Multiple environmental tracer approaches to resolve the age structure of groundwater in European aquifers

Inauguraldissertation der Philosophisch-naturwissenschaftlichen

Fakultät der Universität Bern

vorgelegt von

Jose Antonio Corcho Alvarado

von Fomento, Sancti Spiritus, Kuba

Leiter der Arbeit: Prof. Dr. Thomas Stocker

Abteilung für Klima- und Umweltphysik Physikalisches Institut der Universität Bern

1

2

Multiple environmental tracer approaches to resolve the age structure of groundwater in European aquifers Inauguraldissertation der Philosophisch-naturwissenschaftlichen Fakultät der Universität Bern vorgelegt von

Jose Antonio Corcho Alvarado von Fomento, Sancti Spiritus, Kuba Leiter der Arbeit: Prof. Dr. Thomas Stocker Dr. Roland Purtschert Abteilung für Klima- und Umweltphysik Physikalisches Institut der Universität Bern Von der Philosophish-naturwissenschftlichen Fakultät angenommen. Der Dekan Bern, den 17 November 2005 Prof. Dr. P. Messerli

3

4

“Water for people, water for life”

5

Contents 1. Introduction and thesis summary……………………………………………........ 72. 36Cl in modern groundwater dated by a multi-tracer approach (3H/3He, SF6,

CFC-12 and 85Kr): a case study in quaternary sand aquifers in the Odense Pilot River Basin, Denmark………...……………………………...……………... Abstract 2.1. Introduction and aquifer characterization ..…………………………………….. 2.2. Methods………………………………………………………………………… 2.3. Results and Discussion…………………………………………………………. 2.3.1. Measurements………………………………………………………………… 2.3.2. Tracer dating………………………………………………………………….. 2.3.3. Origin of 36Cl in young groundwaters………………………………………… 2.4. Conclusions……………………………………………………………………... 2.5. Acknowledgements……………………………………………………………... 2.6. References……………………………………………………………………….

21

222424242527303131

3. Constraining groundwater age distribution using 39Ar: A multiple environmental tracer (3H/3He, 85Kr, 39Ar and 14C) study in the semiconfined Fontainebleau Sands aquifer (France)…………………………………………… Abstract 3.1. Introduction……………………………………………………………………... 3.2. Site characterization…………………………………………………………...... 3.3. Methods…………………………………………………………………………. 3.3.1. Field and laboratory investigations…………………………………………… 3.3.2. Strategy of interpretation of tracer data………………………………………. 3.3.2.1. Lumped parameter models and input function……………………………… 3.3.2.2. The inverse fitting procedure……………………………………………….. 3.4. Results and discussion…………………………………………………………... 3.4.1. Investigation of the young water components………………………………… 3.4.2. Investigation of the old water components……………………………………. 3.4.3. Analysis of the tracer methods………………………………………………... 3.5. Conclusions……………………………………………………………………... 3.6. Acknowledgements……………………………………………………………... 3.7. References………………………………………………………………………. Annex 1. 37Ar and 39Ar underground production……………………………………. Annex 2. Noble gases………………………………………………………………...

35

36373939404144464749515252525657

4. Groundwater dating in the Turonian and Cenomanian aquifers of the Bohemian Cretaceous Basin: A first step in getting insights on underground processes and recharge conditions……………………………………………...... Abstract 4.1. Introduction……………………………………………………………………... 4.2. Study area……………………………………………………………………….. 4.3. Methods………………………………………………………………………….

59

606062

6

4.4. Results and discussion…………………………………………………………... 4.4.1. Hydrochemistry……………………………………………………………….. 4.4.1.1. Turonian sandstone aquifer…………………………………………………. 4.4.1.2. Cenomanian sandstone aquifer……………………………………………… 4.4.2. Groundwater dating…………………………………………………………… 4.4.2.1. Turonian sandstone aquifer…………………………………………………. 4.4.2.2. Cenomanian sandstone aquifer……………………………………………… 4.4.3. Investigating underground processes and recharge conditions……………….. 4.4.3.1. Groundwater flow velocity in the Cenomanian aquifer…………………….. 4.4.3.2. Mineral dissolution rates. Chemical tracers as time indicators……………... 4.4.3.3. 3He and 4He in groundwater from the Cenomanian aquifer………………… 4.4.3.4. Recharge conditions and palaeoclimate…………………………………….. 4.5. Conclusions……………………………………………………………………... 4.6. Acknowledgements……………………………………………………………... 4.7. References……………………………………………………………………….

636363646666687171727375777878

5. The age and origin of the Bath thermal waters…………………………………. 5.1. Introduction…………………………………………………………………….. 5.2. Methods………………………………………………………………………… 5.3. Results………………………………………………………………………….. 5.4. Comparison with earlier analyses………………………………………………. 5.5. Interpretation of the data………………………………………………………... 5.6. Conclusions……………………………………………………………………... 5.7. References……………………………………………………………………….

8383858588909696

6. Publications, conference abstracts and reports………….……………………… 997. Acknowledgements………………………………………….…….………………. 1018. Curriculum Vitae…………………………………………….……………………. 103

7

Chapter 1

Introduction

Some two thirds of the freshwater on Earth is locked up in glaciers and permanent snow cover, and only about one third is accessible in lakes, rivers, man-made reservoirs and aquifers. Water accumulated in aquifers (groundwater) is by far the largest reservoir of liquid freshwater on Earth. This makes groundwater the key freshwater resource and its assessment and sustainable management a major concern worldwide. Problems such as over-exploitation and pollution are the forefront of the groundwater research. But other problems such as the impacts of urbanisation, deforestation and climate change on groundwater resources receive as well a significant attention.

As a consequence of the increasing interest in groundwater reservoirs and their dynamics, a large number of techniques and methods have been developed in many different Earth science areas. Hydrogeological research is strongly interdisciplinary and based on geological, hydraulic, hydrochemical, microbial, chemical and isotopic tracer research. Each of them with its specific efficiency and scope of application. This thesis concentrates mainly on the application of tracer methods to investigate groundwater systems, but also on their combination with other traditional methods and simple mathematical modelling for answering important questions concerning the systems.

The use of environmental tracer methods for investigating hydrological processes has been widely developed in the past few decades. Nowadays, the tracer methods are for example routine tools for obtaining information about the flow dynamics of groundwater (Fritz and Fontes, 1980; Fritz and Fontes, 1986; Mazor, 1991; Clark and Fritz, 1997; Cook and Herczeg, 1999). The most frequently used environmental tracers are: 3H/3He (Schlosser, 1989; Solomon and Cook, 1999); 85Kr and 39Ar (Loosli, 1983; Loosli et al., 1999); CFCs - chlorofluorocarbons (Plummer and Busenberg, 1999); SF6 (Busenberg and Plummer, 1999); 14C (Kalin, 1999); the stable noble gases (Stute and Schlosser, 1999) and the stable isotopes of water 2H and 18O (Coplen et al., 1999).

The timescales of application of the environmental tracers expands from a few seconds to more than one million years (Fig. 1) allowing the study of a large range of hydrogeological processes. One of the main applications of environmental tracer methods in groundwater hydrology is for determining the timescale of the water transport in the soil subsurface, i.e., the groundwater age (groundwater dating). The parameter groundwater age is of prime interest for quantifying other aquifer parameters such as the groundwater flow velocity, and the recharge rate, which are of significant importance for effective groundwater management and for assessing the consequences of anthropogenic contamination. The tracer methods have

8

been as well used for investigating mixing of different groundwater bodies, for determining the sources of groundwater and location of recharge areas, for studying the sources and transport rates of contaminants and dissolved elements, for calculating the degradation rates of dissolved compounds in groundwater (Cook and Böhlke, 1999) and for calibrating groundwater numerical flow and transport models (Mattle et al., 1999; Stute and Schlosser, 2000).

Figure 1. Dating range of the most important groundwater tracers.

Thesis summary

In the frame of this thesis the tracer methods were applied in several groundwater bodies that range from shallow and vulnerable aquifers to deep and confined aquifers, and also in a geothermal system. The groundwater bodies investigated are: a) the shallow and semiconfined Odense sands aquifer, located in Denmark; b) the semiconfined Fontainebleau sands aquifer, located in France; c) the shallow and semiconfined Turonian aquifer and the deep and confined Cenomanian aquifer of the Bohemian Cretaceous Basin, located in the Czech Republic; and d) the Bath thermal water, located in Great Britain (Fig. 2). The research carried out in these systems, with the exception of the Bath thermal water, was partially supported by the European Project “Baseline” (EVK1-CT-1999-00006). The project aimed to establish criteria for defining the natural groundwater quality background and to develop a standardised Europe-wide approach which may be used in emerging water directives. Such a standard, based on geochemical principles, is needed as a reference to be able to assess quantitatively whether or not anthropogenic pollution is taking place. The baseline water quality is mainly controlled by the duration of water-rock interaction, a parameter that is directly related to the age of groundwater. This work was meant to provide information about this last parameter.

9

Figure 2. Map of Europe with the locations of the investigated aquifers.

The research of the groundwater systems, of which very scarce or vague information was available before this work, relies on measurements in groundwater of a large set of environmental tracers (e.g. 3H, 3He, 85Kr, SF6, CFCs, 39Ar, 14C and 4He). The combination of several tracers enables other processes or parameters of the aquifer system than the mean residence time (e.g. mixing processes, dispersion, sorption, degradation, degassing, contamination, etc.) to be investigated (Ekwurzel et al., 1994; Plummer et al., 2003, Corcho et al., 2005); and also refining the interpretation of groundwater age. The investigated groundwater bodies show different hydrogeological conditions; therefore slightly different methods were required in each case. More details about the specific methods applied are given further below. The interpretation of the tracer concentrations is commonly carried out by models that try to mathematically describe the age distribution of sampled groundwater. The limited number of sampling sites in the aquifer systems investigated in the frame of this thesis suggested the use of lumped parameter approaches for the assessment of groundwater dynamics (Zuber, 1986; Zuber and Maloszewski, 2001). The choice of an age weighting function in the lumped parameter models that appropriately represents the hydrogeological situation in each study case is validated using the measured tracer data. The use of more complicated models such as numerical flow and transport models is not justified in most of the systems due to the lack of sufficient information.

A brief introduction to the subject areas is given in the following. The thesis is organized in four main chapters which are written in an article form. Each article is intended for publication or has already been submitted. A short summary of the four main chapters of this thesis and some general remarks on the findings are also given. Each chapter concentrates in the characterization of one aquifer system, and they are named as follows:

10

a) Chapter 2: 36Cl in modern groundwater dated by a multi tracer approach (3H/3He, SF6, CFC-12 and 85Kr): A case study in Quaternary sand aquifers in the Odense Pilot River Basin, Denmark. This chapter has been published in: Applied Geochemistry (Corcho Alvarado et al., 2005)

b) Chapter 3: Constraining groundwater age distribution using 39Ar: a multiple environmental tracer (3H/3He, 85Kr, 39Ar and 14C) study in the semi-confined Fontainebleau Sands aquifer (France). This chapter has been submitted to: Journal of Hydrology (Corcho Alvarado et al., submitted).

c) Chapter 4: Groundwater dating in the Turonian and Cenomanian aquifers of the Bohemian Cretaceous Basin: A first step in getting insights on underground processes and recharge conditions. This chapter is in preparation for publication (Corcho Alvarado et al., in preparation).

d) Chapter 5: The age and origin of the Bath thermal waters. This chapter is in preparation for publication (Edmunds et al., in preparation).

Odense sands aquifer, Denmark

In many areas of the world, daily water use is based mainly on groundwater recharged in the last decades. Contamination (e.g. from industry, agriculture) is a typical signature in most of these resources. This is the case of the shallow semi-confined Odense Sands aquifer, which is located on the Island of Funen around the city of Odense, in Denmark (Fig. 2). Since one or two decades ago, the quality of the abstracted groundwater has been affected by the increasing contamination by pesticides, pesticide degradation products and chlorinated solvents (Hinsby et al., 2003). This contamination has created severe problems for the water supply company of the city of Odense. Also the increase of groundwater mineral contents (e.g. hardness) led to the closure of wells. These problems deteriorate groundwater quality and deplete the potable water reserves in the area, and need therefore to be investigated.

As part of this thesis, environmental tracers (3H/3He, 85Kr, SF6 and CFC-12) were measured in groundwater samples from the Odense sands aquifer in order to obtain information about the age structure of groundwater and mixing processes. This information is required in a first instance to define the baseline quality of groundwaters in the aquifer, which is further needed to be able to evaluate the impact of anthropogenic pollution. In a second instance which is not in the scope of this thesis, the tracer and dating results were used for calibration of an integrated transient 3D hydrological model that includes groundwater/surface water interaction (Hinsby et al., 2003; Troldborg, 2005). The calibrated model was applied for analysis of groundwater/surface water interaction and the general response in the hydrological system to long-term aquifer exploitation and changes in abstraction.

The investigated area is situated in a complex setting of Quaternary glaciofluvial sand aquifers with confining sandy and clayey tills. The Odense sands which constitute the main aquifer on the island overlie a sequence of mainly Palaeocene marls and clays. Recharge occurs through sand windows and lenses and through fractures and root holes in the tills at a rate of about 240 mm per year (Hinsby et al., 2003). A Palaeocene Limestone aquifer is underlying the Odense shallow aquifer and the Palaeocene marls and clays, and in some areas

11

deep wells may extract water from this aquifer or create hydraulic contact to the sand aquifers above.

The simultaneous application of several tracer methods (3H/3He, 85Kr and SF6) enabled not only to date the groundwater samples, but also to identify and quantify processes like mixing of different groundwater bodies and dispersive mixing within the aquifer (Corcho Alvarado et al., 2002; Corcho Alvarado et al., 2005). Moreover, the multiple tracers approach provided a cross-check of the behaviour of CFC-12. Hence, it was identified that chemical degradation of CFC-12 is occurring in the aquifer. It is known that this tracer can be degraded under reducing conditions (Busenberg and Plummer, 1992; Plummer and Busenberg, 1999), which are the conditions that prevail in this aquifer. Based on the tracer dating results, the degradation rates of CFC-12 in the aquifer conditions were estimated resulting in values that are comparable to degradation rates found at other aquifer sites in Denmark (Hinsby et al., 2004).

The behaviour of the radioisotope 36Cl in shallow groundwaters of the Odense sands aquifer was also investigated. 36Cl (half-life: 301,000 years) is a well established method to date very old groundwater from 50,000 up to 1 million years (Bentley et al., 1986a; Bentley et al., 1986b; Phillips et al., 1986; Andrews et al., 1994; Lehmann et al., 2003) and to investigate groundwater infiltration processes (Phillips et al., 1988; Fabryka-Martin, 1993; Guerin, 2001). As a consequence of the thermonuclear bomb testing elevated concentrations of 36Cl by over two orders of magnitude were generated in the environment in the late 1950s to early 1960s (Bentley et al., 1986a; Synal et al., 1990). Consequently, this isotope has been proposed to trace modern recharged groundwaters (Bentley et al., 1982; Bentley et al., 1986a; Clark and Fritz, 1997). Some studies have indicated that bomb produced 36Cl is recycled in the environment (Milton et al., 1994; Milton et al., 1997; Cornett et al., 1997; Scheffel at al., 1999; Blinov et al., 2000; Milton et al., 2003). The occurrence of this process has been further confirmed by the present investigation. Reconstructed local atmospheric 36Cl fallout rates based on 36Cl measurements in groundwater samples from the Odense aquifer agree within the range of variations with direct measurements of the 36Cl fallout rates in Europe, but exceed by almost a factor of two the magnitude of the fallout rates predicted from ice core measurements (Corcho Alvarado et al., 2005). Such a large disagreement is attributed to the recycling of 36Cl, a process that is difficult to quantify. It was concluded that 36Cl can only be used as an indication for the presence of modern recharged groundwater.

Fontainebleau sands aquifer, France

Many shallow aquifers in the world are nearly exhausted due to the high water demand for public supply. In these sites, recent recharged groundwaters are not replenished fast enough compared to the velocity of water abstraction. Also, shallow aquifers are, for example, very often chemically and/or biologically polluted. As a result of these facts and of the increasing demand for water, a special interest has focused in groundwaters with longer residence times. An aquifer that contains groundwater with this characteristic was investigated in the Paris region (France) (Fig. 2). The semiconfined Fontainebleau sands aquifer, located in the shallower part of the Paris Basin, represents a major resource for water supply in the area of Paris. However, previous to this work only minor studies were carried out to characterize the aquifer (Bergonzini, 2000).

12

The Oligocene Fontainebleau sands aquifer is embedded between two clayey layers: above is the Beauce formation which was altered by diagenesis from limestone to millstone and clay; and below are Oligocene marls which separate the Fontainebleau Sands from the underlying Eocene multi-layered aquifer. The aquifer is constituted by very fine, well-sorted silica grains. It has a thickness of 50-70 m. The hydrogeological situation in the Fontainebleau aquifer (e.g. spatially extended recharge, large screen intervals, and possible leakage from deeper aquifers) suggested a large spread of groundwater ages from modern to a few hundreds years. Therefore, the frequently used 3H, 3He and 85Kr were combined to identify and date groundwater recharged in the last 50 years. The highlight of this investigation is the analysis of 39Ar (half-life of 269 years) which further constrained the distribution of groundwater ages in an intermediate range from 50 to 1000 years. Presently, 39Ar is the only method suitable to precisely investigate groundwater with ages between 50 and 1000 years. The 39Ar dating method was developed at the Physics Institute of the University of Bern more than 20 years ago (Oeschger et al., 1974; Loosli and Oeschger, 1980; Loosli; 1983) and still today is the only laboratory in the world capable of applying this method. 39Ar has been used to study groundwater in, for example, the East Midlands Triassic Sandstone aquifer, in UK (Andrews et al., 1984); the Stripa granites in Sweden (Andrews et al., 1989); the Muschelkalk formation in northern Switzerland (Pearson et al., 1990) and the confined gravel aquifer in the Glatt Valley, Switzerland (Beyerle et al., 1998). Measurements of 14C in dissolved inorganic carbon were used to investigate the occurrence of groundwater components with ages between 1000-30000 years.

Recharge to the Fontainebleau Sands aquifer occurs through a relatively thick unsaturated zone of more than 20 meters; therefore the time lag of the young environmental tracers (e.g. 3H, 85Kr) from the soil surface to the water table had to be taken into account for dating groundwater (Cook and Solomon, 1995). Time-lags from 1 to 6 years for 85Kr and from 10 to 40 years for 3H were calculated with a one-dimensional transport model which considers advection, diffusion and decay processes in the unsaturated zone (Corcho Alvarado et al., 2003, Corcho Alvarado et al., 2004). Then, the lumped parameter approach was applied to investigate aquifer parameters using as input functions the tracer concentrations at the water table calculated with the one-dimensional model. The parameters that best fitted the model to the measured tracer concentrations were determined by an inverse approach that allowed the parameter errors to be estimated.

It was concluded that small fractions of the sampled groundwater (<50%) follow exponential distributions of ages with mean residence times varying from 1 to 20 years. Large fractions of groundwater follow different age distributions with older mean residence times. The characterization of these fractions was not possible with the applied young tracers. Previous works have demonstrated that in well mixed and highly dispersive aquifers where a large spread of residence times is occurring the use of a dating tracer like 39Ar that can precisely date intermediate ages is indispensable (Weissmann et al., 2002). Hence, the use of 39Ar was required to resolve the whole age structure of the abstracted groundwater (Corcho Alvarado et al., 2003; Corcho Alvarado et al., submitted). Ages for the old groundwater components below 600 years were derived from the interpretation of the 39Ar data. Finally, mean residence times for the abstracted groundwater ranging between 120 and 450 years were calculated. 14C measurements confirmed that leakage of very old groundwater from underlying aquifers is not occurring in the area, contrary to what was suggested by previous studies (Bergonzini, 2000). Based on the tracers’ results, a conceptual model for the investigated area of the aquifer was developed. This model is important for the future management of the groundwater resources in the area.

13

0 100 200 300 400 500 6004

5

6

7

8

9

10

11

0 100 200 300 400 500 6009.0

9.5

10.0

10.5

11.0

SM

CGEBSA

SLP5

SLP4

IMR

LRN10

[D-exc] = (0.005±0.003) MRT + (6.3±0.9) R=0.64

Deu

teriu

m e

xces

s (‰

)

Mean residence time (yrs)

LRN10

SM

CGEB

SA SLP4

IMR SLP5

[NGT] = (0.003±0.001) MRT + (9.2±0.2) R=0.98

Nob

le g

as te

mpe

ratu

re (o C

)

Mean residence time (yrs)

Figure 3. Correlation between a) the deuterium excess and b) the Noble gas temperatures with the groundwater mean residence time in the Fontainebleau sands aquifer, France.

Although they are not included in the discussed chapter, other results derived from the present study provide interesting hints for future research. We can mention for example, the correlations observed between the deuterium excess and the noble gas temperature with the calculated groundwater mean residence time (Fig. 3). These trends could be caused by temporal variation of certain processes or parameters, but also due to spatial variations. The trend observed in the deuterium excess for example could have different origins: a) changes in the atmospheric circulation patterns over Europe (a different moisture source component with lower d-excess); b) a steady increase of the moisture content in the location of the vapour formation in the past hundreds of years; and c) different origins for groundwater such as direct rain water infiltration, water evaporated in ponds and lakes, or mixtures of both sources. The noble gas temperatures indicate that groundwater recharge took place at different temperatures, decreasing about 1oC from older to younger groundwaters (Fig. 3b). It could be that recharge varied spatially along the surface of the aquifer. Other explanations may be related, for example, to temporal changes in climatic conditions or to variations of the soil coverage due to deforestation. A more detailed study is necessary to properly identify the origins of the observed trends.

Turonian and Cenomanian aquifers, Czech Republic

Pollution of groundwater resources has become a major problem in many areas of the world. Once groundwater is polluted, recovery is normally very slow due to its long residence time in the subsurface and slow renewal rate. We must mention also that cleaning up contaminated aquifers is usually extremely difficult and expensive, or even unfeasible. Contamination of groundwater, for example, is an important feature in some parts of the Turonian and Cenomanian aquifers of the Bohemian Cretaceous Basin which are very important sources of water in the Czech Republic (Fig. 2). The groundwater quality in these aquifers was affected as a consequence of the uranium mining that took place in the Stráz Tectonic Block, located in the northern margin of the Basin. The understanding of the groundwater flow and its age structure was therefore needed to investigate the future dispersion and impact of the contamination. Hence, environmental tracer methods (3H, 3He, 85Kr, 39Ar and 14C) were used to investigate these aquifer properties in both sites. The tracer

14

measurements were used in conjunction with other methods such as hydrochemical measurements, the noble gases (He, Ne, Ar, Kr and Xe) and the stable isotopes of the water molecule 2H and 18O.

The regional deeper artesian aquifer in Cenomanian sandstones and the regional aquifer with a free and partly confined water table in Middle Turonian sandstones are separated by low-permeability marlstone and claystone aquitards. The semiconfined Turonian sands aquifer recharge all over its supece extension and has thicknesses varying between 10 m to about 190 m. The Cenomanian aquifer is recharged by infiltrating rainwater in a 1-2km wide zone where the sands outcrop (Fig. 4). Groundwater flows from the north to south direction. This aquifer is typically 30-80 m thick in the studied area. The base of the aquifer is formed by impermeable Permian rocks and, in the region Mladá Boleslav by schists and granitic rocks. Groundwater samples were taken in wells located along similar flow paths in the region exploited for water supply and located to the east and south of the mining area.

It was concluded in this study that the semiconfined Turonian sands aquifer contains groundwater with mean residence times varying from modern to about 200 years (Corcho Alvarado et al., 2004). Elevated nitrate concentrations in groundwater indicated the vulnerability of the aquifer to surface pollution. The vertical distribution of dissolved substances suggested a stratification of the aquifer.

Figure 4. Scheme of the Cenomanian aquifer with indication of the location of the wells and the

calculated groundwater age. Blue arrows indicate the flow direction north-south.

In the Cenomanian aquifer, groundwaters with ages ranging from a few hundreds of years to more than 20 000 years were dated (Fig. 4) (Corcho Alvarado et al., 2004; Corcho Alvarado et al., in preparation). No indication of pollution in groundwater was detected in the investigated area. The calculated tracer ages showed a good linear correlation with the

15

distance from recharge, suggesting that piston flow is a good approximation for the flow type in the aquifer. The groundwater chemical evolution along the flow path and the main geochemical processes occurring in the aquifer were as well studied. A vertical flux of helium from deeper layers (crust and mantle) into the aquifer was detected, with a value approximately one order of magnitude lower than that predicted for the degassing of the whole continental crust. Using typical crustal and mantle helium ratios, different sources of helium were distinguished and quantified. It was concluded that the main external source of helium are crustal rocks, although a large contribution of mantle helium was found in one well. The noble gas temperatures and the stable isotope signature showed that groundwater recharged under different climatic conditions. In one well, for example, recharge took place at cooler temperatures than the present day annual average temperature. This result agrees with the dating, which predicted that groundwater in this well recharged during the end of the Pleistocene when climate with lower air temperatures prevailed.

The nearly similar deuterium excess values observed in groundwater samples from the Cenomanian aquifer suggested that relative humidity over the subtropical regions of the North Atlantic Ocean and circulation patterns of the atmosphere over Europe have not changed considerably during the investigated period (Corcho Alvarado et al., in preparation). This result agrees with the observations made in other aquifers over Europe (Rozanski, 1985, Huneau et al., 2001). The deuterium excess in this aquifer is characterized by a slight increasing trend during the Holocene. Somewhat similar trends were observed in Greenland records and were attributed to an increase of low-altitude annual mean insolation (warming ocean temperatures) and a decrease of high-latitude annual mean insolation (cooling ocean temperatures) in response to the progressive Holocene decrease in obliquity (Masson-Delmotte et al., 2005). This phenomenon should enhance the relative contribution of low-latitude moisture sources.

Lower deuterium excess values in the samples from the Turonian aquifer containing young groundwater (up to 200 yrs) than in the samples from the Cenomanian aquifer containing old groundwater (older than 600 years) suggested either a significant change in the atmospheric circulation patterns over Europe (a different moisture source component with lower d-excess) or an increase of the moisture content in the location of the vapour formation (this would increase the kinetic fractionation and consequently would lower the d-excess) in the last 200 to 600 years. The stable isotope ratios δ2H and δ18O in groundwater from the Cenomanian aquifer (Holocene and Pleistocene waters) followed the same continental gradient observed by Rozanski (1985) in recent infiltration waters and in precipitation; therefore a similar atmospheric circulation pattern was deduced. It was concluded then that the most probable reason for the decrease of the d-excess at present day is a slight increase in the moisture content (<5%) at the ocean surface. No evidence about this process was found in any other investigation. Future studies with a larger number of samples are necessary to fully understand the origin of this tendency in the aquifer.

Bath thermal waters, United Kingdom

Bath has the only hot springs in Great Britain with temperatures in excess of 40 oC (Fig. 2). It was believed that the source of these thermal waters is rainfall which takes probably up to a few thousand years to sink to a depth of about 2km, where it is heated by high temperature rocks and then resurfaces through a fracture zone or fault beneath the centre of Bath (Figure

16

5). Geochemical studies have suggested that the water could be around 6000±2000 years old. However hydrogeological modelling has predicted an age much lower than this, possibly as young as 500 years (Edmunds and Miles 1991). In recent years a new project was devised to research, explore and monitor the thermal springs to achieve greater understanding of their source/s in order to ensure their protection for future generations. As part of this project, detailed chemistry, radioactive and stable isotope studies on the water and the solutes, and investigation of dissolved gases were conducted to investigate the origin and age of the water.

The present work confirmed earlier studies that suggested that there may be a degree of mixing of younger near-surface water with a much older and deeper source. It is shown in this work that the discharge contains up to 5% of modern water less than 50 years old based on 85Kr and CFCs data, which is probably derived from Mesozoic strata some tens of metres below the point of emergence. The interpretation of the 39Ar data indicated that the age of the old component must be older than 1000 years. However dissolved noble gas and stable isotope contents constrained further this age to be less than 12,000 years. Qualitative evidence (for example enriched 13C and a likely negligible, residual 14C) suggests the water to be nearer the upper age limit, and a range of 6-10 kyr is proposed.

Figure 5. Scheme of the Bath thermal system.

References Andrews J.N., Balderer W., Bath A.H., Clausen H.B., Evans G.V., Florkowski T., Goldbrunner J.E., Ivanovich

M., Loosli H.H., Zojer H. (1984) Environmental isotope studies in two aquifer systems: A comparison of groundwater dating methods. In: Isotope Hydrology 1983 (IAEA-SM-270). IAEA, Vienna, p 535-577

17

Andrews J.N., Davis S.N., Fabryka-Martin J., Fontes J-CH., Lehmann B.E., Loosli H.H., Michelot J-L., Moser H., Smith B., Wolf M. (1989) The in situ production of radioisotopes in rock matrices with particular reference to the Stripa Granite. Geochim. Cosmochim. Acta 53, 1803-1815.

Andrews J., Drimme R., Loosli H.H., and Hendry M. (1991a) Dissolved gases in the milk river aquifer, Alberta, Canada. Applied Geochemistry, 6, 393-403.

Andrews J., Florkowski T., Lehmann B., and Loosli H.H. (1991b) Underground production of radionuclides in the milk river aquifer, Alberta, Canada. Applied Geochemistry, 6, 425-434.

Andrews J., Edmunds W.M., Smedley P.L., Fontes J.-Ch., Fifield L.K. and Allan G.L., (1994). Chlorine-36 in groundwater as a palaeoclimatic indicator: the East Midlands Triassic sandstone aquifer (UK). Earth Planet. Sci. Lett., 122: 159-171.

Bentley H.W.; Phillips F.M.; Davis S.N.; Gifford S.; Elmore E.; Tubbs L.E. and Gove H.E. (1982). Thermonuclear 36C1 pulse in natural waters. Nature 300 737-740.

Bentley H.W., Phillips F.M. and Davis S.N. (1986a). Chlorine-36 in the terrestrial environment, in Handbook of Environmental Isotope Geochemistry, vol. 2, ed. by P.Fritz and J.-Ch. Fontes, pp. 427-480.

Bentley H.W.; Phillips F.M.; Davis S.N.; Airey P.L.; Calf G.E.; Elmore D.; Habermehl M.A. and Torgenson T. (1986b) Chlorine-36 dating of very old ground water: I. The Great Artesian Basin, Australia. Water Resour. Res. 22 1991-2002.

Bergonzini L. (2000) Caractérisation géochimique de la nappe des Sables de Fontainebleau (abstract). 18ème Réunion des Sciences de la Terre, Paris.

Beyerle U., Purtschert R., Aeschbach-Hertig W., Imboden D.M., Loosli H.H., Wieler R., Kipfer R. (1998) Climate and groundwater recharge during the last glaciation in an ice-covered region. Science 282:731-734

Blinov A., Massonet S., Sachsenhauser H., Stan-Sion C., Lazarev V., Beer J., Synal H.-A., Kaba M., Masarik J., Nolte E. (2000). An excess of 36Cl in modern atmospheric precipitation. Nucl. Instrum. Meth. Phys. Res. B 172, 537-544.

Busenberg E. and Plummer L.N. (1992). Use of chlorofluorocarbons (CCl3F and CCl2F2) as hydrologic tracers and age-dating tools: the alluvium and terrace system of Central Oklahoma. Water Resour. Res. 28, 2257-2283.

Busenberg E. and Plummer L.N. (1999). Dating young groundwater with sulfur hexafluoride: natural and anthropogenic sources of sulfur hexafluoride. Water Resour. Res. 36, 3011-3030.

Clark, I.D. and Fritz, P. (1997). Environmental isotopes in hydrogeology. Lewis Publishers, Boca Raton, FL, 328 pp.

Cook P.G. and Solomon D.K. (1995) Transport of atmospheric trace gases to the water table: Implications for groundwater dating with chlorofluorocarbons and krypton-85. Water Resour. Res. 31(2), 263-270.

Cook P.G. and Böhlke J-K. (1999). Determining timescales for groundwater flow and solute transport. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 1-30.

Cook P.G. and Herczeg A.L. (2000). Environmental tracers in subsurface hydrology. Kluwer Academic Publishers, Boston, MA, 529 pp.

Coplen T.B., Herczeg A.L. and Barnes Ch. (1999). Isotope engineering- using stable isotopes of the water molecule to solve practical problems. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 79-110.

Corcho Alvarado J.A., Purtschert R., Hofer M., Aeschbach-Hertig W., Kipfer R., Troldborg L., Hinsby K. (2002). Comparison of residence time indicators (3H/3He, SF6, CFC-12 and 85Kr) in shallow groundwater: a case study in the Odense aquifer, Denmark. Goldschmidt conference, Davos. Geochim. Cosmochim. Acta 66: A152.

Corcho Alvarado J.A.; Purtschert R.; Chabault C.; Barbecot F.; Rueedi J.; Schneider V., Aeschbach-Hertig W.; Kipfer R.; Loosli H.H.; Dever L. (2003). Origin and temporal evolution of the groundwater in the Fontainebleau Sands Aquifer (France) investigated using 3H, 85Kr, 39Ar, 14C and stable noble gases. Int. Symposium on Iso. Hydrol. and Integrated Water Res. Manag., IAEA, Austria.

18

Corcho Alvarado J.A., Purtschert R., Barbecot F., Chabault C., Rüedi J., Schneider V., Aeschbach-Hertig W., Kipfer R., Loosli H.H. (2004). Tracer transport in the unsaturated zone of the Fontainebleau sands aquifer. International Workshop on the Application of Isotope Techniques in Hydrological and Environmental Studies, IAEA/UNESCO, Paris, France.

Corcho Alvarado J.A., Purtschert R. and Pačes T. (2004). Establishing timescales of groundwater residence times based on environmental tracer data: a study of the Turonian and Cenomanian aquifers of the Bohemian Cretaceous Basin, Czech Republic. In: 32nd International Geological Congress, Florence, Italy.

Corcho Alvarado J.A., Purtschert R., Hinsby K., Troldborg L., Hofer M., Kipfer R., Aeschbach-Hertig W., Arno-Synal H. (2005). 36Cl in modern groundwater dated by a multi tracer approach (3H/3He, SF6, CFC-12 and 85Kr): A case study in Quaternary sand aquifers in the Odense Pilot River Basin, Denmark. Applied Geochemistry, V. 30, 3, 599-609.

Corcho Alvarado J.A., Purtschert R., Barbecot F., Chabault C., Rueedi J., Schneider V., Aeschbach-Hertig W., Kipfer R. and Loosli H.H. Constraining groundwater age distribution using 39Ar: a multiple environmental tracer (3H/3He, 85Kr, 39Ar and 14C) study in the semi-confined Fontainebleau Sands aquifer (France). submitted to Journal of Hydrology.

Corcho Alvarado J.A., Purtschert R, Pačes T., Kipfer R. and Leuenberger M. Groundwater dating in the Turonian and Cenomanian aquifers of the Bohemian Cretaceous Basin: A first step in getting insights on underground processes and recharge conditions. In preparation.

Cornett R.J., Andrews H.R., Chant L.A., Davies W.G., Greiner B.F., Imahori Y., Koslowsky V.T., Kotzer T., Milton J.C.D., Milton G.M. (1997). Is 36Cl from weapons’ test fallout still cycling in the atmosphere?. Nucl. Instrum. Meth. Phys. Res. B 123, 378-381.

Edmunds W.M. and Miles D.L. (1991). The geochemistry of the Bath thermal waters. In Kellaway G A (ed) Hot Springs of Bath pp 143-156.

Edmunds W.M., Darling W.G., Purtschert R.and Corcho Alvarado J.A.. The origin and age of the Bath thermal waters. in preparation.

Ekwurzel B., Schlosser P., Smethie Jr., Plummer L.N., Busenberg E., Michel R.L., Weppernig R. and Stute M. (1994). Dating of shallow groundwater: Comparison of the transient tracers 3H/3He, chlorofluorocarbons and 85Kr. Water Resources Research, v. 30, no. 6, p. 1693-1708.

Fabryka-Martin J. T., Wightman S. J., Murphy W. J., Wickham M. P., Caffee M. W., Nimz G. J., Southon J. R., and Sharma P. (1993). Distribution of chlorine-36 in the unsaturated zone at Yucca Mountain: An indicator of fast transport paths. In Proc. FOCUS ‘93: Site Characterization and Model Validation, Las Vegas, NV, American Nuclear Society, La Grange Park, Ill., 58-68.

Fritz P. and Fontes J.C. (1980). Handbook of environmental isotope geochemistry, Vol. 1. Elsevier, Amsterdam, 545 pp.

Fritz P. and Fontes J.C. (1986). Handbook of environmental isotope geochemistry, Vol. 2. Elsevier, Amsterdam, 557 pp.

Guerin M (2001). Tritium and 36Cl as constraints on fast fracture flow and percolation flux in the unsaturated zone at Yucca Mountain. J Contam Hydrol.; 51(3-4):257-88.

Hinsby K., Troldborg L., Purtschert R., Corcho Alvarado J.A. (2003). Integrated transient hydrological modelling of tracer transport and long-term groundwater/surface water interaction using four 30 year 3H time series and groundwater dating for evaluation of groundwater flow dynamics and hydrochemical trends in groundwater and surface water. Report to the IAEA (TECDOC in preparation by the IAEA).

Hinsby K.; Thomsen A.; Engesgaard P.; Larsen F.; Jensen K.H.; Laier T.; Busenberg E., Plummer L.N. (2004). CFC dating, and transport and degradation of CFCs in a partly anaerobic sand aquifer, Rabis Creek, Denmark, in prep.

Huneau, F., B. Blavoux, W. Aeschbach-Hertig & R. Kipfer, 2001. Palaeogroundwaters of the Valréas Miocene aquifer (Southeastern France) as archives of the LGM/Holocene climatic transition in the Western Mediterranean region. In International Conference on the Study of Environmental Change Using Isotope Techniques, IAEA, Vienna. IAEA-CN-80/24: 27-28.

Kalin R.M. (1999). Radiocarbon dating of groundwater systems. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 111-144.

19

Lehmann B.E., Love A., Purtschert R., Collon P., Loosli H.H., Kutschera W., Beyerle U., Aeschbach-Hertig W., Kipfer R., Frape S.K., Herczeg A.L., Moran J., Tolstikhin I., and Groening (2003). A comparison of groundwater dating with 81Kr, 36Cl and 4He in 4 wells of the Great Artesian Basin, Australia. Earth and Planetary Science Letters 211, 237-250.

Loosli H.H. and Oeschger H. (1980). Use of 39Ar and 14C for groundwater Dating. Radiocarbon 22(3): 863

Loosli H.H. (1983) A dating method with 39Ar. Earth Planet. Lett. 63, 51.

Loosli H.H., Lehmann B.E., Balderer W. (1989) Argon-39, argon-37 and krypton-85 isotopes in Stripa groundwaters. Geochim. Cosmochim. Acta 53, 1825-1829.

Loosli H.H., Lehmann B.E. and Smethie W.M., (1999). Noble gas radioisotopes: 37Ar, 85Kr, 39Ar, 81Kr. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 379-396.

Masson-Delmotte, V., Landais A., Stievenard M., Cattani O., Falourd S., Jouzel J., Johnsen S. J., Dahl-Jensen D., Sveinsbjornsdottir A., White J. W. C., Popp T., and Fischer H. (2005). Holocene climatic changes in Greenland: Different deuterium excess signals at Greenland Ice Core Project (GRIP) and NorthGRIP, J. Geophys. Res., 110, D14102.

Mattle N., Kinzelbach W., Beyerle U. and Huggenberger P. (1999). Calibration of a 3-dimensinal hydrodynamic transport model with tritiogenic 3He data. In (ed.), International symposium on isotope techniques in water resources development and management, IAEA, Vienna. IAEA-SM-361/45, IAEA-CSP-2/C (CD): 45-53

Mazor E. (1991) Applied chemical and isotopic groundwater hydrology. Halsted Press, New York, 273 p.

Milton J.C.D., Andrews H.R., Chant L.A., Cornett R.J.J, Davies W.G., Greiner B.F., Imahori Y., Koslowsky V.T., McKay J.W., Milton G.M. (1994). 36Cl in the Laurentian Great Lakes basin. Nucl. Instrum. Meth. Phys. Res. B 92, 440-444.

Milton J.C.D., Milton G.M., Andrews H.R., Chant L.A., Cornett R.J.J, Davies W.G., Greiner B.F., Imahori Y., Koslowsky V.T., Kotzer T., Kramer S.J., McKay J.W. (1997). A new interpretation of the distribution of bomb-produced chlorine-36 in the environment, with special reference to the Laurentian Lakes. Nucl. Instrum. Meth. Phys. Res. B 123, 382-386.

Milton G.M., Milton J.C.D., Schiff S., Cook P., Kotzer T.G. and Cecil L.D. (2003). Evidence for chlorine recycling – hydrosphere, biosphere, atmosphere – in a forested wet zone on the Canadian Shield. Appl. Geochem. 18, 1027-1042.

Oeschger H.; Gugelmann A. ; Loosli H. ; Schotterer U. ; Siegenthaler U. ; Wiest W. (1974). 39Ar dating of groundwater. In Symposium on isotope techniques in groundwater hydrology; Vienna, Austria, Vol. II, 179-190.

Pearson F.J. Jr., Balderer W., Loosli H.H., Lehmann B.E., Matter A., Peters Tj., Schmassmann H. and Gautschi A. (1991). Applied Isotope Hydrology: A Case Study in Northern Switzerland. Studies in Environmental Science #43, Elsevier Science Publishers B.V., Amsterdam, 436 p.

Phillips F.M.; H.W. Bentley; S.N. Davis; and D. Elmore (1986) Chlorine-36 dating of very old ground water: II. Milk River Aquifer, Alberta, Canada: Water Resour. Res. 22 2003-2016.

Phillips F.M.; Mattick J.L.; Duval T.A.; Elmore N.; and Kubik P.W. (1988) Chlorine-36 and tritium from nuclear- weapons fallout as tracers for long-term liquid and vapor movement in desert soils: Water Resour. Res. 24 1877-1891.

Plummer L.N. and Busenberg E. (1999). Chlorofluorocarbons. In: Cook, P.G. and Herczeg, A.L. (eds.), Environmental tracers in subsurface hydrology. Kluwer Academic Publishers, Boston, 441-478.

Plummer L.N., Böhlke J.K., Busenberg E. (2003). Approaches for groundwater dating. In Lindsey B.D., Phillips S.W., Donnelly C.A., Speiran G.K., Plummer L.N., Böhlke J.K., Focazio M.J., Burton W.C., Busenberg E. Residence times and nitrate transport in ground water discharging to streams in the Chesapeake Bay Watershed: U.S. Geological Survey Water Resources Investigations Report 03-4035, 12-24.

Purtschert R., Lehmann B.E. and Loosli H.H. (2001a). Groundwater dating and subsurface processes investigated by noble gas isotopes (37Ar, 39Ar, 85Kr, 222Rn and 4He). In: Water Rock Interaction, WRI-10, Vol. 2(ed. by R. Cidu), 1569-1573, Villasimus, Italy.

Purtschert R., Aeschbach-Hertig W., Beyerle U., Kipfer R. and Loosli H.H. (2001b) Palaeowaters from the Glatt Valley, Switzerland. In: Palaeowaters in coastal Europe: Evolution of groundwater since the late

20

Pleistocene (eds. W.M. Edmunds and C.J. Milne). Geological Society Special Publication. Geological Society, London, pp. 155-162.

Rozanski, K., 1985. Deuterium and oxygen-18 in European groundwaters - links to atmospheric circulation in the past. Chem. Geol. (Isotope Geoscience Section), 52: 349-363.

Scheffel C., Blinov A., Massonet S., Sachsenhauser H., Stan-Sion C., Beer J., Synal H.A., Kubik P.W. and Nolte E. (1999). 36Cl in modern atmospheric precipitation. Geophys. Res. Lett. 26, No.10, 1401-1404.

Schlosser P. (1989). Tritium/3He dating of waters in natural systems. In: Isotopes of noble gases as tracers in environmental studies, IAEA, Vienna, 123-145.

Solomon D.K. and Cook P.G. (1999). 3H and 3He. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 397-424.

Stute M. and Schlosser P. (1999). Atmospheric noble gases. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 349-377.

Stute M. and Schlosser P. (2000). Tritium/3He measurements as calibration tools in groundwater transport modelling. In: Tracers and Modelling in Hydrogeology, Proceedings of the TraM´2000 Conference, Belgium. IAHS 262, 33-38.

Synal H.-A., Beer J., Bonani G., Suter M., Wölfli W. (1990). Atmospheric transport of bomb-produced 36Cl. Nucl. Instrum. Meth. Phys. Res. B 52, 483-488.

Troldborg L. (2005) The influence of conceptual geological models on the simulations of flow and transport in quaternary aquifer systems. PhD Thesis, GEUS and Environment & Resources, Technical University of Denmark in collaboration Geological Institute, Copenhagen University.

Weissmann G.S., Y. Zhang, Y., E.M. LaBolle & G.E. Fogg (2002) Dispersion of groundwater age in an alluvial aquifer system. Water Resources Research 38 (10):1198,

Zuber A. (1986). Mathematical models for the interpretation of environmental radioisotopes in groundwater systems. In Handbook of Environmental Isotope Geochemistry (Ed. Fritz P. and Fontes J.Ch.), V.2, The Terrestrial Environment B, pp. 1-59.

Zuber A. and Maloszewski P. (2001). Lumped-parameter models. In Environmental isotopes in the hydrological cycle, vol.6. Modelling. UNESCO/IAEA, Tech. Doc. Hydrol. 39, 5–35.

21

Chapter 2 36Cl in modern groundwater dated by a multi tracer approach (3H/3He, SF6, CFC-12 and 85Kr): A case study in Quaternary sand aquifers in the Odense Pilot River Basin, Denmark. J.A. Corcho Alvarado1 (*), R. Purtschert1, K. Hinsby2, L. Troldborg2, M. Hofer3, R. Kipfer3, W. Aeschbach-Hertig4, H. Arno-Synal5

1 Climate and Environmental Physics, Physic Institute, University of Bern, Switzerland 2 Geological Survey of Denmark and Greenland, GEUS, Copenhagen, Denmark 3 Water Resources and Drinking Water, EAWAG, Dübendorf, Switzerland 4 Institute of Environmental Physics, University of Heidelberg, Heidelberg, Germany 5 ETH Hönggerberg, Zürich, Switzerland Published in Applied Geochemistry 20 (2005) 599-609 Abstract

36Cl produced by thermonuclear bomb testing has been proposed as an additional tool to date or at least to identify recent groundwater components. In order to investigate the behaviour of 36Cl in shallow groundwater a multi-tracer approach (3H/3He, SF6, CFC-12 and 85Kr) was used to characterise and date the groundwater of a Quaternary Sands aquifer which is located on the Island of Funen near the city of Odense, Denmark. Recharge to the semi-confined shallow aquifer occurs through permeable sand windows and fractured tills at the surface. Locally, however, mixing with older pre-bomb water from the underlying limestone aquifer may occur. The integrated analyses of the available tracer data allowed a well constrained age structure determination of the investigated water system.

The 36Cl/Cl ratios measured in groundwater were used to reconstruct the fallout rates for radioactive 36Cl at Odense. The calculated fallout values exceeded the fallout estimated based on data from the Dye-3 ice core in Greenland. Recycling of the bomb peak fallout seems to be the most probable reason of the high values measured. The local extent of this process is difficult to quantify, which impedes the use of 36Cl for dating.

22

1. Introduction and aquifer characterization. The investigated area is located on the Island of Funen around the city of Odense,

Denmark. Geologically the site is situated in a complex setting of Quaternary glaciofluvial sand aquifers with confining sandy and clayey tills (figure 1). The semi-confined glaciofluvial sands which constitute the main aquifer on the island overlie a sequence of mainly Palaeocene marls and clays of varying thickness (typically 10-20 m) which form the lower boundary of the quaternary aquifer system. Recharge occurs through sand windows and lenses and through fractures and root holes in the tills. The potentiometric head is at depths between 4 and 10 m.b.s. Screened intervals (5-14 m) within the sands are at depths ranging from 18 to 56 m.b.s.. A Palaeocene Limestone aquifer is underlying the Odense shallow aquifer and the Palaeocene marls and clays (see the aquifer cross-section in figure 1) and in some areas deep wells may extract water from this aquifer or create hydraulic contact to the sand aquifers above.

Figure 1. a) Location of the investigated area. The area includes the southern part of the city of

Odense. b) Geological cross section with well locations (Geological Survey of Denmark and Greenland).

23

Denmark and the island of Funen (55o N) are located in a humid temperate zone with an

estimated average annual precipitation of about 780 mm. The estimated net recharge of the shallow semiconfined aquifer is 240 mm per year (Hinsby et al., 2003).

A set of modern residence time isotopic indicators (3H/3He, 85Kr, SF6 and CFC-12) was used to resolve the age structure of the shallow groundwater. In the past, several studies demonstrated the applicability of each of the adopted tracer methods (Schlosser, 1989; Busenberg and Plummer, 1992; Plummer et al., 1993; Cook and Solomon, 1997; Beyerle et al., 1999; Plummer and Busenberg, 1999; and Bauer et al., 2001). The benefit of the simultaneous application of several methods lies in the possible identification of processes that could erroneously affect the interpretation of data of a single method (for example CFC degradation, Plummer and Busenberg, 1999).

1940 1950 1960 1970 1980 1990 20000

100

200

300

400

500

600

36Cl

3H

CFC

-12

(ppt

V), 3 H

[TU

/3],

36C

l (10

5 ato

ms

cm-2 y

-1)

0

5

10

15

20

25

30

85Kr

CFC-12

SF6

85Kr [dpm cm

-3 /5], SF6 [pptv]

Figure 2. 3H concentration in precipitation at Odense (IAEA/GNIP database, 2001), 36Cl fallout

predicted at Dye-3 ice core, Greenland, by Synal et al. (1990) and atmospheric concentrations of 85Kr (IAR Freiburg), CFC-12 (Walker et al., 2000) and SF6 (Maiss et al., 1998) used as input functions in the area of Odense.

In this paper the 36Cl/Cl ratio and the Cl concentration in groundwater were analysed to

reconstruct a budget of 36Cl that was produced by testing thermonuclear weapons during 1952-58. The main objective of this part is to investigate the use of the 36Cl fallout for tracing and dating groundwater. High levels of 36Cl in groundwater indicate, such as tritium, that recharge has occurred since the bomb peak tests. Records of the 36Cl fallout were reconstructed by Bentley et al. (1986) and Synal et al. (1990) from the Dye-3 ice core, Greenland (see figure 2). The 36Cl isotope has a half-life of 301 000 years; therefore radioactive decay can be neglected for the time scale involved in this study. According to the reconstructed fallout at Dye-3, 36Cl was washed out from the atmosphere since the bomb tests and the actual fallout rates reach nearly the natural levels. However, according to recent studies (which included direct measurement of the 36Cl fallout at different sites, and the 36Cl concentration in different environmental components such as plants, soil, surface waters, groundwaters, etc.), storage and recycling of Cl in the biosphere seems to maintain a

24

background 36Cl activity chronically elevated above natural levels (Milton et al., 1994; Cornett et al., 1997; Milton et al., 1997; Scheffel et al., 1999; Blinov et al., 2000; and Milton et al., 2003). The present study contributes new data to a better understanding of the behaviour of this isotope in nature. 2. Methods

Three wells (34h, 74e and Od J) were sampled in the first half of 2001 for the analysis of 3H/3He, 85Kr, SF6, CFC-12 and 36Cl/Cl ratio (figure 1). Two depths were sampled from Od J, one mixed sample from the complete screened section, and one sample from the bottom of the screen. Field measurements of dissolved O2 concentration, water temperature, pH and Eh were also performed.

Water samples for noble gases, 3H and CFC analysis were immediately transferred to 45ml copper tubes and sealed with pinch-off clamps. Water for SF6 analysis was sampled in 500ml gas sampling cylinders. The analyses were performed at the EAWAG and ETH laboratories (Switzerland) following the procedures described by Hofer and Imboden (1998), Beyerle et al. (2000) and Hofer et al. (2002).

For 85Kr analysis several hundreds litres of water were degassed in the field and the extracted gas was compressed into evacuated cylinders. Krypton was separated by gas chromatography and the 85Kr activities were measured by low level gas proportional counting at the Physics Institute, University of Bern, Switzerland (Loosli et al., 1986 and Loosli et al., 1999).

The analyses of 36Cl/Cl ratios were performed at ETH Zurich according to the method described by Synal et al. (1990). Chloride analyses in rain water were performed by the Geological Survey of Denmark and Greenland on samples collected by the Odense Water Company (1963-1995). Precipitation rate at Odense was taken from the IAEA/WMO (GNIP database accessible at http://www.iaea.org/). 3. Results and Discussion 3.1. Measurements

Groundwater temperatures range between 9 and 10.3 oC, and pH between 6.8 and 7.2. The excess air, expressed as the Ne excess, ranges between 17 and 25%. This parameter was used to correct the SF6, CFC-12 and 3He concentrations for concentrations above gas solubility limits. The aquifer has generally anoxic conditions (table 1), (O2: < 0.1 mg/l; Eh: < -33 mV). Nitrate concentrations vary, but are generally low or very low (< 1 mg/l) indicating on-going or recent nitrate reduction. Sulphate reduction is generally not observed (Hinsby et al., 2003).

Table 1. Field measurements and noble gas data. Noble gas concentrations expressed in cm3 STP per g of water.

Well Screen depth [m.b.s.]

T [oC]

Eh [mV]

O2 [mg/l]

pH

4He [10-8]

3He/4He [10-6]

Ne [10-7]

Ne excess[%]

34h 18-23 9.0 -33 <0.1 6.8 6.44 ± 0.08 2.24 ± 0.02 2.53 ± 0.04 25 74e 26-39 10.3 -78 <0.1 7.0 8.42 ± 0.10 2.13 ± 0.02 2.45 ± 0.04 22 Od J 46-56 9.2 -56 <0.1 7.2 6.54 ± 0.08 2.03 ± 0.02 2.51 ± 0.04 24

Od J deep 46-56 9.2 -78 <0.2 7.1 6.44 ± 0.08 2.02 ± 0.02 2.37 ± 0.04 17

25

Table 2. Tracer concentrations in groundwater samples. Well 85Kr

[dpm/cm3 Kr] 3H

[TU] 3HeTritiogenic

[TU] SF6

[pptV]CFC-12 [pptV]

Cl [ppm]

36Cl/Cl [10-12]

36Cl [107 atoms/l]

34h 18.5 ± 5.8 10.8 ± 0.2 24.4 ± 1.1 1.36 791 (c) 66.7 0.09 10.3 ± 0.6 74e 5.9 ± 1.1 11.0 ± 0.3 39.4 ± 1.2 <DL 98 54.4 0.15 13.5 ± 0.8 Od J 4.5 ± 0.7 10.2 ± 0.2 20.1 ± 1.0 0.25 80 27.4 0.17 8.0 ± 0.3 Od J deep 4.8 ± 1.3 10.2 ± 0.3 21.4 ± 1.0 0.10 13 27.3 0.18 8.3 ± 0.9

85Kr, 3H and CFC-12 above the detection limit (DL) were observed in all the samples, in

accordance with the geological situation which points to modern recharge (table 2). A CFC-12 concentration in excess of air equilibrated water was observed in one sample and attributed to contamination (c). The concentration of SF6 in one sample was below DL.

The 36Cl/Cl ratios and the Cl concentrations in groundwater were measured at all of the wells and the results are presented in table 2. The lowest chloride concentrations were observed in the deepest well, Od J.

3.2. Tracer dating

Measured concentrations were converted into corresponding atmospheric input concentrations based on the in situ water temperature, elevation of the recharge area and the excess air content (Schlosser, 1989; Busenberg and Plummer, 1992; Plummer et al., 1993; Aeschbach-Hertig et al., 1999; Busenberg and Plummer, 2000; Holocher et al., 2002). No corrections are necessary for 85Kr, if no significant isotopic fractionation occurs during gas/water partitioning in the quasi-saturated zone of the aquifer. Noble gas data support this assumption as the measured 22Ne/20Ne and 40Ar/36Ar ratios agree with the atmospheric ratios and hence exclude any significant isotope fractionation. As a first approximation, the tracer data were interpreted assuming piston flow (PF) and neglecting hydrodynamic dispersion and mixing of different water components. Using the local input functions (figure 2) residence times between 14 and 48 years were calculated (figure 3) (Corcho et al., 2002). Low SF6 concentrations close to the detection limit allowed only the estimation of a minimum age of about 30 years for wells 74e and Od J. The comparisons of tracer ages depict two obvious discrepancies: CFC-12 ages are generally older compared to the other tracers and 3H/3He ages for well Od J are younger. These considerable discrepancies between the apparent tracer ages indicate that additional processes have to be taken into account which are not included in the simple Piston-Flow model. The most probable ones are presented in figure 3 and include mixing, dispersion and tracer degradation.

CFC-12 is expected to behave non-conservative under reducing conditions. It is known from similar studies that CFCs can be degraded in such conditions (Busenberg and Plummer, 1992; Plummer and Busenberg, 1999), and degradation has been found in other reducing sand aquifers in Denmark (Hinsby et al., 1997; Hinsby et al., 2004). As sorption is estimated to be insignificant for the investigated type of aquifer sediments (Engesgaard et al., 2004; Hinsby et al., 2004), degradation stays the most probable explanation for the CFC deviations. Because of the increasing input function this results in an overestimation of residence times.

3H/3He ages correspond to the residence time of the original tritium-bearing water component while the other tracers indicate a mean age which is the combined result of the age of the young water component and dilution with an older tracer-free water component. Mixing of different water components, either by dispersion or by different origins, is therefore potentially the reason for the deviation of the 3H/3He ages at well Od J (Kipfer et al., 2002).

26

Figure 3. SF6, CFC-12 and 3H/3He PF ages vs. 85Kr PF age at three wells of the Odense Sands aquifer

(2 screen intervals at Od J). Arrows indicate direction of age correction if the simple piston flow model is complemented with additional processes. Deviations from the 1:1 correlation can mainly originate from degradation for CFC12, dispersion and admixture of an old water component for 3H/3He.

In the following exercise, the data were interpreted with a simplified model that

analytically accounts for dispersion (described by the parameter Pe, which defines the relative importance of advective and dispersive flow), the ratio of mixed older (tracer free) water component (m) and the mean residence time (τ) of the young component (Zuber and Maloszewski, 2001). The three model parameters pi (τ, Pe and m) are determined by inverse modelling (Purtschert et al., 1999), minimizing the sum of weighted squared deviations,

( )( )∑

−=

n

i i

imeasj

iout CpC

2

22

σχ (1)

where C is the tracer concentration (meas: measured, out: modelled), i the 85Kr, 3H, 3He or SF6, σI the experimental 1σ-errors

The weights σ2i give preference to the most precise data. CFC-12 was excluded from the

fitting procedure considering that an additional process (degradation) is modifying the concentration in groundwater which is not included in our dispersion model. In all cases, the very low values of χ2 (see table 3) are indicating a good fitting.

The resulting mean residence times of the young component (τ) range from 17 to 27 years. In the deepest well Od J, an admixture of between 30 and 40% of old water was concluded. This is in perfect agreement with the fact that the screen of the well Od J is located in the deepest part of the aquifer (table 1, figure 1) where a hydraulic contact with the underlying limestone aquifer can be expected.

27

1960 1970 1980 1990 20000

400

800

1200

1600

2000

~ 20-25 years

~ 10 years

well 74e

well 34h

3H bomb peak in precipitation

3 H fa

llout

(TU

)

0

30

60

90

120

150

3H

at w

ells

(TU

)

Figure 4. 3H time series for Odense precipitation and for wells 34h (○) and 74e (∆).

Dispersion seems to be of minor importance for the analytical solution (table 3). The

values of the Pe number indicate that advection transport predominates over dispersion. In the shallower part of the aquifer (34h) dispersion is more pronounced, than in the deeper part (74e).

Assuming that degradation is the only process affecting the CFC-12 concentration the degradation rate of CFC-12 at the Odense site can be estimated to be in the range of 0.8 to 3.6 10-4 day-1. This is agreement with the CFC-12 degradation rate of 0 – 6 10-4 day-1 (best fit 3 10-4 day-1), corresponding to a half life of about 6 years, which was found in a similar Danish aquifer (Hinsby et al., 2004). Table 3. Parameters that best fit the dispersion model to the measurements. τ- Mean residence time of the young

component [years], m - Mixing ratio of the young component [%] and Pe- Peclet number (vx/D). The dimensionless χ2-value indicates the goodness of the fit as the weighted sum of deviations between measured and modelled concentrations (see equation 1)

Dispersion model Well τ m Pe χ2

34h 17 100 23 0.4 74e 27 100 5500 0.0 Od J 23 66 9562 2.7 Od J deep 22 70 6431 1.4

In the above described multi tracer analysis the aquifer is treated as a steady state system.

The resulting age structure is representative for the most recent, spatially averaged groundwater residence time. A several years record of tritium in precipitation and in two wells (34h and 74e) offers a possibility to reconstruct residence times in the past (figure 4). The time lag between the bomb peak in precipitation and in groundwater indicates residence times of approximately 10 and 20-25 years at the wells 34h and 74e, respectively. Within the uncertainties these results indicate faster travel times to the wells (about 2-7 years) than those obtained from the other isotopic tracer dating, but are still in relative good agreement. For

28

well 34h, where the highest disagreement was obtained, groundwater dating was performed more than 10 years after the well was shut down and this may be part of the explanation as the fraction of young groundwater may change during pumping. The relative high dispersion impact estimated by the model might be explained by the non-stationarity not included in the analytical expression. The screen of this well is positioned in the upper shallow part of the aquifer system, where seasonal variation in recharge is influencing the flow system and consequently may affect the groundwater age distribution. 3.3. Origin of 36Cl in young groundwaters.

36Cl has different origins in the environment (figure 5). In groundwater two sources can be distinguished: i) 36Cl in precipitation and ii) 36Cl added in the subsurface. i) 36Cl concentrations in recharge can be calculated from the fallout rate FCl (atoms/m2 s), the

annual precipitation P (mm) and the evapotranspiration E (%) according to (Bentley et al., 1986):

[ ] ⎟⎠⎞

⎜⎝⎛

−⋅

⋅=

EPFCl Cl

Nat 1001003153600036 (2)

The fallout rate is the sum of a natural fallout (Fnat), due to interaction of cosmic rays with

40Ar, 36Ar and 35Cl in the atmosphere ,and 36Cl that was produced by atmospheric nuclear weapon testing (Fbomb) in the years 1952-1958 (Bentley et al., 1986). A globally averaged Fnat of 48 atoms m-2 s-1 was calculated by Phillips (1999), a value that is up to four times higher than other estimates (Lal and Peters, 1967; Bentley et al., 1986; Masarik and Beer, 1999). The Fbomb was reconstructed from ice core analyses (figure 2) and reached values up to 104 atoms m-2 s-1 (Bentley et al., 1986; Synal et al., 1990). Hence, in young groundwater bomb 36Cl completely overwhelms the natural fallout by several orders of magnitude.

Chloride concentrations in rainwater in the area range between 1-73 ppm with a mean of 15 ppm. An evaporation rate of 70% will increase the mean input concentration to about 50 ppm, which is comparable to the measured concentrations found in the investigated groundwater (table 2). However, subsurface chloride sources with high 36Cl/Cl ratios may significantly contribute to the total 36Cl, even if only small amount of Cl is added to the water. ii) After recharge, 36Cl in groundwater may originate from the following sources

• 36Cl produced cosmogenically in shallow depths through interaction of secondary particles of cosmic radiation in both the rock and the water phases. The most important reactions are spallation of K and Ca and neutron activation of 35Cl. The in-situ production in the water phase depends on the Cl content and the depth below surface. Resulting 36Cl/Cl ratios are commonly negligible considering the small exposure times of the water in the environment (see table 4). Additionally, 36Cl released from rocks into the water due to weathering produces a 36Cl flux which varies in different geological environments and for different altitudes a.s.l. (Phillips, 1999). Typical values at latitude 40oN are below 25 atoms m-2 s-1 and therefore comparable to the natural atmospheric deposition fluxes.

• At greater depths, 36Cl is produced by neutron irradiation of 35Cl either in the rock matrix or in the chloride dissolved in the groundwater. The subsurface n-flux is mainly controlled by the U and Th concentrations in the rock. Calculated and measured

29

equilibrium 36Cl/Cl ratios commonly range from 5-30 10-15 in most rocks (Bentley et al., 1986) and below 1 10-15 in halites (Fabryka-Martin et al., 1987).

• Other possible 36Cl sources could be the addition of chlorinated solvents such as TCE and BTEX, which were observed in some of the groundwater at Odense; and road salting, which can be significant locally because a large proportion of the catchment lies within the city limits. Most solvents are manufactured with chloride derived from brines which were made from evaporite deposits (sea water, halites). Equilibrium 36Cl/Cl ratios for sea water are below 4 10-15 (Finkel et al., 1980) and for halite formations below 1 10-15 (Fabryka-Martin et al., 1987). Salts used for treating roads during winter time have the same origin from evaporite deposits; therefore we could expect similar 36Cl/Cl ratios.

Spontaneous fissionU (n)Th Al, Mg..(α,n)

α- decay

40Ar(p, nα)36Cl

40Ca(p, sp)36Cl39K(p, sp)36Cl40Ca(μ-, α)36Cl39K(μ-, p2n)36Cl40Ca(n, pα)36Cl39K(n, α)36Cl35Cl(n, γ)36Cl

Surface production by cosmic ray particles and

neutrons from natural decay series

Soil surface

Atmospheric production

by cosmic rays

Ocean

35Cl(n, γ)36Cl

Cosmic raysCosmic rays

Marine aerosols (36Cl)

Atmospheric36Cl

Atmospheric fallout 36Cl

(wet and dry deposition)

Groundwater36Cl

Bomb produced - 36ClOcean: 35Cl(n, γ)36Cl

Atmosphere: 40Ar(n, nα)36Cl

40Ca(n, pα)36Cl39K(n, α)36Cl35Cl(n, γ)36Cl

36Cl from the nuclear industry

Spontaneous fissionU (n)Th Al, Mg..(α,n)

α- decay

40Ar(p, nα)36Cl

40Ca(p, sp)36Cl39K(p, sp)36Cl40Ca(μ-, α)36Cl39K(μ-, p2n)36Cl40Ca(n, pα)36Cl39K(n, α)36Cl35Cl(n, γ)36Cl

Surface production by cosmic ray particles and

neutrons from natural decay series

Soil surface

Atmospheric production

by cosmic rays

Ocean

35Cl(n, γ)36Cl

Cosmic raysCosmic rays

Marine aerosols (36Cl)

Atmospheric36Cl

Atmospheric fallout 36Cl

(wet and dry deposition)

Groundwater36Cl

Bomb produced - 36ClOcean: 35Cl(n, γ)36Cl

Atmosphere: 40Ar(n, nα)36Cl

40Ca(n, pα)36Cl39K(n, α)36Cl35Cl(n, γ)36Cl

36Cl from the nuclear industry

Figure 5. A schematic representation of the main sources of 36Cl in the environment and the input to

groundwater. The contributions of the different 36Cl sources are summarized in table 4. Subsurface

sources are relatively low compared to precipitation sources. 36Cl fluxes (equation 4) and 36Cl/Cl ratios were converted into 36Cl concentrations in water using precipitation and evaporation rates and Cl concentrations, respectively, which are valid in the area of investigation. Bomb 36Cl was estimated using the fallout rate depicted in figure 6 (Synal et al., 1990) and the corresponding residence times from the tracer data. Bomb derived 36Cl is clearly the dominant 36Cl source in the Odense groundwater. However the observed 36Cl concentration found in groundwater cannot be explained with the above mentioned sources alone. In figure 6, reconstructed 36Cl fallout fluxes (through equation 4) at Odense are compared with fallout rates obtained from Dye-3 ice core (Greenland). Also shown are direct measurements of the 36Cl fallout rates in annual atmospheric precipitation samples (wet and dry deposition) at different sites in Europe (Scheffel et. al., 1999). There is a general disagreement between the ice core data and data from groundwater or from annual

30

precipitation samples. On the other hand, calculated fluxes at Odense agree within the range of variation with direct measurements in Europe. Our measurements also reproduce the decrease of the fallout rate as function of time but at a smaller rate than concluded from the Dye-3 ice core.

1940 1950 1960 1970 1980 1990 2000101

102

103

104

105

101

102

103

104

105

Natural atmospheric fallout(Phillips, 1999)

Representation of the recycling of 36Cl

Reconstructed fallout at Odense (Denmark) Fallout at Dye-3 ice core (Greenland) Fallout with a latitudinal correction factor of 2.5 with respect to Dye-3 Direct fallout measurements in Europe

Od Jdeep

Od J

34h74e

36C

l (10

3 ato

ms/

cm2 /y

)

Figure 6. 36Cl fallout predicted at Dye-3 ice core (Synal et al., 1990) and fallout calculated after

considering a latitudinal correction factor of 2.5. (●) fallout rates estimated at Odense from groundwater measurements (the error bars of the 36Cl fallout based on evaporation rates from 60-80 percent) and (■) direct measurements of fallout rates at different sites in Europe (Scheffel et al., 1999). A schematic fallout rate considering recycling is shown by the dashed line. The dotted line represents the level of the natural atmospheric fallout (Phillips, 1999).

Recycling of the 36Cl bomb peak fallout (Milton et al., 1994; Milton et al., 1997; Cornett et al., 1997; Scheffel et al., 1999; Blinov et al., 2000; and Milton et al., 2003) might be the explanation for the extra 36Cl measured in groundwater. Thereby, a part of the bomb 36Cl is stored in the biosphere and reemitted into the troposphere in the form of CH3Cl. This process might possibly reintroduce sufficient amounts of 36Cl in the atmosphere, and would be responsible for a large fraction of the present day atmospheric flux at Odense.

31

Table 4. Summary of sources of 36Cl in groundwater at Odense and estimations of their magnitudes. 36Cl flux

[atoms/m2/s]

36Cl/Cl ratio[10-15]

[Cl] [ppm]

[36Cl] [106 atoms/l]

Present in precipitation Natural atmosphere-produced fallout 48 a 15 b <7 c Bomb produced fallout 22-428 d 15 b 3-58 c

Subsurface sources In situ produced in shallow groundwater < 1 e 27-67 < 1 Weathering from shallow rocks <25 a <40 f <3 g Dissolution of Cl bearing minerals (e.g. halite) 5 h <40 f <4 Addition of chlorinated solvents 1-4 h <40 f <3 Road salting 1-4 h <40 f <3

Calculated input to groundwater (mean value) i 41 Measured values (mean value) j 120 61 123

Input from other sources k 82 a) Phillips (1999). b) Mean value of chloride concentration in precipitation. Data from the Geological Survey of Denmark and Greenland. c) Calculated using equation (2) from Bentley et al. (1986). A precipitation rate of 780 mm and an evaporation of 70 % were used (Hinsby et al., 2003). d) Calculated from the Dye-3 ice core data, Greenland (Synal et al., 1990) considering a latitudinal correction factor of 2.5 (Bentley et al., 1986). The range originates from the different recharge years estimated with the box model approach (between 1974 and 1984). e) Calculated using an averaged neutron flux at shallow depths (between soil surface and 20 m.b.s.) taken from Fabryka-Martin (1988). f) Assuming an input of chloride to groundwater of 40 ppm which is most probably an overestimation. g) Calculated using a recharge rate of 280 mm/y (Hinsby et al., 2003). h) Data from Bentley et al. (1986), Fabryka-Martin et al. (1987) and Finkel et al. (1980) i) Sum of the mean value of each subsurface and precipitation source. J) Value corresponds to the mean of the youngest (34h) and the oldest (74e) groundwaters investigated at Odense. k) Difference between “Measured values” and “Calculated input to groundwater”.

4. Conclusions

The combined application of a set of different residence time indicators for young groundwater allowed a consistent dating of 4 samples from 3 boreholes in the Odense aquifer. Considering the mixture of an older, tracer free, water component and dispersive mixing, it was possible to consistently interpret 3H/3He, 85Kr, and SF6 concentrations in terms of groundwater residence time. Groundwater residence times inferred from analytical data are slightly younger than those implied by the delay in arrival of 3H peaks for two wells. For CFC, neither mixing nor dispersion could provide a satisfactory explanation. Under reducing conditions, groundwater residence times obtained from CFC measurements tend to be too high due to degradation. Because it is difficult to quantify such process, CFCs were excluded in the age estimations. Assuming degradation as the only responsible process for the discrepancy, a half life of between 5 and 24 years for CFC-12 is estimated under the prevailing conditions in the investigated area.

A 36Cl budget of the tracer dated groundwater reveals some additional evidence that 36Cl produced during atmospheric detonations of nuclear weapons is recycled in the environment. From groundwater samples, reconstructed 36Cl fallout rates exceed estimations from ice core measurements by almost a factor of two. As a consequence, quantitative groundwater dating using thermonuclear 36Cl is complicated lacking a well constrained local input function. However, because bomb 36Cl exceeds natural background levels by several orders of

32

magnitude it can be used, similar to 3H in the last several decades, as an indicator for the presence of recent water components.

Acknowledgement

This study was supported by the EU project BASELINE “Natural Baseline Quality of European Groundwaters: A Basis for Aquifer Management” and the IAEA coordinated research program “Isotope response to the dynamic changes in groundwater systems due to long-term exploitation”. The Odense Water Company is acknowledged for access to wells and their water sample archive.

We are grateful to PSI and ETH who jointly operate the Zurich AMS facility. References Aeschbach-Hertig, W., Peeters, F., Beyerle, U., Kipfer, R., 1999. Interpretation of dissolved atmospheric noble

gases in natural waters. Water Resour. Res. 35, 2779-2792. Bauer, S., Fulda, C., Wolfgang, S., 2001. A multi-tracer study in a shallow aquifer using age dating tracers 3H,

85Kr, CFC-113 and SF6 – indication for retarded transport of CFC-113. J. Hydrol. 248, 14-34. Bentley, H.W., Phillips, F.M., Davis, S.N., 1986. Chlorine-36 in the terrestrial environment. In: Fritz P. and

Fontes J.Ch. (eds.), Handbook of Environmental Isotope Geochemistry, vol. 2, The Terrestrial Environment B. Elsevier Science Publishers, 427-480.

Beyerle, U., Aeschbach-Hertig, W., Hofer, M., Imboden, D.M., Baur, H., Kipfer, R., 1999. Infiltration of river water to a shallow aquifer investigated with 3H/3He, noble gases and CFCs. J. Hydrol. 220, 169-185.

Beyerle, U., Aeschbach-Hertig, W., Imboden, D.M., Baur, H., Graf, T., Kipfer, R., 2000. A mass spectrometric system for the analysis of noble gases and tritium from water samples. Environ. Sci. Technol. 34, 2042-2050.

Blinov, A., Massonet, S., Sachsenhauser, H., Stan-Sion, C., Lazarev, V., Beer, J., Synal, H.-A.., Kaba, M., Masarik, J., Nolte, E., 2000. An excess of 36Cl in modern atmospheric precipitation. Nucl. Instrum. Meth. Phys. Res. B 172, 537-544.

Busenberg, E., Plummer, L.N., 1992. Use of chlorofluorocarbons (CCl3F and CCl2F2) as hydrologic tracers and age-dating tools: the alluvium and terrace system of Central Oklahoma. Water Resour. Res. 28, 2257-2283.

Busenberg, E., Plummer, L.N., 2000. Dating young groundwater with sulfur hexafluoride: natural and anthropogenic sources of sulfur hexafluoride. Water Resour. Res. 36, 3011-3030.

Cook, P.G., Solomon, D.K., 1997. Recent advances in dating young groundwater: chlorofluorocarbons, 3H/3He and 85Kr. J. Hydrol. 191, 245-265.

Corcho Alvarado, J.A., Purtschert, R., Hofer, M., Aeschbach-Hertig, W., Kipfer, R., Troldborg, L., Hinsby, K., 2002. Comparison of residence time indicators (3H/3He, SF6, CFC-12 and 85Kr) in shallow groundwater: a case study in the Odense aquifer, Denmark. Goldschmidt conference, Davos. Geochim. Cosmochim. Acta 66: A152.

Cornett, R.J., Andrews, H.R., Chant, L.A., Davies, W.G., Greiner, B.F., Imahori, Y., Koslowsky, V.T., Kotzer, T., Milton, J.C.D., Milton, G.M., 1997. Is 36Cl from weapons’ test fallout still cycling in the atmosphere?. Nucl. Instrum. Meth. Phys. Res. B 123, 378-381.

Engesgaard, P.; Højberg, A.L.; Hinsby, K., Laier, T., Jensen, K.H, Larsen, F, Busenberg, E., Plummer, L.N., 2004. Transport of CFCs in the unsaturated zone, Rabis Creek, Denmark. Accepted for publication in Vadose Zone Journal.

Fabryka-Martin, J.T., Davis, S.N., Elmore, D., 1987. Applications of 129I and 36Cl to hydrology. Nucl. Instrum. Meth. Phys. Res. B 29, 361-371

Fabryka-Martin, J.T., 1988. Production of Radionuclides in the Earth and their Hydrogeologic Significance, with Emphasis on Chlorine-36 and Iodine-129. Unpublished PhD thesis, Department of Hydrology and Water Resources, University of Arizona, Tucson, USA.

Finkel, R.C., Nishiizumi, K., Elmore, D., Ferrazo, R.D., Gove, H.E., 1980. 36Cl in polar ice, rainwater and seawater. Geophys. Res. Lett. 7, 11, 983-986.

Geological Survey of Denmark and Greenland, Denmark. Hinsby, K.; Thomsen, A.; Engesgaard, P.; Larsen, F.; Jensen, K.H.; Laier, T.; Busenberg, E., Plummer, L.N.,

2004. CFC dating, and transport and degradation of CFCs in a partly anaerobic sand aquifer, Rabis Creek, Denmark, in prep.

33

Hinsby, K., Troldborg, L., Purtschert, R., Corcho Alvarado, J.A., 2003. Integrated transient hydrological modelling of tracer transport and long-term groundwater/surface water interaction using four 30 year 3H time series and groundwater dating for evaluation of groundwater flow dynamics and hydrochemical trends in groundwater and surface water. Report to the IAEA (TECDOC in preparation by the IAEA).

Hinsby, K., Laier, T., Thomsen, A., Engesgaard, P., Jensen, K.H., Larsen, F., Jakobsen, R., Busenberg, E., Plummer, L.N., 1997. CFC-dating, and transport and degradation of CFC-gases in different redox environments: two case studies from Denmark. Abstracts with programs, Geological Society of America, Annual Meeting, Salt Lake City, Utah, USA.

Hofer, M., Imboden, D.M., 1998. Simultaneous determination of CFC-11, CFC-12, N2 and Ar in water. Anal. Chem. 70, 4, 724-729.

Hofer, M., Peeters, F., Aeschbach-Hertig, W., Brennwald, M., Holocher, J., Livingstone, D.M., Romanovski, V., Kipfer R., 2002. Rapid deep-water renewal in Lake Issyk-Kul (Kyrgyzstan) indicated by transient tracers. Limnol. Oceanogr. 47, 1210-1216.

Holocher, J., Peeters, F., Aeschbach-Hertig, W., Hofer, M., Brennwald, M., Kinzelbach, W., Kipfer, R., 2002. Experimental investigations on the formation of excess air in quasi-saturated porous media. Geochim. Cosmochim. Acta 66, 4103-4117.

International Atomic Energy Agency/World Meteorological Organization (2001). The Global Netwrok of Isotopes in Precipitation Database. Accessible at: http://isohis.iaea.org.

Institute of Atmospheric Radioactivity (IAR), Freiburg (Germany) Kipfer, R., Aeschbach-Hertig, W., Peeters, F., Stute, M., 2002. Noble gases in Lakes and Ground Waters. In

Noble gases in geochemistry and Cosmochemistry; ed. Porcelli D., Ballentine C.J. and Wieler R.; Reviews in Mineralogy and Geochemistry, V. 47, 615-700.

Lal, D., Peters, B., 1967. Cosmic ray produced radioactivity on the Earth, In: Handbuch der Physik, v.46/2, Springer-Verlag, Berlin-Heidelberg-New York, 551-612.

Loosli, H.H.; Möll, M.; Oeschger, H., Schotterer, U., 1986. Ten years low-level counting in the underground laboratory in Bern, Switzerland. Nuclear Instruments and Methods in Physics Research Section B 17, 402-405.

Loosli, H.H., Lehmann, B.E., Smethie, W.M., 1999. Noble gas radioisotopes: 37Ar, 85Kr, 39Ar, 81 Kr. In : Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 379-396.

Maiss, M., Brenninkmeijer, C. A. M., 1998. Atmospheric SF6: Trends, sources, and prospects. Environmental Science & Technology 32, 3077-3086.

Masarik, J., Beer, J., 1999. Simulation of particle fluxes and cosmogenic nuclide production in the earth's atmosphere. J. Geophys. Res. D104 (10), 12,099-13,012.

Milton, J.C.D., Andrews, H.R., Chant, L.A., Cornett, R.J.J, Davies, W.G., Greiner, B.F., Imahori, Y., Koslowsky, V.T., McKay, J.W., Milton, G.M., 1994. 36Cl in the Laurentian Great Lakes basin. Nucl. Instrum. Meth. Phys. Res. B 92, 440-444.

Milton, J.C.D., Milton, G.M., Andrews, H.R., Chant, L.A., Cornett, R.J.J, Davies, W.G., Greiner, B.F., Imahori, Y., Koslowsky, V.T., Kotzer, T., Kramer, S.J., McKay, J.W., 1997. A new interpretation of the distribution of bomb-produced chlorine-36 in the environment, with special reference to the Laurentian Lakes. Nucl. Instrum. Meth. Phys. Res. B 123, 382-386.

Milton, G.M., Milton, J.C.D., Schiff, S., Cook, P., Kotzer, T.G., Cecil, L.D., 2003. Evidence for chlorine recycling – hydrosphere, biosphere, atmosphere – in a forested wet zone on the Canadian Shield. Appl. Geochem. 18, 1027-1042.

Phillips, F.M., 1999. Chlorine-36. In Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 299-348.

Plummer, L.N., Busenberg, E., 1999. Chlorofluorocarbons. In: Cook, P.G. and Herczeg, A.L. (eds.), Environmental tracers in subsurface hydrology. Kluwer Academic Publishers, Boston, 441-478

Plummer, L.N., Michel, R.L., Thurman, E.M., Glynn, P.D., 1993. Environmental Tracers for Age Dating Young Ground Water. In Alley, W.M. (Ed.), Regional Ground-Water Quality. Van Nostrand Reinhold, New York, 255-294.

Purtschert, R., Loosli, H.H., Beyerle, U., Aeschbach-Hertig, W., Imboden, D., Kipfer, R., Wieler, R., 1999. Dating of young water components by combined applications of 3H/3He and 85Kr measurements. In: International Symposium on Isotope Techniques in Water Resources Development and Management, Vienna, Austria, 59-60

Scheffel, C., Blinov, A., Massonet, S., Sachsenhauser, H., Stan-Sion, C., Beer, J., Synal, H.A., Kubik, P.W., Nolte, E., 1999. 36Cl in modern atmospheric precipitation. Geophys. Res. Lett. 26, No.10, 1401-1404.

Schlosser, P., 1989. Tritium/3He dating of waters in natural systems. In: Isotopes of noble gases as tracers in environmental studies, IAEA, Vienna, 123-145.

34

Synal, H.-A., Beer, J., Bonani, G., Suter, M., Wölfli, W., 1990. Atmospheric transport of bomb-produced 36Cl. Nucl. Instrum. Meth. Phys. Res. B 52, 483-488.

Walker, S.J., Weiss, R.F., Salameh, P.K., 2000. Reconstructed histories of the annual mean atmospheric mole fractions for the halocarbons CFC-11, CFC-12, CFC-113 and carbon tetrachloride. J. Geophys. Res., C, Oceans, 105, 285-290.

Zuber, A., Maloszewski, P., 2001. Lumped-parameter models. Environmental Isotopes in the Hydrological Cycle. Vol. 6: Modelling, UNESCO/IAEA, Tech. Doc. in Hydrology 39, 5-35.

35

Chapter 3 Constraining groundwater age distribution using 39Ar: a multiple environmental tracer (3H/3He, 85Kr, 39Ar and 14C) study in the semi-confined Fontainebleau Sands aquifer (France). J.A. Corcho Alvarado1, R. Purtschert1, F. Barbecot2, C. Chabault2, J. Rueedi3, V. Schneider2, W. Aeschbach-Hertig4, R. Kipfer5,6, H.H. Loosli1

1 Climate and Environmental Physics Division, Phys. Institute, Univ. Bern, Bern, Switzerland 2 Laboratoire d'Hydrologie et de Géochimie Isotopique, Université Paris-Sud, Orsay, France 3 Robens Centre for Public and Environmental Health, University of Surrey, United Kingdom 4 Institute of Environmental Physics, University of Heidelberg, Heidelberg, Germany 5 Water Resources and Drinking Water, EAWAG, Dübendorf, Switzerland 6 Isotope Geology, ETHZ, Zürich, Switzerland Submitted to Journal of Hydrology, October 2005

Abstract

A multiple tracer approach is used to investigate the age structure of groundwater in the semiconfined Fontainebleau Sands aquifer that is located in the shallower part of the Paris Basin (France). The hydrogeological situation which is characterized by spatially extended recharge, large screen intervals and possible leakage from deeper aquifers lead to expect a wide range of residence times and pronounced mixing of different water components. Consequently, a large set of tracers (3H/3He, 85Kr, 39Ar and 14C) with corresponding dating ranges was adopted. We examine the use of 39Ar, a noble gas radioisotope with a half-life of 269 years, to constrain the age distribution of groundwater in an intermediate range (< 1000 years).

By inverse modelling the recharge rate and depth of water and the groundwater age structure are estimated. The obtained recharge rates of 100-200 mm/a are comparable to estimations using hydrograph data. Best agreement between the modelled and measured tracer concentrations was achieved if tracer transport through a 30-40 meter thick unsaturated soil zone is assumed. This result fits well with the hydrogeological boundary conditions in the area. Transport times of the water and gas from the soil surface to the deep water table range between 10-40 and 1-6 year, respectively. Reconstructed concentrations of 85Kr and 3H at the water table were used for saturated flow modelling. Consistently with the spatially extended recharge and large screened intervals, provides the exponential model with mean residence times between 2-28 years good fits of 3H, 3He and 85Kr. Both 39Ar measurements as well as the box model approach indicate the presence of older waters (3H and 85Kr free). The mean residence times of the old water components are in the order of about 100-600 years. Finally, a simple conceptual model of the aquifer is constructed based on the results of the tracer methods.

36

1. INTRODUCTION Environmental tracer methods are nowadays routine tools for obtaining information about

the flow dynamics of groundwater. One of the most important applications is for groundwater dating. Amongst the most frequently used dating tracers we found: 3H/3He (Schlosser, 1989; Solomon and Cook, 1999), 85Kr (Loosli et al., 1999), CFCs (Plummer and Busenberg, 1999), 14C (Kalin, 1999), the stable noble gases (He, Ne, Ar, Kr and Xe; Stute and Schlosser, 1999) and the stable isotopes of the water 2H and 18O (Coplen et al., 1999).

Many processes determine the concentrations of isotopes or chemical tracers in groundwater (e.g. radioactive decay, hydrodynamic dispersion, mixing, chemical degradation, recharge date, transport time through unsaturated and saturated zones, etc.). Tracers suitable for groundwater dating are particular sensitive to time dependent processes (e.g. transport time through the saturated zone) where other processes are in the ideal case less dominant. However, in particular cases factors that lead to a misinterpretation of tracer concentrations in terms of residence time can provide important and additional information such as mixing and dispersion in the aquifer. The interpretation of the tracer concentrations is commonly carried out by models that try to mathematically describe the age distribution of sampled groundwater. The estimation of a set of free model parameters requires a corresponding number of measurements. This can be achieved with a high spatial sampling density or by applying multitracer measurements at selected locations. The latter technique, applied in the presented study, is particularly useful to consistently interpret single well measurements.

The study area is located in a semi-confined subsystem of the Paris Basin where the main objective was to determine the age structure of groundwater. The unconfined character of the aquifer leads to an age stratification with depths rather than to an aging along decreasing piezometric heads (Vogel, 1967). The limited number of sampling sites in the project area necessitates the application of multiple groundwater dating tracers and the use of lumped parameter approaches for the assessment of groundwater dynamics where the choice of an appropriate age weighting function that appropriately represents the hydrogeological situation can be validated using the measured tracer data. Thus, the results of several models can be compared and also the sensitivity of the model to parameters such as the mean residence time (MRT) or recharge rate can be investigated (Rueedi et al., in press).

The commonly adopted approach for dating and quantifying the portion of young (post-bomb) water components is the combination of either 3H and 3He or 85Kr (Schlosser, 1989; Solomon and Cook, 1999; Loosli et al., 1999). However, these tracers preferentially detect the “young tail” of broad age distributions. Hence, to verify to which extent the extrapolation of the interpretation with these tracers to the “old tail” fits the real situation in the aquifer an intermediate-age (< 1000 years) dating technique is required, which is what 39Ar enables in this study. This tracer has been proposed for dating groundwater (Loosli, 1983) for its ideal characteristics: a constant and well known atmospheric input concentration, no local contamination, an isotope ratio (39Ar/Ar) that is insensitive to degassing or incomplete gas extraction yield, and an important dating range for groundwater hydrology. 39Ar was used previously in selected studies (Andrews et al., 1984; Loosli et al., 1989; Pearson et al. 1991; Loosli, 1992; Beyerle et al., 1998; Loosli et al., 1999; Purtschert, 2001a).

Crucial for the application of environmental tracers for groundwater dating is the knowledge of a reliable local input function. The input functions may vary spatially but also as function of the depth of the water table below ground surface (Cook and Solomon, 1995). Both the delay and phase shift of transport through the unsaturated zone (USZ) of different tracers become significant in thick unsaturated zones. This is the always true in the investigated area with recharge depths between 20 and 45 meters. Therefore a one

37

dimensional transport model was integrated in the box model in order to calculate the tracer input at the water table. This procedure introduces additional parameters namely the porosity and tortuosity of the unsaturated soil and the recharge rate. Some of these parameters can be estimated based on complementary methods, others have to be determined by fitting to the tracer data. With the proposed inverse procedure mean recharge rates in the area of investigation could be estimated. The approach presented to investigate parameters in the Fontainebleau sands aquifer can be generalized and used for example for other tracers and for determining other parameters at many other sites.

2. SITE CHARACTERIZATION

The area of investigation is located in the shallower zone of the Paris Basin (France), which is the largest sedimentary basin of Western Europe (Fig. 1). The Oligocene sandy aquifer is embedded between two clayey layers (Fig. 2): above is the Beauce formation which was altered by diagenesis from limestone to millstone and clay; and below are Oligocene marls (Ménillet, 1988) which separate the Fontainebleau Sands from the underlying Eocene multi-layered aquifer.

Figure 1. Location of the study area showing the isopiezometric heads in the Fontainebleau Sands

Aquifer (modified after Rampon, 1965) and the location of the sampling wells. Arrows indicate the most probable flow directions.

Constituted by very fine, well-sorted silica grains with an average diameter of 100 μm, the

Fontainebleau Sands formation has a thickness of 50 to 70 m (Fig. 2), a hydraulic transmissivity of 1.10–3 to 5.10–3 m2.s–1 and a mean total porosity of about 25% (Mégnien et al., 1979; Mercier, 1981; Vernoux et al., 2001). The upper part of the formation is made of up to 99 % of pure quartz sands (white facies), while the content of organic matter, carbonates, sulphides, feldspar and clays (dark facies) increases with depth (Bariteau, 1996). It is assumed that the transition of the “white facies” to the “dark facies” is discrete rather than continuous (Schneider, 2005).

38

Figure 2. Geology of one section of the aquifer (left). Schematic display of the applied conceptual

model including the unsaturated zone (right). The mean precipitation rate in the area is about 700 mm per year (Station Trappes of

Meteo France: period of observation 1991 and 2000). The estimated recharge rates varies between 80 and 210 mm/a and they are based on hydrograph data (Mercier, 1981; Bariteau, 1996, Schneider, 2005). It was found that recharge rates in the outcrop areas of the sands in the valleys tend to be higher compared to the ones of the plateaus which are covered by the Beauce formation. Hence, as the sampling points of this study are located in the plateau areas, recharge rates are expected to be in the lower part of the estimated range.

Groundwater tables of the investigated wells lay between 20 and 45 meters below ground level (mbgl) (Table 1). There are at least four different areas with high piezometric heads (Fig. 1, modified after Rampon (1965)), indicating different flow regimes within the investigated region. The groundwater head distribution is mainly a consequence of the topography where water flows from the elevated plateaus to the lower valleys where groundwater discharges. During recent decades the aquifer has suffered a substantial abstraction to meet the water supply needs of the region. However, water tables decreased only slightly because of the high yield and conductivity of the aquifer.

Nitrate concentrations in groundwater between 0.27 and 0.45 mmol per liter have been reported in previous studies (Baseline Report, 2004). Agriculture is one of the main sources of nitrate in the region introducing considerable amounts through fertilizers usage. The observed elevated nitrate concentrations confirms the vulnerability of the aquifer to surface pollution and the presence of recent water components.

Previous studies suggested that, based on sulphate measurements, leakage could possibly occur from the underlying Eocene aquifers to the Fontainebleau Sands (Bergonzini, 2000). However, in the area of investigation, the potentiometric surface of the Eocene aquifer is far below that of the Fontainebleau aquifer preventing any upward seeping through the confining lower Oligocene (Baseline Report, 2004; Schneider, 2005).

39

Table 1. Characteristics of the wells of the Fontainebleau sands aquifer selected for the investigation. Groundwater was not abstracted from SM by more than 10 years before the sampling.

Soil Surface

Depth of the water Table

Depth of the water Table

Screen elevation (present days data)

Elevation in 1965a) today Screen 1 Screen 2 Screen 3 Well (masl) (mbgl.) (mbgl) (mbgl) (mbgl) (mbgl) SM 175.8 25-20 21 21.0-37.9 44.4-46.9 51.0-53.5CGEB 176.0 31-26 35 35.0-46.1 b) SA 171.0 31-26 38 38.0-49.2 b) LRN10 171.0 40-35 39 38.5-61.0 IMR 169.5 40-35 38 46.0-58.0 SLP4 154.5 40-35 55 55.4-56.0 60.5-61.0 66.8-67.3SLP5 141.5 26-21 43b) 50.0-50.5 52.5-53.0 a) Data from Rampon (1965). b) Uncertainty in the value of the depth. masl – meters above sea level. mbgl – meters below ground level.

The well screens in the studied region intercept the water table in most cases and some wells are screened over a large depth interval of the aquifer (Table 1).

3. METHODS 3.1 Field and laboratory investigations

Seven boreholes from the Fontainebleau Sands aquifer were sampled in October 2001 for extensive tracer investigation. Most of the sampling points are located in the central part of the aquifer, which is the part most heavily exploited for water supply. Field measurements of pH, dissolved O2, water temperature and electric conductivity were carried out and used to determine when the wells were sufficiently purged for sampling.

2 to 5 m3 of groundwater were degassed in the field to analyse the radioactive noble gases 37Ar, 39Ar and 85Kr. In the laboratory, the gases argon and krypton were separated from the samples; and 39Ar, 37Ar and 85Kr activities were measured by low level gas proportional counting in the Deep Laboratory of the Physics Institute, University of Bern, Switzerland (Loosli, 1983; Loosli et al., 1986; Forster et al., 1992).

Water samples for noble gas analyses were immediately transferred to 45 ml copper tubes and sealed with pinch-off clamps. The copper tubes were connected to the point of water withdrawal by flexible plastic tubing secured with hose clamps (gas tight) and the water was flushed by several minutes (until no bubbles were detected in the plastic tube) through the copper tube at high pressures before the steel clamps were closed. The measurements were carried out in the noble gas laboratory of ETH Zurich (Switzerland) according to the procedures described by Beyerle et al. (2000). Recharge temperatures (NGTs) are calculated from noble gas concentrations by accounting for their temperature-dependent solubilities and for the common excess air component found in groundwater (Aeschbach-Hertig et al., 1999, 2000).

The measurements of 3H were performed at the Physics Institute of the University of Bern by liquid scintillation after an enrichment step. 3H measurements were combined with the measurements of the decay product 3He to obtain 3H/3He ages (Schlosser et al., 1989). 3He concentrations of tritiogenic origin were calculated from the noble gas data (Beyerle et al., 2000).

One litre samples were collected for the analyses of the carbon isotopes (14C and 13C) content of the dissolved inorganic carbon. Radiocarbon is a well-established dating tool for water with recharge date between 1000 and 30000 years ago. The radiocarbon activities were

40

measured by AMS (graphite sources, Université Paris-Sud and measurements at Tandetron, Gif sur Yvette) and are expressed as a percentage of modern carbon (pmC). The δ13C contents were measured by mass spectrometry at the Laboratoire d'Hydrologie et de Géochimie Isotopique of the Université Paris-Sud and are expressed in permil variations from the Vienna Peedee Belemnite Standard [‰ VPDB].

3.2 Strategy of interpretation of tracer data

The three isotopes 3H, 3He and 85Kr are in particular sensitive for young groundwater components with residence times less than about 50 years. Lacking more detailed data in the area of investigation simple age frequency distributions are assumed for these young waters. This so called lumped parameter models (LPM) or box models (Zuber, 1986; Zuber and Maloszewski, 2001) are constrained by few parameters like the mean residence time and a parameter describing the dispersion of the age distribution. We can expect reliable results from this approach if a) the input concentrations at the water table are correct, b) the assumed age distribution is adequate, and c) a considerable fraction of the water is marked with tracer or with other words is younger than 50 years. In our study case these factors are taken into account as follows:

a) The local temporal evolution of atmospheric concentrations of 3H and 85Kr are already afflicted by some uncertainties. Additionally transport through the thick unsaturated soil zone has to be considered. We assume that this effect is similar for all of the investigated wells and that spatial variations are averaged out due to mixing and dispersion because of spatially distributed recharge.

b) Box model age distributions are simple idealisations of complex flow patterns in aquifers (Zuber, 1986; Zuber and Maloszewski, 2001). In the present case the exponential model (EM) seems to be the appropriate approach but also other models have to be examined by sound statistical criteria. Large scale heterogeneities and multi component mixing can lead to a even larger age dispersion than predicted by the EM (Weissmann et al., 2002). This tailing towards older ages can only be examined with a tracer which is sensitive to water ages older than 50 years. 39Ar is a suitable isotope for this purpose.

c) The portion of water that is younger than 50 years defines the sensitivity of the estimated over all age distribution to 3H/3He and 85Kr concentrations. Even in the case of an exponential age distribution which weights the youngest waters most the portion of tracer bearing water is less than 20% if the mean residence time is higher than 224 years. The resulting uncertainties due to the extrapolation to the water portion which contains no modern tracers can again be reduced by 39Ar data.

Our approach of data analyses includes therefore the following stages. A general transport model of unsaturated zone flow (one-dimensional advection-diffusion-decay transport model: 1D-ADDTM) is combined with lumped parameter models (LPM) for saturated flow at each individual well. Some parameters of 1D-ADDTM are known from literature. Others like a mean unsaturated zone thickness (for all wells) and the recharge rate are included in the inversion procedure. LPM parameters for each well are then estimated by a χ2 fitting routine. Thereby 3H, 3He and 85Kr data are treated symmetrically and a priory no tracer or tracer ratio is favoured (e.g. the 3H/3He ratio). The parameters which best explain the data are then checked for consistency with the 39Ar measurements.

41

3.2.1. Lumped parameter models and input function

The LPM is given by a weighting function h(t,pj) with parameters pj that describes the age distribution of the water (Table 2). The convolution of tracer input cin to tracer output cout for a certain sampling date Ts is calculated according to the formula,

∫∞

⋅⋅−⋅−=0

),()exp()(),( dtpthttTcpTc jsinjsout λ (1)

where t is the integration time, λ is the decay constant and pj are the model parameters given in Table 2. A more detailed description of the LPM is found in Zuber (1986) and Zuber and Maloszewski (2001). Table 2. Description of the lumped parameter models (LPM), their weighting functions and parameters. τ is the

mean residence time (MRT), δ is the Dirac delta function (PFM), η is the ratio of the total volume to the volume with exponential distribution of transit times (η is applied in the EPFM). Pe is Peclet number and defines the relative importance of advective and dispersive flow. (Pe is applied in the DM). Mixing of different groundwater components can potentially occur in the Fontainebleau sands aquifer. This mixing requires the introduction of an additional parameter m which accounts for the fraction of younger water (< 50 years). A fraction of water (1-m) is older due to reasons that have to be investigated.

LPM Weighting function Parameters Piston flow model (PFM) )(),( τδτ −⋅= tmmh τ, m

Exponential model (EM) ⎟⎠⎞

⎜⎝⎛ −

⋅=ττ

τ tmmth exp),,( τ, m

Dispersion model (DM) ( )

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

⋅⋅−⋅−

⋅⎟⎠

⎞⎜⎝

⎛⋅⋅

=t

tPet

PemPethτ

τπ

ττ4

exp14

),,,(22

3

τ, m, Pe

Exponential-Piston flow model (EPFM) ⎟

⎠⎞

⎜⎝⎛ −+

⋅−⋅

⋅= 1exp),,,( η

τη

τηητ tmmth for t ≥ τ(1-1/ η),

0 for t < τ(1-1/ η)

τ, m, η

It is crucial to select or determine the correct input function in Eq. 1. In confined aquifers with long groundwater residence times and comparable short transit times through the USZ the delay of the environmental tracers in the USZ can be neglected (Cook and Solomon, 1995). In these cases the atmospheric concentrations of the tracers can be used as starting values. However this simplification is probably not valid in the case of the Fontainebleau Sands aquifer because of the extended USZ with a thickness of 20-45 meters (Fig. 2, Table 1). Therefore, the transport of the tracers through the USZ has to be taken into account when dating young groundwater components (Cook and Solomon, 1995). The tracers (e.g. 3H, 85Kr) are transported through the unsaturated zone by advection and diffusion in both the water and gas phases. While water-bound tracers like 3H are transported mostly advectively with the water seepage, the transport of gaseous tracers (85Kr) is mainly diffusion controlled within the unsaturated soil pores.

A one-dimensional advection-diffusion-decay transport model (1D-ADDTM) was used to simulate the variations of the 85Kr and 3H concentrations C in the USZ as function of depth z and particularly to estimate their concentrations above the groundwater table (Cook and Solomon, 1995; Rueedi et al., in press). Assuming homogeneous physical conditions throughout the USZ, the equation can be expressed as:

42

llll C

zC

qzC

Dt

C⋅−⎟

⎠

⎞⎜⎝

⎛∂

∂⋅−⎟⎟

⎠

⎞⎜⎜⎝

⎛

∂∂

⋅=∂

∂λ*

2

2* (2)

with

( )*

*

θρρ KDD

D ggll ⋅⋅+⋅= [m2/y]

**

θρllqq ⋅

= [m/y]

Kggll ⋅⋅+⋅= ρθρθθ * [kg/m3]

with

lg CKC ⋅=

θl and θg are the water filled and the gas filled porosities respectively;

λ is the decay constant of the radioisotopes [y-1];

ρ is the density [g/cm3];

ql is the advective flow velocity [m/y];

D is the diffusion coefficient in the gaseous (g) and liquid (l) phase respectively [m2/y].

K is the equilibrium partition coefficient between the liquid and gas phase. Note that an instantaneous equilibration between the two phases is assumed.

The diffusion coefficients D, in the water and gas phase are estimated by,

Gas phase: ggogg DD τθ ⋅⋅= [m2/y]

Liquid phase: l

lll

oll

qDDθ

αθτ ⋅+⋅⋅= [m2/y]

where Do is the self-diffusion coefficient of the species in air (g) and in water (l); τg or l are the

gaseous (g) and liquid (l) tortuosity (value: 0 < τg or l < 1); and α is the dispersivity [m].

Equation 2 was solved using an implicit Crank-Nicholson scheme. The upper and lower boundary conditions are given by the atmospheric input concentration at the soil surface and zero diffusive flux at the water table in depth Z (Rueedi et al., in press). The latter condition is justified by the fact that the diffusion coefficient D* drops by about four orders of magnitude at the groundwater surface. A detailed description of the numerical solution can be found in Rueedi et al. (in press).

The 85Kr atmospheric activities are measured routinely at Freiburg (IAR - Institute of Atmospheric Research, Freiburg, Germany). These activities have been taken as the 85Kr input function in several groundwater studies. However, direct measurements of 85Kr in soil gas samples from the USZ of the Fontainebleau sands aquifer and in air samples in the region of location of the aquifer showed activities consistently 1.5 times higher than those measured in Freiburg (Corcho et al., 2004). Such higher activities in the Fontainebleau area than in Freiburg have been attributed to a closer location to the 85Kr emission sources (La Hague and Sellafield) (Corcho et al., 2004). Therefore, the 85Kr atmospheric activities measured at Freiburg were multiplied by a correction factor of 1.5 in order to estimate the input function at the soil surface in the Fontainebleau area. This scaling factor agrees with values derived from

43

atmospheric circulation model calculations (Winger et al., 2005) based on 85Kr emission data from different sources (e.g. nuclear reprocessing plants) in the world.

3H fallout was constructed averaging the 3H fallout data reported for the stations located in Le Mans and Orleans-La-Source (data taken from IAEA/WMO database). Previous studies have shown that 3He produced by radioactive decay of 3H during the transport through the unsaturated zone is completely lost to the atmosphere (Rueedi et al., in press).

According to sensitivity analysis performed by Rueedi et al. (in press) on an analytical solution of the advective-diffusive-decay transport equation for a semi-infinite medium, the parameters that most strongly influence the model results for the water dominated tracer 3H are the water filled porosity (θl); the dispersivity (α); and recharge rate (ql) whereas the gaseous tracer 85Kr depends mostly on the gas filled porosity (θg) and the tortuosity in the gaseous phase (τg). Therefore, results from previous studies were used to better constrain the soil parameters used in the 1D-ADDTM. Values of about 0.10 for θl and of about 0.25 for the total porosity were reported for this formation (Mégnien et al., 1979; Vernoux et al., 2001; Schneider, 2005). The dispersivity α, was set to 0.1 m (Cook and Solomon, 1995). In fact, 3H profiles of the sand dominated unsaturated zone in northwestern Senegal show that dispersion is at most 0.1 meters (Gaye and Edmunds, 1996). The tortuosities are assumed to be about 0.6 for the gaseous phase (Millington, 1959) and about 0.25 for the liquid phase (Barraclough and Tinker, 1982).

1950 1960 1970 1980 1990 20000

15

30

45

60

75

90

105

120

135

150

85Kr

(dpm

/ccK

r)

1950 1960 1970 1980 1990 2000

0

500

1000

1500

2000

2500

3000

3500

3H at the water table

85Kr at the water table

3H fallout

Atmospheric 85Kr(Correction factor f=1.5)

3H

(TU

)

Figure 3. The plot presents the 3H (constructed from nearby monitoring stations, GNIP: IAEA data

base) and 85Kr (the activities are a factor of f=1.5 higher than the ones from the IAR, Freiburg) atmospheric concentrations at Fontainebleau since 1950. Also shown are the 3H and 85Kr input concentrations at the water table of the Fontainebleau Sands aquifer (estimated using the 1D-ADDTM under the assumptions of a water-filled porosity of 10%, a total porosity of 25%, a gas tortuosity of 0.6, a water tortuosity of 0.25, a dispersivity of 0.1 m, a mean recharge rate of 150 mm/a and an USZ thickness of 25 m).

In Figure 3, the result of modelling the tracer transfer through the USZ for the specific

case of a water table at 25 m depth and a recharge rate of 150mm/a is presented. The 3H input at the water table is smoothed out and flat compared to the atmospheric fallout curve mainly

44

due to hydrodynamic dispersion and radioactive decay. The resulting transport times of 85Kr and 3H through the USZ range between 1 and 6 years and 10 and 40 years respectively depending on the water table depth and recharge rate. Since 3H is a part of the water molecule, its time lag represents the average transport time of the water from the soil surface to the water table.

3.2.2. The inverse fitting procedure

Estimates of the parameters θl, θg, τg, τl and α of the 1D-ADDTM are reasonable known in the Fontainebleau Sands aquifer (Table 3) and were fixed in the calculations. On the other hand, estimated values of the recharge rate ql in the area of investigation vary between 80 and 210 mm/a (Mercier, 1981; Bariteau, 1996; Schneider, 2005). This parameter was therefore included in the fitting as a global variable and was allowed to vary in a range between 50 and 500 mm/a. The thickness of the unsaturated soil zone is known at each well (i). However, recharge occurs spatially distributed potentially at any location between the wells. It was decided to select the mean recharge depth Z also as a global fitting parameter varying manually in the observed range between 20 and 40 m. Table 3. Parameters used for the one dimensional advection-diffusion decay transport model. Also it is indicated

whether the parameter is free or fixed in the modelling. Parameter Value Free or fixed Water filled porosity (θl) 0.1 a) Fixed Gas filled porosity (θg) 0.15 a) Fixed Dispersivity (α) 0.1 m b) Fixed Tortuosity in the gaseous phase (τg) 0.6 c) Fixed Tortuosity in the liquid phase (τl) 0.25 d) Fixed Recharge rate (ql) 50-500 mm/a a) Free Recharge depth (Z) 20-40 mbgl Free

a) Mégnien et al., 1979; Vernoux et al., 2001; Schneider, 2005; b) Cook and Solomon, 1995; c) Millington, 1959; d) Barraclough and Tinker, 1982

The combination of the LPM with the 1D-ADDTM can then be summarized by

considering a vector modeli containing 2 + (2 or 3) model parameters (the number of parameters depend of the LPM selected):

modeli=[ ql, Z, τi, mi, ηi or Pei]

and the measured tracer data for one well (i) as a vector of three parameters,

datai=[C(3H), C(3He), C(85Kr)]

The parameters (τi, mi, ηi or Pei) that best fit the LPM to the observations are estimated by minimizing the χ2

i function,

( )∑ −=

n

y y

yel

ymeasured

iCC

2

2mod2

δχ (3)

where Cymeasured are the measured tracer concentrations (y: 85Kr, 3H or 3He), δy is the error of

the measurements and Cymodel are the concentrations (cin) predicted by the LPM after

convoluting the input concentrations of the tracers calculated with the 1D-ADDTM Weighting with 1/δ2

y favours accurate measurements in the fitting routine (Press et al., 1986; Aeschbach-Hertig et al., 1999). The number of degrees of freedom V of the χ2 distribution is defined by V=N-M, where N is the number of measurements and M the number of free parameters of the model. This inverse approach offers the possibility of an error estimation of

45

the fitted parameters of the model. They are calculated from the covariance matrix based on the propagation of the experimental errors (Press et al., 1986; Aeschbach-Hertig et al., 1999).

Similar hydrogeological conditions are expected in all wells. Therefore, for the sake of simplicity, all the tracer data are interpreted simultaneously with the LPMs. Hence, a total χ2

T, which is defined by χ2

T=Σχ21-7 (where the suffixes 1-7 represent the respective well: SM,

CGEB, SA, LRN10, IMR, SLP4 and SLP5), was calculated and used to select the best fit of the LPM to the whole set of modern tracer data (3H, 3Hetri and 85Kr).

In order to reduce the correlations between parameters and to facilitate the inverse fitting procedure with the DM and the EPFM, the parameters (Pe) and (η) are constrained. This is done because the parameters calculated with these models of three unknown parameters often show correlations when fitting a set of three tracer data. Both lumped parameters (Pe,η) depend highly on other aquifer parameters such as the hydrodynamic dispersion, the size of the recharge area, the length of the screen interval and the depth of the screen below the water table. Hence different values of (Pe) and (η) are in principle expected at each well. However, taking into account the high homogeneity of the sands where the aquifer is located and the relative analogous hydrogeological conditions in the sampled wells, it is decided to assume similar values for the parameters (Pe) and (η) in all wells. This constrain allows to fit some parameters (τ,m) individually for each well (i) and some (Pe or η) for all the wells, which increases the total degrees of freedom of the DM and EPFM to six.

The number of tracers measured in each sample is three (3H, 3Hetri and 85Kr) and the total

number of wells is seven. A total of seven degrees of freedom for the EM and PFM (both with 14 parameters), and of six degrees of freedom for the DM and EPFM (both with 15 parameters) is obtained.

Summarizing, the modern tracers (3H, 3Hetri and 85Kr) data are interpreted in three main steps as follows:

1. The 1D-ADDTM is used to calculate the input functions of the tracers 3H and 85Kr into groundwater (cin) for different values of the parameters recharge rate ql (50-500 mm/a) and depth Z (20-40 mbgl). The other parameters of this model are fixed as explained above in this section (θl = 0.1, θg = 0.15, α = 0.1 m, τg = 0.6 and τl = 0.25) (Table 3).

2. By minimizing χ2T Eq.3, the LPMs parameters were adjusted to reproduce the tracer

data (3H, 3Hetri and 85Kr) for the different input functions calculated (1).

3. Finally, the extrapolation of the calculated age distribution to waters older than 50 years must be validated by 39Ar data. In cases where the parameter m is equal to 1, which means that one water component is predicted in the well, the 39Ar concentration calculated with the LPM must be comparable to the measured 39Ar concentration. In the other way the individual fit and consequently the whole set of fits (χ2

T) are rejected. If the parameter m is lower than 1, which means that two water components are predicted in the well; then the 39Ar concentration calculated with the LPM for the young water fraction must be comparable to or lower than the measured concentration. If the calculated 39Ar activity in the young water fraction is higher than the measured activity, then a negative value for the 39Ar concentration in the old water component is derived. A result that has no physical sense. Therefore the fit and consequently the whole set of seven fits (χ2

T) are rejected.

46

Table 4. Radioactive and stable isotope amounts measured in samples from the Fontainebleau Sands aquifer. Also shown the 85Kr ages assuming piston flow (PF) and the 3H/3He ages (3He concentrations calculated from the noble gas data, Table 5). The 39Ar ages calculated assuming piston flow and exponential distribution are as well listed.

Well 85Kr [dpm/cm3Kr]

85Kr PF age [yrs]

3H [TU]

3H/3He age [yrs]

39Ar [%modern]

39Ar PF age [yrs]

39Ar EM age [yrs]

14C [pmC]

13C [‰ vs.

VPDB]

SM 43.0 ± 5.0 11 ± 1 10.0 ± 0.8 8 ± 1 79 ± 7 91 ± 35 103 ± 45 80.2 ± 0.6 -16.3 CGEB 6.8 ± 0.7 29 ± 1 8.5 ± 0.8 9 ± 1 73 ± 5 122 ± 27 144 ± 37 75.1 ± 0.6 -14.0 SA 16.1 ± 4.1 20 ± 2 15.1 ± 0.8 11 ± 1 69 ± 5 144 ± 28 174 ±41 84.2 ± 0.6 -14.3 LRN10 6.1 ± 4.8 30 ± 5 7.8 ± 0.8 15 ± 1 77 ± 5 101 ± 25 116± 33 73.7 ± 0.6 -14.1 IMR 2.9 ± 0.4 35 ± 1 3.1 ± 0.8 9 ± 2 55 ± 5 232 ± 36 318 ± 65 69.8 ± 0.6 -13.8 SLP4 6.2 ± 2.5 30 ± 3 7.8 ± 0.8 2 ± 1 59 ± 5 205 ± 33 270 ± 56 75.5 ± 0.6 -13.9 SLP5 5.6 ± 2.8 31 ± 5 4.0 ± 0.8 13 ± 2 51 ± 5 261 ± 38 373 ± 76 73.8 ± 0.6 -13.7

Table 5. Noble gas data (Concentrations expressed in cubic centimetres of gas at STP per gram of water, [cm3/g]). Noble gas temperatures obtained with the unfractionated air model, together with the amount of excess air expressed as ∆Ne and the 3He concentration of tritiogenic origin.

Well [He] 10-8

3He/4He 10-6

[Ne] 10-7

[Ar] 10-4

[Kr] 10-8

[Xe] 10-8 χ2

Probability [%]

NGT [°C]

∆Ne [%]

3He tritiogenic

[10-14]

He radiogenic

10-9 [cm3/g]SM 6.45 1.57 2.71 4.18 9.50 1.34 0.18 91.3 9.68 ± 0.21 36.9 1.29 ± 0.08 -2.7 ± 0.9 CGEB 6.00 1.61 2.58 4.09 9.36 1.33 0.44 80.4 9.84 ± 0.22 30.5 1.44 ± 0.05 -2.0 ± 0.9 SA 6.12 1.87 2.61 4.08 9.32 1.32 0.65 72.3 10.08 ± 0.22 32.0 3.09 ± 0.08 -2.3 ± 0.9 LRN10 6.65 1.75 2.77 4.20 9.53 1.35 0.77 68.0 9.59 ± 0.21 39.4 2.56 ± 0.08 -1.3 ± 0.9 IMR 7.45 1.44 3.13 4.33 9.54 1.32 0.03 98.6 10.51 ± 0.23 58.9 0.52 ± 0.05 -4.2 ± 1.1 SLP4 6.49 1.40 2.76 4.17 9.39 1.32 0.06 97.0 10.27 ± 0.21 39.6 0.17 ± 0.05 -3.2 ± 1.0 SLP5 6.67 1.53 2.84 4.18 9.27 1.31 0.52 77.0 10.63 ± 0.22 44.2 1.10 ± 0.05 -3.7 ± 1.0 1σ error 0.03 0.01 0.03 0.02 0.14 0.01

4. RESULTS AND DISCUSSION 85Kr and 3H were detected at all sampling sites at levels well above the detection limits,

indicating the presence of modern water. Measured concentrations of 3H vary from 3.1 to 15.1 TU, and the specific activities of 85Kr show large variation between 2.9 and 43.0 dpm/cm3 Kr (Table 4). The noble gas composition of the water samples are shown in Table 5. In general, the interpretation of noble gas concentrations, which is described in more detail in Annex 2, provides evidence that degassing of atmospheric gases within the aquifer is most probably not a significant process. 3Hetrit concentrations, NGT and excess air components were calculated based on a model that assumes that excess air is pure atmospheric air (unfractionated air or UA model). However, a model that assumes partial diffusive re-equilibration (PR model) and correct for a fractionation parameter provides as well acceptable fits, with results that do not differ significantly from those calculated with the UA model. The PR model predicts an excess air lost of less than 17%.

39Ar activities range between 51 and 79 %modern, while 14C activities lay between 69 and 84 pmC (Table 4). The relatively high 85Kr and 3H concentrations together with relatively low 39Ar values is a clear indication for pronounced mixing of water components with different ages. Piston flow and exponential ages deduced from 39Ar activities range between 91-261 years and 103-373 years, respectively. Underground production of 39Ar, which most probably can be neglected in this aquifer (Annex 1), would shift these ages to even higher values. In any case, it can be assumed that a considerable portion of water in the investigated wells must be older than 50 years and is therefore free of 85Kr and 3H.

47

In a first step, the modern tracers (3H, 3Hetri and 85Kr) are used to identify, quantify and date the young groundwater components present in the aquifer. This first part is performed following the methodology presented in section 3. Then in a second part, the tracer 39Ar is used to date the old groundwater components, or in other words, to date the groundwater components that do not contain modern tracers (water that recharged before the year 1950). Measurements of 14C are also used to further constrain the ages of the old groundwater components.

4.1 Investigation of the young water components.

The comparison of uncorrected 85Kr (the atmospheric activities of 85Kr are used as input function) and 3H/3Hetri piston flow ages (Table 4) depict large discrepancies which are a consequence of different time lags of 85Kr and 3H, degassing of 3He in the unsaturated zone and mixing of waters. It is evident that only an integrated approach can consistently interpret all data.

Table 6. Mean residence time (τ) and mixing ratio (m) that best fitted the EM, the EPFM (η is also shown) and

the DM (Pe is also shown) to the measurements of 3H, 3He and 85Kr. The χ2T of the fits and the

correlation factor between the calculated parameters are also reported. Model parameters Correlation factors

Sample χ2 τ (years) m Pe η τ - Pe τ - η τ - m Pe - m η - mEM, recharge rate of 35 mbgl and recharge rate of 150 mm/a

SM 17 4.4 ± 0.3 0.36 ± 0.03 -0.826 CGEB 8 15 ± 2 0.18 ± 0.01 0.363

SA 1 11 ± 1 0.41 ± 0.01 -0.750 LRN10 2 117 ± 335 1.0 ± 1.0 1.000

IMR 1 13 ± 3 0.07 ± 0.01 0.303 SLP4 27 2.4 ± 0.4 0.17 ± 0.02 -0.531 SLP5 0 26 ± 28 0.16 ± 0.04 0.995 χ2

T 56 EM, recharge rate of 30 mbgl and recharge rate of 150 mm/a

SM 4 4 ± 1 0.46 ± 0.07 0.817 CGEB 18 13 ± 2 0.16 ± 0.03 -0.523

SA 9 9 ± 2 0.44 ± 0.08 -0.934 LRN10 0 28 ± 10 0.30 ± 0.04 0.992

IMR 2 12 ± 3 0.06 ± 0.01 -0.173 SLP4 41 3 ± 3 0.16 ± 0.15 -0.547 SLP5 0 13 ± 1 0.17 ± 0.01 -0.932 χ2

T 74 DM, recharge rate of 30 mbgl and recharge rate of 150 mm/a

SM 8 5 ± 1 0.46 ± 0.13 0.23 ± 0.12 -0.802 -0.589 0.121 CGEB 17 22 ± 50 0.20 ± 0.09 0.23 ± 0.12 0.976 0.994 0.988

SA 7 10 ± 3 0.45 ± 0.16 0.23 ± 0.12 0.878 0.550 0.869 LRN10 5 30 ± 110 0.34 ± 0.23 0.23 ± 0.12 0.879 0.990 0.935

IMR 3 17 ± 21 0.06 ± 0.02 0.23 ± 0.12 0.914 0.938 0.920 SLP4 40 4 ± 1 0.14 ± 0.10 0.23 ± 0.12 -0.437 -0.511 -0.104 SLP5 1 30 ± 140 0.15 ± 0.12 0.23 ± 0.12 0.741 0.984 0.844 χ2

T 81 EPFM, recharge rate of 40 mbgl and recharge rate of 150 mm/a

SM 32 5 ± 2 0.27 ± 0.02 1.01 ± 0.07 0.067 -0.783 0.192CGEB 3 28 ± 26 0.28 ± 0.16 1.01 ± 0.07 -0.886 0.997 -0.912

SA 2 62 ± 164 1.0 ± 1.0 1.01 ± 0.07 -0.925 1.000 -0.931LRN10 40 20 ± 9 0.39 ± 0.1 1.01 ± 0.07 -0.604 0.955 -0.792

IMR 1 20 ± 15 0.09 ± 0.03 1.01 ± 0.07 -0.718 0.975 -0.751SLP4 14 2 ± 2 0.17 ± 0.02 1.01 ± 0.07 0.342 -0.584 -0.501SLP5 3 26 ± 28 0.20 ± 0.12 1.01 ± 0.07 -0.455 0.987 -0.574χ2

T 95

48

Best fits according to the procedure explained in section 3 are listed in Table 6 in order of increasing values of the total χ2

T. The best results could be obtained with the EM, DM and EPFM. The PFM leads to the larger deviations between the modelled and observed values, in agreement with the expectation. Best estimates of the parameters recharge rate and unsaturated zone thickness are in the ranges 100-200 mm/a and 30-40 mbgl, respectively. The estimated spatially averaged recharge rate for the aquifer (100-200 mm/a) is comparable to previous estimates of this parameter based on hydrological balances (80-200 mm/a) (Mercier, 1981; Bariteau, 1996; Schneider, 2005).

Mean residence times and mixing ratios obtained from the EM, DM and EPFM do agree in most cases within the calculated errors. This can be understood if the fitted dispersion parameters Pe and η are considered. The low values of the Pe (high dispersion) as well as η’s close to one (no PF part) correspond to age distributions which look very similar to the EM and must therefore lead to the same fitting results within uncertainties. The inverse modelling procedure confirms therefore that an exponential age distribution is the best approximation of the real situation in the investigated aquifer.

Figure 4. Contour plots of the χ2 surface in the parameter space calculated for the tracer data of the

wells a) IMR and b) LRN10, with the EM assuming the water table at 35 mbgl and a recharge rate of 150 mm/a. In the first case, the χ2 surface is circular and a well defined minimum is obtained. The parameter that best fitted the EM to the tracer data are: τ = 13 ± 3 years and m = 0.07 ± 0.01. In the second case, the model parameters show a strong correlation, and the confidence interval of the parameters become large. The parameter that best fitted the EM to the tracer data are: τ = 117 ± 335 years and m = 1.0 ± 1.0.

Correlation between parameters leads to large uncertainties of the estimation of these

parameters (Aeschbach-Hertig et al., 1999; Press et al., 1986). This can be visualised in a contour plot of the χ2 surface. Two examples are compared in Figure 4 where lines of equal χ2 are plotted as function of the mean residence time (τ) and portion of young water (m). In the first case the χ2 surface shows a circular shape with a well defined minimum and corresponding small parameter uncertainties (Sample IMR analysed with the EM: τ = 13±3 yrs, m = 0.07±0.01). A high correlation is shown in the second example. The calculated mean residence time for this case is 117±335 years, and the mixing ratio 1±1 (Sample LRN10 analysed with the EM, Table 6). Residence time and mixing portion are in particular correlated when an increase of residence time (more water above the 50 years border) can be compensated by a corresponding increase of the young water portion or vice versa. In such

49

cases the parameters can not be resolved without constraining one of them (Poeter and Hill, 1997; Carrera et al., 2005) or by adding further constrains such as 39Ar measurements.

0 2 4 6 8 10 12 14 16 18 200

2

4

6

8

10

12

14

16

18

20

1E-15 1E-14 1E-131E-15

1E-14

1E-13

0 10 20 30 40 500

10

20

30

40

50

SLP41

SM1

IMR SLP5

LRN10

SM

SA

CGEB

SLP4

Measured = Modelled * 1.02R = 0.93

Mea

sure

d 3 H

(TU

)

Modelled 3H (TU)

IMR

SLP5

SM, SM1 CGEB

LRN10SA

Measured = Modelled - 3.2E-16R = 1.0

SLP41, SLP4Mea

sure

d tri

tioge

nic

3 He

(cm

3 STP

/g)

Modelled Tritiogenic 3He (cm3STP /g)

SM1

SM

IMRSLP5

LRN10

CGEB

SA

SLP41, SLP4

Measured = Modelled * 0.94 R = 0.87

Mea

sure

d 85

Kr (

dpm

/cm

3 Kr)

Modelled 85Kr (dpm/cm3 Kr) Figure 5. Comparison between the modelled (obtained with the EM) and the measured concentrations

of 3H, 3He and 85Kr. Larger disagreements are observed in the samples SM and SLP4. Better agreements are observed at these wells when a shallower recharge depth of 20 mbgl and a deeper recharge depth of 40 mbgl are assumed, respectively. The points SM1 (at 20 mbgl) and SLP41 (at 40 mbgl) represent these conditions.

Overall there is a relative good agreement between the modelled and the measured tracer

concentrations for the best fits (Fig. 5). However, one has to keep in mind that the modelling is based on averaging some parameters like the recharge depth and recharge rate. In the area of SM the unsaturated soil zone tends to be thinner than average and in the area of SLP4 thicker than average (see depths of the water tables in Table 1). Taking this into account would also improve considerably the fit for these wells (Fig. 5).

Between 7 and 41 % of the groundwater follow exponential age distributions with MRTs between 2 and 28 years (Table 6). However, this implies that in order to describe the whole water mass an even broader age dispersion has to be assumed with larger fractions of groundwater older than 50 years.

4.2 Investigation of the old water components.

The older tail of the age distribution could not be characterized with the approach described above; therefore a different technique is required. The environmental tracers 39Ar and 14C are suitable to better constrain the ages in this timescale range

The measured concentrations of 39Ar are used to date the old groundwater components after correction for the contribution of the young groundwater components. Mean ages ranging between 138 and 341 years, with a mean value of 240 years are calculated if piston flow conditions are assumed (Table 7). A similar result is derived from the correlation line presented in Figure 6a, where 39Ar concentration is plotted against the fraction of young water (Fig. 6a). If an exponential-piston flow age distribution, where waters younger than 50 years are not included, is assumed; then 39Ar ages between 149 and 474 years are predicted for the old water components (Table 7). It can then be concluded from these results that old water fractions follow an age distribution with mean residence times of a few hundres of years.

50

0 20 40 600

20

40

60

80

100

0 20 40 6050

60

70

80

90

100

50

60

70

80

90

100

Mean 39Ar activity of old water components (55.0 %Modern = 232 years)

SM

SA

LRN10

CGEB

SLP5

IMR SLP4

39Ar = (0.5±0.1) * (%) + (55.0±3.8) R = 0.61

39Ar

(%M

oder

n)

Fraction of young groundwater (%)

Mean 14C activity of old water components (68.6±0.4 pmC)

Contribution of the old water component

Contribution of the modern component (bomb peak of 14C)

IMR

SLP5

CGEB

SLP4

LRN10

SA

SM

14C = (0.32±0.01) * (%) + (68.6±0.4) R = 0.96

14C

in T

DIC

(pm

C)

Fraction of young groundwater (%)

Figure 6. Correlation between the fractions of young water and the measured activities of: a) 39Ar (a

mean 39Ar activity for the old groundwater component of 55.0 %Modern is deduced by extrapolation to 0% young water the correlation line presented); and b) 14C in TDIC (a mean 14C activity for the old component of 68.6 pmC is deduced by extrapolation to 0% young water the correlation line presented). The contribution of bomb 14C is also indicated.

Some fractions of the abstracted groundwater recharged after 1950, hence some

contribution of the bomb 14C (14C produced during atmospheric testing of nuclear weapons) is expected in the groundwater samples. This effect can be better observed in Figure 6b, where the 14C activity in TDIC is plotted against the percent of young groundwater in the mixtures. The good correlation confirms the influence of the bomb 14C. The intercept of the correlation line most likely represents the mean 14C activity of the old groundwater components (68.6 pmC). A good estimate of the initial 14C activity in recharging groundwater of 65-75 pmC is obtained by the Fontes-Garnier model (Fontes and Garnier, 1979) assuming the following data (the water chemistry was taken from the Baseline report, 2004): a) an average δ13CCO2(g) value of -25 ‰ for the soil CO2 (Gillon et al., 2004), b) an assumed 14C activity of 100 pmC for soil CO2, and c) a δ13C and a 14C activity in carbonate minerals from the rock matrix equal to 0 ‰ and 0 pmC, respectively. The calculated initial 14C activity is comparable to the estimated mean 14C activity in old groundwater (68.6 pmC) which indicates a relative short residence time for the old groundwater fraction. This conclusion is in agreement with the interpretation of the 39Ar data. Furthermore, it confirms that leakage of very old groundwater from the underlying Eocene aquifer, where much lower 14C activities (<10 pmC) have been reported (Schneider et al., 2004), is not occurring in the investigated area.

Finally, the MRT of the whole abstracted groundwater are estimated by weighting the mean age of each water component in the mixture by their percents (Table 7). Combining the MRT calculated by the EPFM in the old water fractions (Table 7) with the ones obtained with the EM for the young groundwater fractions (Table 6), MRTs for the abstracted groundwater ranging between 107 and 396 years are calculated (last column in Table 7).

The calculated MRTs of the abstracted groundwater show a trend to higher values from the plateaus (SM, CGEB, SA and LRN10) towards the valleys (SLP4, SLP5). This patter is consistent with the local directions of the groundwater flows in the aquifer as indicated by the distribution of potentiometric surfaces in the investigated region (Fig. 1).

51

Table 7. 39Ar specific activities measured in groundwater, the ones predicted with the EM in the young water fractions (for the conditions shown in table 6), and the 39Ar activities of the old components after correction for mixing. Last column contain the resulting mean residence times of the abstracted groundwater which are calculated by weighting the age of each component in the mixture by the their percents.

Well 39Ar measured

39Ar predicted with EM

Old component 39Ar activity 39Ar ages 39Ar ages

MRT1)

(%Modern) (%Modern)

Fraction of old water

(%) (%Modern) PFM (years) EPFM (years) (years) SM 79 ± 7 36 ± 3 64 ± 3 68 ± 12 151 ± 68 164 ± 92 107 ± 59 CGEB 73 ± 5 17 ± 1 82 ± 1 68 ± 6 150 ± 36 164 ± 45 137 ± 37 SA 69 ± 5 40 ± 1 59 ± 1 49 ± 9 274 ± 68 358 ± 133 216 ± 79 LRN10 77 ± 5 28 ± 4 70 ± 4 70 ± 10 138 ± 53 149 ± 72 113 ± 51 IMR 55 ± 5 7 ± 1 93 ± 1 52 ± 6 255 ± 41 318 ± 77 297 ± 72 SLP4 59 ± 5 17 ± 2 83 ± 2 51 ± 7 263 ± 50 331 ± 94 275 ± 78 SLP5 51 ± 5 17 ± 1 83 ± 1 42 ± 6 341 ± 57 474 ± 119 396 ± 99 Mean age: 240 ± 52 280 ± 90 220 ± 68

1) Calculated as: )()1( 39ageArmmMRT ⋅−+⋅= τ , where m and τ are taken from table 4, and 39Arage is the

39Ar age of the old groundwater component calculated with the EPFM.

4.3 Analysis of the tracer methods results.

The combined 1D-ADDTM and LPM, which were used to interpret the modern tracers, can explain only a small fraction (less than 20% in four wells, and between 30 and 45% in the other three wells) of the groundwater age distribution in the Fontainebleau sands aquifer. A large fraction of groundwater with an older mean residence time is also present in all of the investigated wells. In some wells the presence of different water components might probably be explained by the existence of more than one screen interval, but this explanation cannot be generalized for all wells (Table 1). Therefore, other processes of mixing must be considered to explain the tracer dating results. The available geological data do not provide evidence (e.g. impermeable layers, regions with faster or lower groundwater flow velocities within the aquifer, etc.) regarding the existence of water components with different MRTs. It can then be concluded that either the formation where the aquifer is located is not as homogenous as it was described before or that the interpretation of the modern tracers is not precisely enough to describe the whole age distribution of groundwater in the aquifer. As stated above there are no evidences to support the first point. For the second point, it should be kept in mind that the interpretation of the modern tracers entails large uncertainties that are sometimes not included in the reported parameter errors and also that the groundwater age distribution is estimated from assumptions regarding flow conditions (see section 3). Misinterpretations are expected if the shape of the real groundwater age distribution in the aquifer differs significantly from the one assumed in the lumped parameter model. Also it should be mentioned that in well mixed and/or highly dispersive flow aquifers, which is the situation observed in the Fontainebleau sands aquifer, age dating techniques that detect young recharged groundwater (e.g. CFCs, 85Kr, 3H) could provide mean residence times that deviate considerably from real groundwater mean residence times (Weissmann et al., 2002). This effect is larger when the amount of groundwater that contain young tracers is small compared to the total sampled groundwater as in the Fontainebleau aquifer. Hence, it can not be excluded that the interpretation of the young tracers provided results that could be deviated from the real situation in the aquifer. The extrapolation of the obtained age distribution to older ages could probably be biased.

The simultaneous interpretation of 39Ar measurements resolved this problematic in the present study. Most of the sampled groundwater is marked by this tracer and consequently a more precise estimation of the mean residence time could be achieved. Groundwater mean residence times of a few hundreds of years were derived from its interpretation, a range that

52

was futher confirmed by 14C data. The application of 39Ar is evidently indispensable to properly evaluate the age distribution of groundwater in aquifers affected by a high mixing due to a high dispersion and/or long screened intervals and where most of the water does not contain modern tracers.

We can then build a simple conceptual model of the aquifer. The Fontainebleau Sands aquifer is homogeneously recharging all over the aquifer surface at rates varying between 100 and 200 mm per year. Due to the spatially extended recharge and the large screened intervals, a large spread of residence times is observed in the wells with a mean value of 100-400 years. Groundwater recharged in the plateaus flows towards the valleys where it discharges. The aquifer contains modern recharged groundwaters with elevated nitrate concentrations, which shows its vulnerability to surface pollution.

5. CONCLUSIONS

In aquifers that contain large portions of groundwater with ages over 50 years, it is necesary to use a tracer that can precisely date in an intermediate timescale range (<1000 years). Nowadays, 39Ar is the only tracer available that accurately date this age range. It was shown in the paper that this tracer may be useful not only for dating groundwater with ages ranging from hundred to thousand years, but also as an additional tool to constrain the interpretation of modern groundwater tracers. The use of 39Ar in this study enabled us to decide whether or not a set of parameters of a lumped parameter model were capable of describing the age distribution of groundwater in the Fontainebleau sands aquifer.

A one dimensional advective-diffusive transport model was successfully combined with the lumped parameter models to investigate parameters in the Fontainebleau sands aquifer. In order to find the values of the parameters that best fitted the lumped parameter models to the environmental tracer data, an error weighted inverse approach was discussed. One of the most important advantages of this approach is the quantification of the errors of the parameters.

A broad range of mean residence times are present within the investigated area of this aquifer. The unsaturated zone model predicted a time lag of between 1 and 6 years for 85Kr and a time lag of between 10 and 40 years for 3H for their transport through the unsaturated soil zone above the Fontaineblea sands aquifer. An excellent agreement between the modelled and the measured concentrations of 3H, 3He and 85Kr was obtained with the exponential model for a recharge that took place at 35 meter of depth. Old water fractions of between 59 and 93 % in the mixed samples were estimated with this model. The resulting spatially averaged recharge rates (100-200 mm/a) are in agreement with previous estimates of this parameter (80-200 mm/a).

6. ACKNOWLEDGEMENTS

This study was partially supported by the EU project BASELINE “Natural Baseline Quality of European Groundwaters: A Basis for Aquifer Management”. We would like to thank two anonymous reviewers for their helpful comments.

7. REFERENCES Aeschbach-Hertig W., Peeters F., Beyerle U. and Kipfer R. (1999) Interpretation of dissolved atmospheric noble

gases in natural waters. Water Resour. Res. 35(9), 2779-2792.

53

Aeschbach-Hertig W., Peeters F., Beyerle U. and Kipfer R. (2000) Palaeotemperature reconstruction from noble gases in ground water taking into account equilibration with entrapped air. Nature 405, 1040-1044.

Andrews J.N., Balderer W., Bath A.H., Clausen H.B., Evans G.V., Florkowski T., Goldbrunner J.E., Ivanovich M., Loosli H.H., Zojer H. (1984) Environmental isotope studies in two aquifer systems: A comparison of groundwater dating methods. In: Isotope Hydrology 1983 (IAEA-SM-270). IAEA, Vienna, p 535-577

Andrews J.N., Davis S.N., Fabryka-Martin J., Fontes J-CH., Lehmann B.E., Loosli H.H., Michelot J-L., Moser H., Smith B., Wolf M. (1989) The in situ production of radioisotopes in rock matrices with particular reference to the Stripa Granite. Geochim. Cosmochim. Acta 53, 1803-1815.

Appelo C. A. J. and Postma D. (1996). Geochemistry, Groundwater and Pollution. A.A. Balkema Publishers, Rotterdam, Netherlands, 536 pp.

Bariteau A. (1996) Modélisation géochimique d'un aquifère : la nappe de l'Oligocène en Beauce et l'altération des Sables de Fontainebleau. Thèse, Ecole des Mines de Paris.

Barraclough P.B. and P.B. Tinker (1982). The determination of ionic diffusion coefficients in field soils. II. Diffusion of bromide ions in undisturbed soil cores. J. Soil Sci. 33:13–24.

BASELINE Report (2004). Natural baseline quality in European aquifers - A basis for aquifer management. EC Framework V Project, EVK1-CT1999-0006.

Bath A.H., Edmunds W.M., Andrews J.N. (1978) Palaeoclimatic trends deduced from the hydrochemistry of a Triassic sandstone aquifer, United Kingdom. In Isotope Hydrology (ed. IAEA). IAEA, Vienna, 545-568.

Bergonzini L. (2000) Caractérisation géochimique de la nappe des Sables de Fontainebleau (abstract). 18ème Réunion des Sciences de la Terre, Paris.

Beyerle U., Purtschert R., Aeschbach-Hertig W., Imboden D.M., Loosli H.H., Wieler R., Kipfer R. (1998) Climate and groundwater recharge during the last glaciation in an ice-covered region. Science 282:731-734

Beyerle U., Aeschbach-Hertig W., Imboden D.M., Baur H., Graf T., Kipfer R. (2000) A mass spectrometric system for the analysis of noble gases and tritium from water samples. Environ. Sci. Technol. 34, 2042-2050.

Carrera J., Alcolea A., Medina A., Hidalgo J., Slooten L.J. (2005) Inverse problem in hydrogeology. Hydrogeol J., 13, 206-222.

Corcho Alvarado J.A., Purtschert R., Barbecot F., Chabault C., Rüedi J., Schneider V., Aeschbach-Hertig W., Kipfer R., Loosli H.H. (2004). Tracer transport in the unsaturated zone of the Fontainebleau sands aquifer. International Workshop on the Application of Isotope Techniques in Hydrological and Environmental Studies, IAEA/UNESCO , Paris, France.

Cook P.G. and Solomon D.K. (1995) Transport of atmospheric trace gases to the water table: Implications for groundwater dating with chlorofluorocarbons and krypton-85. Water Resour. Res. 31(2), 263-270.

Coplen T.B., Herczeg A.L. and Barnes Ch. (1999). Isotope engineering- using stable isotopes of the water molecule to solve practical problems. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 79-110.

Fontes, J. C.; Garnier, J. M. (1979) Determination of the initial 14C activity of the total dissolved carbon; a review of the existing models and a new approach. Water Resour. Res., 15, 399-413.

Forster M., Ramm K. and Maier P. (1992). Argon-39 dating of groundwater and its limiting conditions. In Isotope techniques in water resource development 1991 (ed. IAEA), IAEA, Vienna, 203-214.

Gaye C.B. and Edmunds W.M. (1996) Groundwater recharge estimation using chloride, stable isotopes and tritium profiles in the sands of northwestern Senegal, Env. Geology 27, 246-251.

Gelhar L.W., Welty C. and Rehfeldt K.R. (1992). A Critical Review of Data on Field-Scale Dispersion in Aquifers, Water Res. Res., 28, pp. 1955-1974.

Gillon M., Barbecot F., Gibert E., Marlin C., Massault M. (2004). Variability of CO2 composition in 13C within the unsaturated zone and influence on groundwater dating. In: International Workshop on the Application of Isotope Techniques in Hydrological and Environmental Studies, IAEA/UNESCO, Paris, France.

IAEA/WMO. The Global Network of Isotopes in Precipitation database (Accessible at: http://isohis.iaea.org).

Institute of Atmospheric Radioactivity (IAR), Freiburg, Germany.

54

Kalin R.M. (1999). Radiocarbon dating of groundwater systems. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 111-144.

Kipfer R., Aeschbach-Hertig W., Peeters F. and Stute M. (2002) Noble gases in lakes and ground waters. In Noble gases in geochemistry and cosmochemistry, Rev. Mineral. Geochem. 47 (eds. D. Porcelli, C. Ballentine and R. Wieler). Mineralogical Society of America, Geochemical Society, Washington DC, 615-700.

Lehmann B. and Loosli H.H. (1984) Use of noble gas radioisotopes for environmental research. RIS 1984. Inst. Phys. Conf. Ser. 71, Knoxville, Tennessie

Lehmann B.E. and Loosli H.H. (1991) Isotopes formed by underground production. In Applied Isotope Hydrogeology, a Case Study in Northern Switzerland (eds. F.J. Pearson et al.). Elsevier, Amsterdam, 239-296.

Loosli H.H. (1983) A dating method with 39Ar. Earth Sci. Planet. Lett. 63, 51.

Loosli H.H., Moeli M., Oeschger H. and Schotterer U. (1986) Ten years low-level counting in the underground laboratory in Bern, Switzerland. Nucl. Instr. Meth. B 17, 402-405.

Loosli H.H. and Lehmann B.E. (1989) Transfer of underground produced 37Ar, 39Ar and 40Ar from rock into water. Water-Rock Interaction WRI-6, 445-448

Loosli H.H., Lehmann B.E., Balderer W. (1989) Argon-39, argon-37 and krypton-85 isotopes in Stripa groundwaters. Geochim. Cosmochim. Acta 53, 1825-1829.

Loosli H.H., Lehmann B.E. and Däppen G. (1991) Dating by radionuclides. In Applied Isotope Hydrogeology, a Case Study in Northern Switzerland (eds. F.J. Pearson et al.). Elsevier, Amsterdam.

Loosli H.H., Lehmann B.E., Thalmann C., Andrews J.N., Florkowski T. (1992) Argon-37 and argon-39: measured concentrations in groundwater compared with calculated concentrations in rock. In: Isotope techniques in water resources development (IAEA-SM-319). IAEA, Vienna, p 189-201

Loosli H.H. (1992) Applications of 37Ar, 39Ar and 85Kr in hydrology, oceanography and atmospheric studies. In: Isotopes of noble gases as tracers in environmental studies, IAEA, Vienna, p 73-85

Loosli H.H., Lehmann B.E., Smethie W.M., 1999. Noble gas radioisotopes: 37Ar, 85Kr, 39Ar, 81 Kr. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 379-396.

Mégnien C. (1979) Hydrogéologie du centre du Bassin de Paris. In Mémoire B.R.G.M. 98, B.R.G.M., Orléans.

Ménillet F. (1988) Meulières, argiles à meulières et meuliérisation - historique, évolution des termes et hypothèses génétiques. Bulletin d'Information des géologues du Bassin de Paris 25(4), 71-79

Mercier R. (1981) Inventaire des ressources aquifères et vulnérabilité des nappes du département des Yvelines. Rapport B.R.G.M., 81SGN348IDF, B.R.G.M., Service géologique régional Île de France.

Millington R.J. (1959). Gas Diffusion in Porous Media. Science 130: 100-102.

Oeschger H., Gugelman A., Loosi H.H., Schotterer U., Siegenthaler U., and Wiest W. (1974). 39Ar dating of groundwater, in Isotope Techniques in Groundwater Hydrology, IAEA, Vienna. pp. 179-190.

Pearson F.J. Jr., Balderer W., Loosli H.H., Lehmann B.E., Matter A., Peters Tj., Schmassmann H. and Gautschi A. (1991). Applied Isotope Hydrology: A Case Study in Northern Switzerland. Studies in Environmental Science #43, Elsevier Science Publishers B.V., Amsterdam, 436 p.

Peeters F., Beyerle U., Aeschbach-Hertig W., Holocher J., Brennwald M. S., and Kipfer R. (2002a) Improving noble gas based paleoclimate reconstruction and groundwater dating using 20Ne/22Ne ratios. Geochim. Cosmochim. Acta 67(4), 587-600.

Peeters F., Aeschbach-Hertig W., Holocher J. and Kipfer R., (2002b) Excess air correction in groundwater dating with He isotopes, Goldschmidt conference, Davos, Switzerland. Geochim. Cosmochim. Acta (abstr.), A587.

Plummer L.N. and Busenberg E. (1999). Chlorofluorocarbons. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 441-478.

Poeter E.P. and Hill M.C. (1997). Inverse Methods: A Necessary Next Step in Groundwater Modeling. Ground Water, 35, no. 2, pp. 250-260.

55

Press W.H., Flannery P.F., Teukolsky S.A., Vetterling W.T. (1986). Numerical recipes. Cambridge Univ. Press, New York, 818pp.

Purtschert R., Lehmann B.E., Loosli H.H. (2001a). Groundwater dating and subsurface processes investigated by noble gas isotopes (37Ar, 39Ar, 85Kr, 222Rn, 4He). In: Water Rock Interaction, WRI-10, Vol. 2(ed. by R. Cidu), 1569-1573, Villasimus, Italy.

Purtschert R., Aeschbach-Hertig W., Beyerle U., Kipfer R. and Loosli H.H. (2001b) Palaeowaters from the Glatt Valley, Switzerland. In: Palaeowaters in coastal Europe: Evolution of groundwater since the late Pleistocene (eds. W.M. Edmunds and C.J. Milne). Geological Society Special Publication. Geological Society, London, pp. 155-162.

Rampon G. (1965) Etat de la documentation sur les ouvrages souterrains implantés sur les feuilles topographiques de Nogent le Roi-Rambouillet et synthèse hydrogéologique provisoire. Rapport du Bureau des Recherches Géologiques et Minières, DSGR. 65.A7, Service géologique régional du Bassin de Paris, Paris.

Rueedi J., Brennwald M.S., Purtschert R., Beyerle U., Hofer M., Kipfer R.. (in press) Estimating amount and spatial distribution of groundwater recharge in the Iullemmenden Basin (Niger) based on 3H, 3He and CFC-11 measurements. IAEA Conference on Isotope Hydrology, Vienna, 2003. Accepted by Hydrological Process – Special Edition.

Schlosser P., Stute M., Sonntag C., and Munnich K.O. (1989) Tritiogenic 3He in shallow groundwater. Earth Planet. Sci. Letters 94, 245-256.

Schneider V., Barbecot F., Bergonzini L., Marlin C., Filly A., Massault M., Dever L. (2004). Geochemical-hydrological relationships between two groundwater bodies of the Paris Basin: clues for a conceptual model. In: International Workshop on the Application of Isotope Techniques in Hydrological and Environmental Studies, IAEA/UNESCO, Paris, France.

Schneider V. (2005). Apports de l´hydrodynamique et de la géochimie à la caractérisation des nappes de l`Oligocène et de l`Éocene, et à la reconnaissance de leurs relations actuelles et passées: origine de la dégradation de la nappe de l`Oligocène (sud-ouest du Bassin de Pris). Ph.D. Thesis, Université Paris-Sud, U.F.R. Scientifique d`Orsay, France.

Solomon D.K. and Cook P.G. (1999). 3H and 3He. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 397-424.

Stute M., Forster M., Frischkorn H., Serejo A., Clark J.F., Schlosser P., Broecker W.S., Bonani G. (1995) Cooling of tropical Brazil (5 °C) during the Last Glacial Maximum. Science 269, 379-383.

Stute M. and Schlosser P. (1999). Atmospheric noble gases. In: Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 349-377.

Vernoux J.F., Le Nindre Y.M. and Martin J.C. (2001) Relations nappe-rivière et impact des prélèvements d'eau souterraine sur le débit des cours d'eau dans le bassin de la Juine et de l'Essonne. Rapport BRGM, BRGM/RP-50637-FR, B.R.G.M.

Vogel J.C. (1967) Investigation of groundwater flow with radiocarbon. In: Isotopes in Hydrology (IAEA-SM-83). IAEA, Vienna, p 355-369

Weissmann, G.S., Y. Zhang, .E.M. LaBolle, and G.E. Fogg. 2002. Dispersion of groundwater age in an alluvial aquifer system. Wat. Resour. Res. 38(10):16.1-16.8

Winger K., Feichter J., Kalinowski M., Sartorius H., Sclosser C. (2005). A new compilation of the atmospheric 85Krypton inventories from 1945 to 2000 and its evaluation in a Global Transport Model. J. of Env. Rad., 80, 183–215.

Zuber A. (1986). Mathematical models for the interpretation of environmental radioisotopes in groundwater systems. In Handbook of Environmental Isotope Geochemistry (Ed. Fritz P. and Fontes J.Ch.), V.2, The Terrestrial Environment B, pp. 1-59.

Zuber A. and Maloszewski P. (2001). Lumped-parameter models. In Environmental isotopes in the hydrological cycle, V.6. Modelling. UNESCO/IAEA, Tech. Doc. Hydrol. 39, 5–35.

56

Annex 1. 37Ar and 39Ar underground production.

In the atmosphere 39Ar is mainly produced by the reaction 40Ar(n,2n)39Ar and decays with a half-life of 269 yrs with 39K as decay product. Since nuclear weapon tests have not influenced the atmospheric 39Ar activity, the infiltrating water exhibits a constant 39Ar activity of 100% modern (0.107±0.004 dpm/L Argon) (Lehmann and Loosli, 1984). A possible limitation of the application of 39Ar arises from the presence of subsurface produced 39Ar (Loosli and Lehmann, 1989). Underground produced 39Ar is mainly the result of the interaction of neutrons with 39K atoms in the rock matrix. A high 39Ar production rate requires: i) high U and Th concentrations in the rock (Emitters of neutrons producing α particles via the α,n reaction ii) low concentration of n absorbing elements in the rock (Gd, B, etc) iii) high K concentration in the rock and iv) a high escape rate from rock into the water phase. Such conditions (in particular i) were e.g. fulfilled in the Stripa granite in Sweden, where the mean concentrations of uranium and thorium in the rocks were 44.1 and 33.0 ppm respectively, activity values of 39Ar higher than 1000 %modern were measured (Andrews et al., 1989). However in many aquifers with average crustal U and Th concentrations of ~2 and 6-10 ppm 39Ar activities below or at the detection limit were measured (Bath et al., 1978; Loosli, 1983; Purtschert et al., 2001b).

Lacking 39Ar free old waters in the investigated aquifer, which would indicate negligible underground production, one has two alternative methods to estimate the order of magnitude of underground production. A first method is based on the use of 37Ar. 37Ar is also produced by subsurface neutrons by the 40Ca(n,α)37Ar reaction and its short half-life of 35 days excludes any atmospheric component in old groundwater. 37Ar is therefore a good monitor for high n-fluxes. However, because the target elements K and Ca are elements of different rock minerals with probably different grain sizes and microscopic structure, a large uncertainty of the relative release rates of 39Ar and 37Ar remains. Similar problems are encountered when 37Ar measurements are replaced or complemented with theoretical simulations of the subsurface neutron fluxes and production rates of 37Ar and 39Ar (Lehmann and Loosli, 1991).

In the present study, the underground production of 39Ar was estimated using both methods. Based on measured mean rock composition, an estimated (large) range of escape rates e of 37Ar and 39Ar between 0.1 and 10 % (Loosli et al., 1991) with a rock porosity p of 25 % (Mercier, 1981) equilibrium concentrations of 37Ar and 39Ar in the rock matrix (NR) and in the water NW were calculated using the relationship:

peNN Rw ⋅= (4)

Upper limits of the underground produced activities (e=10%) are 10-5 dpm/l Ar and 6 %modern for 37Ar and 39Ar respectively (Table 4). The low value for 37Ar is in agreement with the measured activity in well IMR (below DL of 0.03 dpm/L Argon; for comparison 37Ar in Stripa groundwater is between 2-13 dpm/L Ar (Loosli et al., 1989)) providing additional evidence for the correctness of the calculated low 39Ar underground production rate. Because this rate is comparable to the detection limit (DL of 5 %modern), subsurface produced 39Ar was therefore neglected in the dating calculations in the Fontainebleau Sands aquifer. Results of the direct measurements of 37Ar and 39Ar in samples from the well IMR. The 37Ar and 39Ar equilibrium concentrations in rocks, calculated from neutron production rates and rock chemical composition, are also shown.

Measured values Calculated values a

Isotope Activity

Equilibrium concentration in rock [atoms/cm3 rock]

Concentration in water [atoms/g water]

Concentration of argon [cm3 STP/g water] Activity in water

37Ar < 0.027dpm/l Ar ~1 · 10-6 4 · 10-9 to 4 · 10-7 10-7 to 10-5 dpm/l Ar

39Ar (55 ± 5) %modern 1.3 0.005 to 0.520

(4.33 ± 0.03) · 10-4 0.05 to 5.5 %modern

a Calculations were made assuming the following data: uranium and thorium concentrations of 0.59 and 1.25 ppm, respectively; escape factors from rocks to the water phase between 0.1 and 10% (Loosli et al., 1991) and a saturated porosity of 25%.

57

Annex 2. Noble gases. The noble gas data were interpreted by fitting various models for noble gas dissolution in

groundwater to the concentrations of the atmospheric gases (Ne, Ar, Kr, Xe) in order to determine the model parameter values that minimize the error weighted deviation (measured by �2, compare eq. 2 and Aeschbach-Hertig et al., 1999). This procedure yielded very good fits with the simplest model, assuming that the excess of the concentrations above solubility equilibrium (excess air) is pure atmospheric air (unfractionated air or UA-model). The low values of χ2 (and correspondingly high values of the probability of finding these values according to a χ2–distribution with 2 degrees of freedom) suggest that the above model describes very well the measured concentrations and that, for example, no degassing in the aquifer or in the samples has occurred.

However, if the derived model parameters are applied for He, and the difference between measured He concentrations and modeled atmosphere-derived He concentrations is interpreted as the radiogenic He component, small but significant negative values are calculated for the radiogenic He in all cases (Table 3). The model predicts between 2 and 6% more helium than the concentrations measured. The problem of apparently negative radiogenic He concentrations is quite frequently encountered in studies applying 3H-3He dating to shallow groundwater, when the atmospheric He is estimated from the measured Ne concentration (Peeters et al., 2002b). The usually adopted solution for the calculation of the tritiogenic 3He component is to assume that the radiogenic He concentrations are equal to zero. The tritiogenic 3He and hence the 3H-3He age can then be calculated solely from the measured He data. The tritiogenic 3He concentrations listed in table 3 and the 3H-3He ages in table 2 were calculated in this way.

The physically senseless result of negative radiogenic He concentrations indicates that the model used to fit the heavy noble gases is incomplete. Such a discrepancy can be explained by a fractionation of the excess air relative to air, as discussed by Peeters et al. (2002b). Two fractionation models are commonly used (see Kipfer et al., 2002 for a review): the CE-model, assuming incomplete dissolution of entrapped air bubbles and closed-system equilibration (Aeschbach-Hertig et al., 2000) and the PR-model, assuming partial diffusive re-equilibration (Stute et al., 1995). Both models can in principle resolve the problem by introducing a fractionation parameter F > 0. However, the fitting procedure based only on Ne, Ar, Kr, and Xe always yields best fits for F = 0, corresponding to unfractionated excess air. Therefore, this approach does not provide a unique and consistent solution.

Adopting the assumption that no radiogenic He is present allows finding a unique and consistent interpretation for all noble gases. With this assumption, He can be treated as a purely atmospheric gas and included in the fitting procedure along with the other noble gases. With He as an additional constraint, the UA-model does not provide acceptable fits (p < 0.01) for 3 of the 7 samples. The CE-model does not perform significantly better. However, the PR-model still yields very good fits, with small values of the fractionation parameter (F < 0.17). The derived NGTs with this fit are slightly (0.04 to 0.20 °C) higher than those derived only from the heavy noble gases (Table 3), but the difference is within the uncertainty of the fitting results. The tritiogenic 3He concentrations calculated with this model differ slightly but not significantly from those calculated from He only (Table 3).

The finding that only the PR-model explains all measured noble gas concentrations is rather surprising, since several previous studies have argued for the use of the CE-model (Aeschbach-Hertig et al., 2000; Peeters et al., 2002a,b). Peeters et al. (2002a) showed that the isotope ratios of Ne and Ar can provide a clear test for the applicability of the PR-model. Therefore, as a final consistency check, the fitting procedure was performed using all five noble gas concentrations and the 22Ne/20Ne (0.102) and 40Ar/36Ar (295.5) ratios as constraints. The PR-model remains the favoured model in this case and still yields a good explanation of all measured data. However, it should be noted that the isotope ratios do not provide strong constraints in the present case of weak fractionation.

The 3Hetri concentrations in groundwater vary between 1.1 10-14 and 5.2 10-14 cm3 He STP per g of water (Table 3). The noble gas temperatures (NGT) range between 9.6 and 10.6 oC, with errors of approximately 0.2 oC, and correspond within the range of measured variation (~1oC) to the present interannual mean air temperature of 11.0 ± 0.6 oC (Meteo France, Trappes station, 1991-2000). Excess air concentrations, expressed as the excess of Ne in the sample compared to the equilibrium concentration (∆Ne) range between 30 and 59 %.

58

59

Chapter 4 Groundwater dating in the Turonian and Cenomanian aquifers of the Bohemian Cretaceous Basin: A first step in getting insights on underground processes and recharge conditions. J.A. Corcho Alvarado1, R. Purtschert1, T. Pačes2, R. Kipfer3, M. Leuenberger1

1 Climate and Environmental Physics Division, Physics Institute, University of Bern, Switzerland 2 Czech Geological Survey, Prague, Czech Republic 3 Water Resources and Drinking Water, EAWAG, Dübendorf, Switzerland (Manuscript in preparation)

Abstract

Environmental tracers (3H, 3He, 85Kr, 39Ar, 14C) are used in a first exercise to get insights

on age and mixing processes of groundwater in the semiconfined Turonian aquifer (TA) and the deeper and confined Cenomanian aquifer (CA) of the Bohemian Cretaceous Basin. The northern section of the Bohemian Basin was under intensive uranium exploitation with a negative impact on groundwater quality of both aquifers. The understanding of the groundwater flow and its age structure are then necessary pieces of information to investigate the future dispersion and impact of the contamination. The resulting mean groundwater residence times in the TA vary from modern to about 200 years. The presence of modern water is correlated with the detection of elevated nitrate concentrations, indicating the vulnerability of the aquifer to surface pollution. The 14C activities decrease along the flow North-South direction in the CA indicating mean groundwater residence times varying from a few hundreds of years to more than 20,000 years. For the radiocarbon dating of the groundwater samples, the measurements of 39Ar are used to identify the initial 14C activity and to estimate its age.

The noble gas temperatures and the stable isotopes signature show that water in one well recharged at cooler temperatures than present day temperature averages. This result agrees with the dating, which predicts that water recharged during the late Pleistocene when climate with lower air temperatures prevailed. 4He and 3He concentrations increase linearly with the age of the water with accumulation rates several orders of magnitude higher than that predicted by the in situ production within the aquifer. The 3He/4He ratios in groundwater suggest that the main source of helium is the earth crust. However, the mantle contribution seems to be more significant in one of the wells. The drill hole is located in an area with volcanic activity in Tertiary period. This volcanic event is known to be precursor of an increased recent flux of CO2 with mantle helium in the Bohemian Massif.

60

1. Introduction

The Turonian and Cenomanian aquifers of the Bohemian Cretaceous Basin are very important sources of high-quality groundwater in the Czech Republic. An intensive withdrawal for public use has been in progress since 1930. The northern section of the Bohemian Basin was under intensive uranium exploitation with negative impact on groundwater quality of both aquifers (Novak et al., 2000). Therefore, the understanding of the groundwater flow and its age structure are required to assess the future dispersion and impact of this contamination.

Periodic groundwater sampling in some wells of the Turonian sandstone aquifer, performed by the Czech Hydrometeorological Institute, have shown a steady increase in nitrate concentrations suggesting anthropogenic contamination from agricultural practices and consequently recent recharged groundwater. The 14C dating of isolated groundwater samples from the confined Cenomanian aquifer resulted in ages of a few thousands of years (Šilar, 1976; Šilar, 1977; Šilar, 1983; Šilar, 1990; Herčík et al., 2003). Considering the above facts, environmental tracers with a wide dating range (3H/3He: Schlosser, 1989; 85Kr and 39Ar: Loosli et al. 1999; 14C: Fontes and Garnier, 1979; and 4He: Solomon, 1999) were measured at different wells along similar flow paths in the region exploited for water supply and located to the east and south of the mining area. