isotopes in groundwater as indicators of climate changes

Trends in Analytical Chemistry, Vol. 30, No. 8, 2011 Trends

Isotopes in groundwater asindicators of climate changesPhilippe Negrel, Emmanuelle Petelet-Giraud

Isotopes of the water molecule (d18O and d2H) are a well-used tool for investigating groundwater origin and history (i.e. tracing

the recharge conditions over time, processes occurring during infiltration of rainwater towards aquifers and those involved in the

water-rock interaction, and mixing of different waters).

This review covers several large European aquifers (Portugal, France, UK, Switzerland, Germany, Hungary, and Poland), which

were investigated in terms of their recharge conditions, and the story of the groundwater at a large scale, involving recent,

Holocene and Pleistocene components and their eventual mixing.

ª 2011 Elsevier Ltd. All rights reserved.

Keywords: Aquifer; Climate change; European waters; Groundwater; Hydrogeology; Isotope; Rainwater infiltration; Recharge; Stable isotope;

Water-rock interaction

Philippe Negrel*,

Emmanuelle Petelet-Giraud

BRGM, Metrology Monitoring

Analysis Division,

Avenue C. Guillemin,

BP 36009,

45060 Orleans Cedex 02,

France

*Corresponding author.

E-mail: [email protected]

0165-9936/$ - see front matter ª 2011

1. Introduction

For several decades, the use of isotopicmethods in groundwater investigationshas been readily accepted by hydrogeolo-gists and scientists scrutinizing ground-water resources and their evolution inaquifer systems [1,2]. Well-establishedtechniques, mainly applying stable iso-topes of the water molecule (hydrogen andoxygen) as tracers of water sources, havebeen applied in investigations, so isotopehydrology and isotope hydrogeology havebeen great challenges since that time[3–6].

It is also clear and really evident to sci-entists and end-users that groundwater isone of the endangered resources of Europe.Groundwater, the main source of fresh-water in the majority of European Union(EU) states, is under increasing threat fromanthropogenic activities (e.g., industry,intensive agriculture, and mass tourism).Because of over-exploitation, presentrecharge cannot fully compensate for theincreasing pumping, and groundwaterresources are declining in many of theimportant aquifers of Europe. Climateprojections for Europe show changes inprecipitation and temperature patternsthat are key variables controlling forma-tion of groundwater resources [6]. Sub-stantial decreases in precipitation havebeen predicted for some parts of Europe,

Elsevier Ltd. All rights reserved. doi:10.1016/j.trac.2011.06.001

while more rainfall is expected in northernEurope. The climate of Europe is diverseand characterized by large variations fromnorth, south, east and west.

In Southern Europe, global warming,will lead to a large reduction in rechargedue to the decrease in precipitation. Thismay lead to impacts on water quality.

In Central Europe, continued depressionof groundwater levels correlates withexcessive use of water resources.

In the Atlantic regions and in NorthernEurope, increase in precipitation and re-charge, and reduction of the unsaturatedzone are expected as a direct consequenceof climate change.

Thus, because aquifers may be subjectof the effects of climate change, which isexpected to decrease precipitation andrecharge rates in large parts of Europe,there is no general agreement on how tomaintain a sustainable development ofEuropean aquifers in the future [7].

The objective of this article is to illus-trate past climatic variations, rechargeover time and water sources using isotopicmethods in groundwater investigations,especially with stable isotopes of the watermolecule (d18O and d2H). We use majorcontinental aquifer systems from all overEurope to illustrate complex histories ofgroundwaters from infiltration to theaquifer through the processes that canaffect their original signature.

1279

mailto:[email protected]

http://dx.doi.org/10.1016/j.trac.2011.06.001

Trends Trends in Analytical Chemistry, Vol. 30, No. 8, 2011

2. What are climate and climate changes?

We are all living in areas of regional climate correspond-ing to the average weather in that place over more than30 years. The regional climate can be described by thetemperatures over the seasons, how windy it is, and howmuch rain or snow falls. The climate of a region dependson many factors, including the amount of sunlight, theheight above sea level, the lie of the land, and the distanceto the oceans. However, considering the entire Earth,global climate is a description of the climate as a whole,including all the regional differences in the average.

Climate variations and change, caused by externalforces, may be partly predictable, particularly on thelarger, continental and global, spatial scales. Becausehuman activities (e.g., emission of greenhouse gases orland-use change) result in external forces, it is believedthat the large-scale aspects of human-induced climatechange are also partly predictable.

The climate system, comprising the atmosphere,the hydrosphere, the cryosphere, the land surface andthe biosphere, is an interactive system, as defined in theIPCC Report [8,9]. This system is influenced by variousexternal forces, the most important of which are the Sunand the direct effect of human activities. In the climatesystem, the atmosphere is the most unstable and reactivepart of the system, and its composition has changed withthe evolution of the Earth.

The most variable component of the atmosphere iswater, and, because the transition between the variousphases (vapor, cloud droplets, and ice crystals) absorbsand releases lot of energy, water vapor is central forclimate variability and change. The hydrosphere is thecomponent comprising all liquid surfaces and subterra-nean waters, both fresh (rivers, lakes and aquifers) andsaline (oceans and seas). Freshwater run-off from theland to the oceans influences ocean composition andcirculation, but, due to the large thermal inertia of theoceans, they act as a regulator of the Earth�s climate anda source of natural climate variability, in particular onlonger time-scales. The cryosphere, including the icesheets, continental glaciers and snow fields, sea ice andpermafrost, derives its importance to the climate systemfrom its albedo, low thermal conductivity, large thermalinertia and critical role in driving deep ocean-watercirculation. Because of the large amount of water storedin ice sheets, their variations in volume are a potentialsource of variations in sea levels.

