balance approach: Appli-cation to interpreting the chemical evolution of hydrologicsystems. Amer J Sci 280:130–142
Plummer LN, Busby JF, Lee RW, Hanshaw BB (1990) Geo-chemical Modelling of the Madison Aquifer in Parts of Mon-tana, Wyoming, and South Dakota. Water Resources Research26:1,981–2,014
Plummer LN, Prestemon EC, Parkhurst DL (1994) An interactivecode (NETPATH) for modelling NET geochemical reactionsalong a flow PATH, version 2.0. US Geol Survey Water-Resources Investigations Report 94–4169
Sahu SJ, Roy S, Jana J, Nath B, Bhattacharya R, Chatterjee D, DeyDalal SS (2001) Water chemistry and sediment-water interac-tion responsible for the mobilisation of arsenic in groundwaterin Bengal delta plain. In: Cidu R (ed) Proc. of the 10th Int Sympon Water-Rock Interaction—WRI-10, A. A. Balkema, Lisse,The Netherlands, pp 1,139–1,142
Salomons W, Frstner U (1984) Metals in the Hydrocycle.Springer, Heidelberg Berlin New York
Schoepfer P (1989) S�dimentologie et stratigraphie de la Molassemarine sup�rieure entre le Gibloux et l’Aare. (Sedimentologyand stratigraphy of the Upper Marine Molasse between theGibloux and the Aare river) PhD, University of Fribourg,Switzerland
Tr�mpy R (1980) Geology of Switzerland, A Guide Book. Wepf,Co., Basel, New York
Wedepohl KH (1978) Handbook of Geochemistry. Springer, Hei-delberg Berlin New York
White AF, Bullen TD, Vivit DV, Schulz MS, Clow DW (1999) Therole of disseminated calcite in the chemical weathering ofgranitoid rocks. Geochim Cosmochim Acta 63:1,939–1,953
Word Health Organisation (WHO) (1993) Guidelines for drinking-water quality. Recommendations. (2nd ed). Geneva
Word Health Organisation (WHO) (1998) Guidelines for drinking-water quality. Recommendations. (2nd ed). Geneva
Wigley TML (1973) The incongruent dissolution of dolomite.Geochim Cosmochim Acta 37:1,397–1,402
661
Hydrogeology Journal (2004) 12:643–661 DOI 10.1007/s10040-004-0366-9