Vegetation and soils control the Sun–atmosphereexchange of energy. Part of the exchange inducesheating of the atmosphere as the land surface warms,part leads to evaporation processes inducing water toreturn back to the atmosphere. Because the evaporationof soil moisture requires energy, soil moisture has astrong influence on the surface temperature.

1280 http://www.elsevier.com/locate/trac

Many physical, chemical and biological interactionprocesses occur among the various components of theclimate system over wide ranges of space and time scales,making the system extremely complex. As an example,the marine and terrestrial biospheres have a major impacton the composition of the atmosphere through uptakeand release of greenhouse gases by the biota. Similarly,the atmosphere and the oceans are strongly coupled and,for example, exchange water vapor and heat throughevaporation. This is part of the hydrological cycle andleads to condensation, cloud formation, precipitation andrun-off, and supplies energy to weather systems.

However, climate varies by region as a result of localdifferences in these interactions [10]. Thus, some of thefactors that have an effect on climate are changes in theamount of solar energy, greenhouse gases, albedo ofsnow and ice, and volcanic eruptions [11]. While theweather can change in just a few hours, climate changesover longer timeframes. Any change, whether natural oranthropogenic, in the components of the climate systemand their interactions, or in the external forces, mayresult in climate variations. Climate has changed in thepast, is changing nowadays and will change in thefuture. The timescale of climate change may vary fromdecades up to hundreds of million years.

3. Why stable isotopes can trace climate changes

3.1. Trace the groundwater rechargeRecharge of aquifers is mainly done by direct infiltrationof rainwater or surface water or by subsurface inflow, soprimarily originates from precipitation. In that way, it isnecessary first to obtain the signature of the recharge(i.e. of the rainfall). For the hydrosphere, increasingglobal surface temperatures lead to changes in precipi-tation and atmospheric moisture [12] and impact therecharge of aquifers. By the 1950s, it was observed thatstable isotopes of the water molecule in rainwater (d18Oand d2H, reflecting the ratio of heavy and light isotopesof 18O and 16O, and 2H and 1H respectively) dependedon several climatic factors, including air temperature,amount of rain, and altitude and latitude of precipita-tions (e.g., [13]). Thus, combining this relationshipbetween isotope ratios and climate and their well-established thermo-dependance, the isotopic signaturesof the water molecule appeared to be an appropriate toolto study the past climates in various continental andmarines archives.

The spatial and temporal variability of d2H and d18O ofmeteoric water results from the isotope-fractionationeffect accompanying the processes of evaporation andcondensation. The latitude effect reflects the rainfallprocess based on the Rayleigh fractionation/condensa-tion model that includes two processes:


(1) the formation of atmospheric vapour by evapora-tion in regions with the highest surface ocean tem-perature; and,

(2) the progressive condensation of vapour duringtransport to higher latitude.

For coastal and continental stations in Europe, thislatitude effect is D18O � –0.6&/degree of latitude (GNIPdata network; [14]). The temperature is a key parameterthat controls the d18O signatures (and thus d2H).

Yurstsever [15] established the following relationbased on amount-weighted means d18O for NorthAtlantic and European stations:

d18O ¼ 0:52t� 15‰:

The temperature effect thus mainly controls the sea-sonal variations of the isotopic signal in rainwater. Thecontinental effect, resulting in a progressive 18O (and 2H)depletion in rainwater with increasing distance from theocean, also largely controls the isotopic signature of

Figure 1. Contour map of amount-weighted mean annual d18O values (&) ias of 1997 (adapted from [16]).

precipitations. Over Europe, from the Irish coast to theUral Mountains, an average depletion of 7& is observedfor d18O, but the extent to which a continental effect alsooccurs depends on the prevailing direction of themovement of air masses.

Finally, there is an altitude effect that is temperaturerelated, as the temperature drops when altitudeincreases. All these parameters controlling the isotopicsignature in rainwater lead to the general relationbetween d18O and d2H, defined as the Global MeteoricWater Line (GMWL) [13]:

d2H ¼ 8 � d18Oþ 10

The isotopic signal of the recharge of aquifer (i.e.rainwater) over Europe is summarized as a map (Fig. 1,[16]) that reflects the continental and latitudinal effectswell. More detailed maps exist for European countries,reflecting especially the local altitudinal effects {e.g.,France [17], UK [18,19], Spain [20], and Italy [21,22]}.

n precipitation derived from the GNIP database, for stations reporting

http://www.elsevier.com/locate/trac 1281


The stable isotopes of the water molecule (d2H andd18O) are generally measured using isotope ratio massspectrometry (IRMS) with a precision of 0.1& vs. stan-dard mean ocean water (SMOW) for d18O and 0.8& ford2H. Isotopic compositions are reported in the usuald-scale in & with reference to Vienna SMOW (V-SMOW),according to:

dsampleð‰Þ ¼ fðRsample=RstandardÞ � 1g � 1000

where R are the 2H/1H and 18O/16O atomic ratios.During the past decade, the ongoing development and

evolution of laser gas analyzers and laser spectroscopypresents an alternative to conventional IRMS and con-tinuous flow IRMS (CF-IRMS) analysis of water isotopesO and H. In particular, the laser-based method is con-ceptually simple [23] and may display some majoradvantages over the IRMS method (e.g., smaller samplesizes, direct measurement of isotope ratios in the watervapor, and avoiding time-consuming sample prepara-tions). However, the main disadvantage of laser spec-troscopy compared to IRMS is the lowering of analyticalflexibility because of the single gas isotopic species ofinterest (e.g., water vapor, CO2, or CH4). However, atpresent, if some studies tend to demonstrate that lasertechnology for the measurement of liquid water isotopesyields comparable or better accuracy to conventionalIRMS and CF-IRMS analysis [24], others highlightdivergences and concern about the capability of laserspectroscopy in the analysis of liquid samples other thanpure water, due to the presence of organic compounds[25].

3.2. What changes in the water molecule can be tracedby O-H isotopes in aquifers?The isotopic composition (d2H and d18O) of the watermolecule can change (isotopic fractionation) during itstravel from the atmosphere, as rainwater, to ground-water, and sometimes within the aquifer. These potentialchanges are controlled by evaporation and exchangeprocesses.

If the isotopic signatures are not affected by anyprocess from surface to groundwater, the measuredisotope ratio in the aquifer strictly reflects the origin ofthe water (location, period and process of recharge) (i.e.the conditions prevailing at the moment of the re-charge). Thus, as the isotopic signatures are highlythermo-dependent, the climate prevailing at themoment of the recharge is preserved in the groundwatersystem, as a typical isotopic signature (under a colderclimate, the d2H and d18O values are depleted in heavyisotopes and are thus more negative). If the isotopicsignatures change along the groundwater paths, thistraces the history of the water, particularly mixing,salinization and discharge processes [14]. Even if theisotopic composition of groundwater is mainly inheritedfrom atmospheric signal, there are some cases where


reactions between groundwater and the rock matrix orgases or surface–subsurface processes (e.g., evaporation)can modify the original meteoric signatures [14,4].

4. Recharge and residence-time assessment ofgroundwater

The Adour-Garonne region covers 116,000 km2 (one-fifth of French territory). It is limited by the MassifCentral and Montagne Noire to the east, the ArmoricanMassif to the north, the Pyrenees Mountains to the southand the Atlantic Ocean to the west (Fig. 2). In thisregion, the Eocene sands water body (a multi-layersystem) constitutes a series of major aquifers used fordrinking water supply (6.7 million inhabitants), agri-culture irrigation and thermo-mineral-water resource[26] and comprises sandy Tertiary sediments alternatingwith carbonate deposits. The Eocene aquifer systempresents high permeability and a thickness of severaltenths of meters to a hundred meters and has at least fiveaquifers: Paleocene, Eocene infra-molassic sands (IMS),early Eocene, middle Eocene, and late Eocene. Ground-water recharge may occur to the east by the edge of theMassif Central, to the south by the edge of the Pyreneesand by inflow from the Paleocene aquifer.

The d18O values of the groundwaters fall in the range�5.6& to �10.6& vs. SMOW, with the d2H valuesvarying between �34.3& and �72.3& vs. SMOW [2].There is no relationship between d18O and d2H values,and the salinity with a correlation coefficient R close to0.30 and 0.45 between salinity and d18O–d2H values,respectively. All the waters analyzed (Fig. 3) fall on ornear the GMWL [13].

The groundwater data present a wide range of stableisotopic composition (d2H, d18O) both between thedifferent aquifers and within a single aquifer (Fig. 3).Comparing independently each aquifer level sampled inlow and high flows of a same hydrological cycle, thefollowing is apparent.

4.1. Paleocene aquiferCollected NW-SE along the Pyrenees border, the Paleo-cene aquifer presents a large variation in the d18O andd2H signature, agreeing with that observed in the region.Such heterogeneity in the d18O and d2H signatures forthe Paleocene aquifer reflects variable recharge in spaceand time. The most depleted value corresponded to awater recharged with a colder climate than the presentone (>10,000 y); the similarity in the d18O and d2Hsignatures along the hydrological cycle confirms ahomogeneous aquifer system without any water inputwith a different signature between the two periods. Thismeans that the aquifer is mostly confined with no sig-nificant recharge.

Figure 2. General setting of the Adour Garonne region (SW France) and schematic map of the Total Dissolved Solids (TDS, mg/L) in the IMSaquifers (adapted from [2]).


4.2. Lower Eocene aquiferThe Lower Eocene aquifer was essentially collected in thenorthern part of the region (Fig. 3). The d18O and d2Hsignatures are less variable than those of the Paleoceneaquifer and fall in the middle range observed as a whole.The values are consistent along the hydrological cycleand are more depleted compared to the present rain-water input in the Massif Central, reflecting an olderrecharge that could have occurred under colder climate.This means that the aquifer is probably semi-confined.

4.3. Middle Eocene aquiferLargely sampled in the northern part of the region, theMiddle Eocene aquifer presents a large range of d18O andd2H signatures among which the most enriched values,very close to that of the rain inputs observed in Dax,reflect a modern recharge. Most of the samples showdepleted values that reflect a recharge under colderclimate. In between these two groups, a third group withintermediate values corresponds either to a mixturebetween the two previous groups or to closed pockets(e.g., small confined parts of the aquifer). The largevariation of the point EM4 between the two surveyssurely reflects hydrological conditions that differ fromone survey to the other.

4.4. Upper Eocene aquiferThe Upper Eocene aquifer was sampled in four locationsin the northern part of the region along a NW-SE profile.Although having similar signatures in the two surveys,they displayed enriched or depleted values. ES-3 isenriched, while ES-4 is depleted (recharge under colderclimate) and ES-2 displays a signature close to present-day rainwater in the Massif Central.

4.5. Infra Molassic Sand aquiferLargely collected in the southern part of the region alonga W–E profile, the Infra Molassic Sand aquifer displays alarge range of d18O and d2H signatures. SIM2 and SIM8have signatures close to those of rainwater in Dax andMassif Central. SIM2 is a shallow bore-well (58 m) and amodern recharge is compatible with the observed values.By contrast, SIM8 is a 1400-m deep bore well screenedat 1030–1040 m and the water is the most saline of theregion [Total Dissolved Solids (TDS) up to 2.5 g/L].The d18O and d2H signatures suggest a modern rechargethat may result in rapid circulation of the groundwaterin the system. As for the remaining points, they aredepleted in 18O and D compared to rainwater, reflectinga recharge under colder climate and a semi-confinedstatus for the aquifer.


Figure 3. d18O–d2H plot for the groundwaters collected in the Eocene sand aquifers (adapted from [2]). The Global Meteoric Water Line (GMWL)is defined as d2H = 8\d18O + 10 [13].


4.6. Adour-Garonne region aquifers taken as a wholeTaken as a whole, the enriched samples clearly correlatewith the present-day recharge as measured in Dax andMassif Central. The most enriched waters (EM1, EM2,ES3) originate from the north of the area, in the vicinityof the Gironde estuary and present signatures quitesimilar to that of present-day coastal precipitations(mean weighted rain in Dax).

Samples SIM2 and 8 also show an enriched signaturethat can be related to the modern recharge.

By contrast, the most depleted sample (P3) originatesfrom the Paleocene aquifer (860 m depth), and mayreflect an old recharge, as its signature is clearly lowerthan that of present-day precipitations from the MassifCentral or Pyrenees. The groundwater presents a widerange of variation along the GMWL which excludessignificant evaporation of infiltrating waters and anycontinental effect on the stable isotope composition.These variations cannot be easily correlated with thedata on spatial location, and are probably mostly due tothe period and the location of the recharge of theaquifer.

The most depleted sample of IMS (SIM4) is located inthe eastern border of the basin and originates from177 m depth. It may represent an old recharge {as theestimated age of some groundwaters is close to 16–35 kausing 14C [26]}.


5. Stable isotopes in some European large aquifers

Variations of the d18O and d2H signature in the Eoceneaquifer system from the Adour-Garonne region in SWFrance can be related to changes in the recharge overvarious climatic periods, from present day to old (e.g., upto 35 ka). Similar behavior can be revealed by severalstudies of groundwaters in aquifers, and such significantisotopic depletion of groundwaters may be due to a lowerrecharge temperature at the time of infiltration [27]. Weconsider several illustrative cases in European aquifers,as summarized in Fig. 4, for which we investigatedgroundwater as an archive of climatic changes or areflection of a more recent recharge period.

5.1. Aquitaine BasinThe Aquitaine Basin, considered by Le Gal La Salle et al.[28] and Jirakova et al. [29], occupies an area of25,810 km2 and extends essentially in the northern partof the Adour-Garonne region (Poitou-Charentes district)and is limited to the west by the Atlantic Ocean. Lowerand Middle Jurassic carbonate formations make up thedeepest aquifer of the northern part of the AquitaineBasin investigated in these studies. Reported in the d18Ovs. d2H graph (Fig. 5), the data fall in the upper range ofvalues measured for the whole Aquitaine Basin {[26]and this article, Section 4}.

Figure 4. Location map of the main European aquifers summarized in this article.


Groundwaters with a wide range of values for d18O(�4.9& to �7.4&) and d2H (�35& to �48&) plotalong the GMWL, reflecting the meteoric origin {Fig. 5[28]}. Some of the groundwaters are significantlyenriched or depleted in comparison with modern waters(d18O around �5.7& and d2H around �36&). Con-sidering the repartition of the values, enriched stableisotopes in groundwaters in the southwest part of thearea were in evidence, while groundwaters were eitherdepleted or similar to modern recharge northeastward[28]. The heterogeneity of the recharge conditions in thearea are attested by the variations observed and may bedue to climate, temperature or infiltration changes, evenif the authors do not preclude mixing with enrichedwaters (e.g., seawater) in the deeper part of the basin.However, groundwaters having a depleted value com-pared to the modern recharge plead in favor of a largepaleorecharge of the aquifer that occurred during acolder period, possibly the last glaciation or deglaciationperiod.

Considering a larger area in the Poitou-Charentesdistrict, Jirakova et al. [29] show an entire range ofstable-isotope values from �7.7& to �4.9& for d18Oand –52.3& to –29.6& for d2H, which agree with the

study of Le Gal La Salle [28] concerning the hetero-geneity of the recharge conditions, as demonstrated inthe d18O vs. d2H diagram (Fig. 5). Enlarging the areaconfirms the existence of depleted d18O and d2H valuesand suggests lower temperatures during the rechargeperiod. This type of water represents the paleorechargeunder cold climatic conditions during the late Pleisto-cene period and was clearly separated from those of theHolocene. The modeling of the water-residence time,using 14C, showed Holocene waters with enriched values(around �6& and �38.5& for d18O and d2H, respec-tively) with radiocarbon ages up to 10 ka Before Present(BP). However, the groundwaters having depleted values(around �7.4& and �48& for d18O and d2H, respec-tively) have radiocarbon ages of 20–15 ka BP that cor-respond to cold recharge.

5.2. Paris BasinMoving north of the Aquitaine Basin, the Paris Basin isthe second major sedimentary basin in France with anextent of around 600 km in diameter and more than3000 m of sediment deposits. The Dogger aquifer (200–300 m, predominantly limestone), confined between theLiassic and Upper Callovian marls, was studied by


Figure 5. d18O–d2H plot for the groundwaters from European aquifers; Global Meteoric Water Line (GMWL) as in Fig. 3: Aquitaine Basin (datafrom [26,28,29], data from the groundwaters collected in the Eocene sand aquifers are in the grey area); Chalk aquifer from the Paris Basin (datafrom [30,31]; Chalk aquifer from the London Basin (data from [32–34]. Gorleben and Laegerdof aquifers (Germany, data from [31,36]; Polandaquifer (data from [38]); Pannonian Basin (data from [37]); Portugal aquifer (data from [39]); Switzerland (data from [40] and reference therein);and, Lorraine aquifer (Eastern France, data from [35]).


Matray et al. [30]. The Cretaceous Chalk aquifer (700 mof fine-grained limestone), either confined or unconfined,is extensively exploited for drinking water and irrigationand was studied by Kloppmann et al. [31].

Groundwater samples from the recharge zone of theDogger aquifer mimic the stable-isotope composition ofthe present-day rainwaters, indicating recent meteoricorigin (Fig. 5). The rest of the groundwater data fall tothe right of the GMWL with low variations in the d2Hand d18O (ranges �40& to�30& for d2H, and �6& to


�3& for d18O). In this part of the basin, watertemperature and salinity are high but seem to have noimpact on the stable-isotope signatures. The authorsdiscussed the possible isotope exchange with the car-bonate matrix for oxygen, and with H2S for deuterium,but they argued that such processes cannot be themajor ones controlling the groundwater-isotope values.Important mixing processes are responsible of boththe stable-isotope composition and the salinity of thegroundwater in the Dogger aquifer and successive


mixings between sedimentary brines with several mete-oric waters led to the observed isotope compositions ofthe groundwater. The authors argue for percolation ofmeteoric water, dissolution of halite in the Triassicaquifer that generated a brine that mixed with a residualprimary brine then migrated via vertical faults into theDogger aquifer.

Stable isotopes in the confined groundwaters of theChalk aquifer of the Paris Basin plead in favor of arecharge during the Holocene period with a watercomponent related to Pleistocene ages in the deepestconfined part of the aquifer. The observed depletedvalues in the confined part of the aquifer can be relatedto a lower recharge temperature at the time of infiltra-tion. However, the continental effect influences thestable-isotope composition of Chalk groundwaters.The observed depletion of the stable isotopes in theunconfined groundwaters from west to east mimics thatobservable in rainfall when air masses penetrated acontinental area. Long-term changes in the inputfunction related to climatic evolution yield the lower-most d2H–d18O values in the confined aquifer, andenriched values in the unconfined aquifer originatesfrom fractionation processes due to variation in therecharge signal (e.g., the rain input).

5.3. London BasinIn the UK, the Chalk aquifers were studied by Hiscocket al. [32] in the Norfolk area (eastern England), Denniset al. [33] in the London Basin (SE England), andElliot et al. [34] in the London Basin and the adjacentBerkshire Basin (east of the London Basin). The geologycomprises Cretaceous Chalk overlain with Tertiaryclastic deposits with a thickness of more than 400 m inthe Norfolk area and up to 250 in the London Basin.

d18O–d2H values fall within the range �8& to �6.6&

and �55& to �43&, respectively, and all data fall on,or close to, the GMWL, as illustrated in Fig. 5.

Both studies argue that the evolved Chalk waterstypically show enriched isotopic signatures that might becompared to the recharge temperatures determinedthrough noble-gas investigations. This may reflect themixing of relatively young groundwater in the fissureswith older groundwater in the matrix. The d18O–d2Hvariation along the GMWL was related by all studies tothe complexity of the recharge over time and mixingprocesses in the aquifer. The conceptual model for theChalk aquifer suggests that the water evolved fromconnate Cretaceous marine water has repeatedly mixedwith fresh meteoric water since the Late Tertiary. Thepresent-day conditions reflect such mixing with paleo-water recharged during a cold period.

5.4. Lorraine BasinMoving eastward of the Paris Basin, the lower Triassic ineastern France is mainly represented by sandstones and

conglomerates and the extent of the Triassic aquiferreaches 3000 km2 for the unconfined and 25,000 km2

for the confined part. In the d2H vs. d18O graph (Fig. 5),data are scattered between two extreme values. Themost enriched values correspond to recent waters closeto the recharge zone and reflect the values of modernrainfall [35]. However, the depleted values may corre-spond to a cooler recharge regime during the Holocene.

5.5. Germany: Gorleben and Laegerdorf BasinsMoving more eastward, the salt structure of Gorleben–Rambow in Germany, crossing the Elbe River about100 km upstream of Hamburg, has been investigated indetail for decades with around 400 boreholes [36]. Inthis framework, all fluids analyzed fall on the GMWL.The shallow groundwaters show d18O values of –8& to–8.5& and d2H of –56& and –60& (Fig. 5) in thetypical range of meteoric waters, and they have beenidentified as modern on the basis of their 14C and 3Hcontents. Most saline waters are depleted with respect toshallow groundwater by up to 20& in 2H and 2& in18O. The lowest observed values are �72& (d2H) and�10.3& (d18O) for samples with salinities around 50 g/L. A group of highly saline groundwaters show stable-isotope contents in the range of shallow freshwaters. Thedepletion of most of the deep saline groundwaters can beexplained by a significant contribution of Pleistocenerecharge, probably through meltwater infiltrationduring or shortly after the last glaciation, this interpre-tation being in agreement with radiocarbon data.Intermediate stable-isotope contents are due to mixingbetween Holocene and Pleistocene components leadingto the large scatter of the reported values.

Also in Germany, Kloppmann et al.[31] presentedisotope data in chalk groundwater (Fig. 5). The Laeger-dorf Chalk outcrop (Campanian-Maastrichtian) isexploited up to a 90-m depth. The d2H and d18O valuesfall along the GMWL but with higher values than inGorleben. Kloppmann et al. [31] related the mostdepleted values in Laegerdorf to a temperature effect.

5.6. Pannonian BasinMoving further eastward, the Pannonian Basin is a largearea formed mainly during the late Tertiary and Qua-ternary periods and covering 100,000 km2 in southeastHungary. The hydrogeological system is a multilayeraquifer system with an intermediate flow regime in thePleistocene sediments that concerns local to regionalscale, and a deeper system (below 2500 m) that con-cerns the regional scale.

Groundwaters plotted on a d2H and d18O graph definetwo groups (Fig. 5 [37]), as follows.(1) One group, corresponding to the deeper part of the

aquifer (500–2500 m, T > 40�C), shows a weakrange in d2H and d18O and falls close to the GMWL.This suggests that infiltration occurred during the



same, and probably the last, cold period thatoccurred between 70–12 ka BP.

(2) The data for the other group of waters fall to theright of the GMWL and the shifts of d2H and d18Oare not related to an evaporation process but tomixing between old waters with water enriched inheavy isotopes [37]. The origin of the enrichedwaters is suspected to be from oilfield water,squeezed from the Pannonian layers underlyingthe aquifer.

5.7. Poland BasinThe Malm aquifer (limestones, sandstones and marls) insouthern Poland around Cracow is the most eastward ofthe investigated aquifers with which we illustrate thisarticle. This aquifer is intensively exploited and serves asa major strategic reserve of potable water for the 1million inhabitants of Cracow. The Malm aquifer isdriven by numerous faults, graben and horsts, and, as aconsequence of this complex geology, the flow patternand ages of water are poorly constrained. Based onseveral tools (e.g., noble gases, tritium and carbon iso-topes), Zuber et al. [38] defined a range of water ages onthe confined and unconfined part of the aquifer frommodern period to glacial waters, and glacial waterspartly mixed with older water (e.g., before 15 ka BP).

In the d2H and d18O graph illustrated in Fig. 5,groundwaters fall along the GMWL. Modern watersshow enriched d2H–d18O values, in agreement with themean annual precipitation. They are from the presentday (e.g., containing tritium) or from pre-Bomb time(e.g., free of tritium). The older the waters, the more thatthey have depleted d7H–d18O values and their status(glacial-Holocene transition period and glacial) is con-strained by the d2H–d18O values and noble-gas temper-atures. The older waters show the largest d2H–d18Odepleted values and are related to other recharge areas,or of deeper origin than the glacial waters [38].

5.8. PortugalMoving south-westward in Europe, to south Portugal,the sedimentary Sado Basin is made of Eocene (sand-stone and carbonate), Miocene (conglomerates, lime-stones and sandstones) and Pliocene (conglomeratesand sands) sediments [39]. Within the Sado Basin, thePlio-Miocene and the Eocene are the two identifiedaquifer systems. The two aquifers have similar d2H–d18Ovalues (e.g., �5.0& to �4.0& in the Eocene aquifer and�5.0& to �4.6& in the Plio-Miocene aquifer for d18O)(Fig. 5). Values are depleted along the flow path, asshown by the Plio-Miocene, which is more depleted nearthe northern limit of the basin compared to the south-ern. According to the range in 14C ages for thegroundwater and the d2H–d18O values in the Plio-Miocene and Eocene aquifers, Galego-Fernandes andCarreira [39] argued that infiltration processes of the


Eocene waters under climatic conditions differed fromthe modern ones.

5.9. A highly complex aquifer system: the SwitzerlandexampleIn the middle of the Europe, northern Switzerland com-prises three main sedimentary stratigraphic groups fromTertiary to Permo-Carboniferous, which include largeaquifers, potentially locally connected through tectonicaccidents. These sediments contain several aquifers,composed of sandstone (Tertiary-Malm group) and Malmlimestones; the upper Muschelkalk comprises limestonesand dolomites; and, the lower Triassic-Permian groupclastic sediments. The two other aquifer systems are theQuaternary cover, largely affected by anthropogenicactivities, and the crystalline basement, where watercirculations in the crystalline basement are mainlycontrolled by tectonic fractures [40]. The synthesispresents waters sampled in the beginning of the 1980s(Fig. 5). Waters from Quaternary, Tertiary and Malmaquifers are not considered, as most of the samplespresent typical values of modern recharge from thenorthern Switzerland. The Keuper aquifer level mainlycontains young groundwaters, with detectable tritiumcontents, defining relatively small ranges of stableisotopic signatures (i.e. �10.5& to �8.7&, and �74&

to �63& for d18O and d2H, respectively).The first Muschelkalk layer contains young waters

(3H > 20 TU) falling on the GMWL and reflectingsuperficial waters of the upper and middle Muschelkalk,and also mixed superficial and deeper waters of theupper Muschelkalk. Samples with low tritium contents(3H < 20 TU) present different stable-isotope signatures;some are identical to the modern samples, while othersare depleted in heavy isotopes or enriched in 18O only.The 18O-enriched samples correspond to hot springs thatcould originate from mixing of depleted deep water withenriched water in both 2H and 18O by evaporation, or byisotope exchange with the rock matrix. Depleted samplesin both 2H and 18O have different stories, and, surpris-ingly, the most depleted, which normally must reflect anold recharge, was identified as modern as or youngerthan 1 ka using dating tools. This sample, close to theRhine River, seems to reflect a recharge from the Rhinewith a signature similar to those of Alpine precipitations.Some other samples have signatures consistent with arecharge in cooler conditions, as they were estimated tobe older than 15–30 ka. Finally, some Muschelkalksamples probably reflect a recharge at higher altitude inthe adjacent Black Forest.

Most of the waters from the Buntsandstein, Permianand Crystalline basement are in the same range for d18Oand d2H as samples of the other upper aquifers. The fourmost depleted samples from tunnels in the Alps presentsignatures consistent with recharge from precipitationat high altitude, and thus do not reflect old waters


recharged under cooler climate. Enriched waters in 2Hand 18O or only in 18O include tritium-free samples andwaters with 3H > 20 TU as well as more or less miner-alized waters. Each reflects a specific story, from isotopeexchange due to water-rock interaction to mixing ofhighly evaporated seawater with meteoric waters similarto young waters currently found in the upper crystallineof northern Switzerland.

This example in Switzerland, mainly in the context oflarge sedimentary aquifers, illustrates well that signa-tures of stable isotopes of the water molecule can reflectlong, complex stories and processes, not only climatevariation through time, even when samples fall alongthe GMWL. It is therefore evidence that isotopic dataalways need to be interpreted jointly with chemical datain the context of general and local hydrogeologies.

6. Conclusions

Based on the literature and recent investigations in SWFrance, this article has highlighted the use of the isotopicmethods in groundwater investigations, applying stableisotopes of the water molecule (hydrogen and oxygen) astracers of water source, recharge over time and pastclimatic variations in various continental aquifers.Parameters controlling the isotopic signature in rain-water (e.g., continental and altitude effect) lead to thegeneral relation between d18O and d2H, defined as theGMWL, and groundwaters generally fall along this line.If no processes affect the water molecule from surface toaquifers, groundwater conditions prevailing at themoment of the recharge are preserved. Thus, the isotopesignature may reflect recent recharge with hydrogenand oxygen isotope values in the range of rainwaters orrecharge under colder climate.

Starting from the recent study of the multi-Eocenesands layer system (five aquifers: Paleocene, Eocene in-fra-molassic sands, Early Eocene, Middle Eocene, andLate Eocene) in the Adour-Garonne region (one-fifth ofFrench territory), this article explored the rechargeconditions for different aquifer systems all over Europe(Portugal, France, UK, Switzerland, Germany, Hungary,and Poland). We highlighted the recharge conditionsand the story of the groundwater at a large scale,involving recent, Holocene and Pleistocene componentsand eventually mixing between them.

The Adour-Garonne region highlighted the story of therecharge over different climatic conditions with enrichedsamples in 18O and 2H that clearly correlated with thepresent-day recharge, while the most depleted samplereflected an old recharge. Such processes of variablerecharge in aquifers over different climatic periods arealso evidenced in the Aquitaine Basin, the Paris BasinChalk aquifer and in Lorraine, Poland and PortugalBasins. Depleted values in the confined part of this aquifer

can be related to a lower recharge temperature at thetime of infiltration corresponding to the Pleistocene.

In addition to this process, the conceptual modeldeveloped from the Chalk aquifer in the London Basinshows the complexity of the recharge over time andmixing processes. The water evolved from connateCretaceous marine water repeatedly mixed with freshmeteoric water since the Late Tertiary, leading to theobserved d18O–d2H variation. Such complex recharge-mixing processes are also evident in the case inGermany, the Pannonian Basin, and, more particularly,the example in Switzerland.

AcknowledgementsThis work was financially supported within the scope ofthe research partnership between BRGM and WaterAgency (Adour Garonne) through the CARISMEAUproject (http//carismeau.brgm.fr). Two anonymous re-viewer are acknowledged for providing helpful reviews ofthis manuscript.

References[1] W.G. Darling, Quat. Sci. Rev. 23 (2004) 743.

[2] Ph. Negrel, E. Petelet-Giraud, A. Brenot, in: Ph. Quevauviller,

A.M. Fouillac, J. Grath, R. Ward (Editors), Groundwater Quality

Assessment and Monitoring, Wiley & Sons Ltd., Chichester, West

Sussex, UK, 2009, p. 331.

[3] P. Fritz, J.C. Fontes, Handbook of Environmental Isotope Geochem-

istry, Vols. 1–5, Elsevier, Amsterdam, The Netherlands, 1980.

[4] I.D. Clark, P. Fritz, Environmental Isotopes in Hydrology, CRC

Press/Lewis Publishers, Boca Raton, FL, USA, 1997.

[5] P.K. Aggarwal, J.R. Gat, K.F.O. Froehlich (Editors), Isotopes in the

Water Cycle; Past, Present and Future of a Developing Science,

Springer, Heidelberg, Germany, 2005.

[6] W.G. Darling, J. Human Evol. 60 (2011) 417.

[7] W. Dragoni, B.S. Sukhija, in: W. Dragoni, B.S. Sukhija (Editors),

Climate Change and Groundwater, vol. 288, Geological Society,

London, 2008, p. 1 Special Publ.

[8] IPCC report, in: IPCC Third Assessment Report - Climate Change

2001 (accessible at http://www.ipcc.ch/publications_and_data/

publications_and_data_reports.shtml#1).

[9] IPCC report, in: IPCC Fourth Assessment Report - Climate Change

2007 (accessible at http://www.ipcc.ch/publications_and_data/

publications_and_data_reports.shtml#1).

[10] D.Z. Sun, F. Bryan, Geophysical Monograph 189, American

Geophysical Union, Washington, D.C., USA, 2010.

[11] P.A. Allen, Earth Surface Processes, Blackwell Science, Oxford,

UK, 1997.

[12] M. Hulme, T.J. Osborn, T.C. Johns, Geophys. Res. Lett. 25 (1998)

3379.

[13] H. Craig, Science (Washington, DC) 133 (1961) 1702.

[14] J.R. Gat, Groundwater, in: J.R. Gat, R. Gonfiantini (Editors). Stable

Isotopes Hydrology, Deuterium and Oxygen 18 in the Water

Cycle, Technical report series N� 210, IAEA, Vienna, Austria,

1981, pp. 223–240.

[15] Y. Yurstsever, Worldwide Survey of Isotopes in Precipitation,

IAEA Report, Vienna, Austria, 1975.

[16] International Atomic Energy Agency (IAEA), GNIP Maps and

Animations, IAEA, Vienna, Austria, 2001 (accessible at http://

isohis.iaea.org).

[17] R. Millot, E. Petelet-Giraud, C. Guerrot, Ph. Negrel, Appl.

Geochem. 25 (2010) 1510.


http://www.ipcc.ch/publications_and_data/publications_and_data_reports.shtml#1




http://isohis.iaea.org

http://isohis.iaea.org


[18] W.G. Darling, J.C. Talbot, Hydrol. Earth Syst. Sci. 7 (2003) 163.

[19] W.G. Darling, A.H. Bath, J.C. Talbot, Hydrol. Earth Syst. Sci. 7

(2003) 183.

[20] A. Plata-Bedmar, Composicion Isotopica de las Precipitaciones y

Aguas Subterraneas de la Peninsula Iberica, Centro des Estudios

de Tecnicas Aplicadas, Madrid, Spain, 1994.

[21] A. Longinelli, E. Selmo, J. Hydrol. 270 (2003) 75.

[22] A. Longinelli, E. Anglesio, O. Flora, P. Iacumin, E. Selmo, J.

Hydrol. 329 (2006) 471.

[23] E.R.Th. Kerstel, R. van Trigt, N. Dam, J. Reuss, H.A.J. Meijer,

Anal. Chem. 71 (1999) 5297.

[24] G. Lis, L.I. Wassenaar, M.J. Hendry, Anal. Chem. 80 (2008) 287.

[25] M. Schmidt, K. Maseyk, C. Lett, P. Biron, P. Richard, T. Bariac, U.

Seibt, Geophys. Res. Abs. 13 (2011) EGU2011-9375.

[26] L. Andre, PhD Thesis, Universite de Bordeaux 3, France, 2002.

[27] W.M. Edmunds, in: P.K. Aggarwal, J.R. Gat, K.F.O. Froelich

(Editors), Isotopes in the Water Cycle; Past, Present and Future of

a Developing Science, Springer, Heidelberg, Germany, 2005, pp.

341–352.

[28] C. Le Gal La Salle, C. Marlin, S. Savoye, J.C. Fontes, Appl.

Geochem. 11 (1996) 433.

[29] H. Jirakova, F. Huneau, H. Celle-Jeanton, Z. Hrkal, P. Le

Coustumer, J. Hydrol. 368 (2009) 1.


[30] J.M. Matray, M. Lambert, J.C. Fontes, Appl. Geochem. 9 (1994)

297.

[31] W. Kloppmann, L. Dever, W.M. Edmunds, Appl. Geochem. 13

(1998) 593.

[32] K.M. Hiscock, P.F. Dennis, P.R. Saynor, M.O. Thomas, J. Hydrol.

180 (1996) 79.

[33] F. Dennis, J.N. Andrews, A. Parker, M. Wolf, Appl. Geochem. 12

(1997) 763.

[34] T. Elliot, J.N. Andrews, W.M. Edmunds, Appl. Geochem. 14

(1999) 333.

[35] H. Celle-Jeanton, F. Huneau, Y. Travi, W.M. Edmunds, Appl.

Geochem. 24 (2009) 1198.

[36] W. Kloppmann, Ph. Negrel, J. Casanova, H. Klinge, K. Schelkes, C.

Guerrot, Geochim. Cosmochim. Acta 65 (2001) 4087.

[37] I. Varsanyi, J.M. Matray, L.O. Kovacs, Chem. Geol. 140 (1997)

89.

[38] A. Zuber, S.M. Weise, J. Motyka, K. Osenbruck, K. Rozanski, J.

Hydrol. 286 (2004) 87.

[39] P. Galego-Fernandes, P.M. Carreira, J. Hydrol. 361 (2008)

291.

[40] F.J. Pearson, W. Balderer, H.H. Loosli, B.E. Lehmann, A. Matter, T.

Peters, H. Schmassmann, A. Gautschi, Studies in Environmental

Science 43, Elsevier, Amsterdam, The Netherland, 1991.

isotopes in groundwater as indicators of climate changes

Documents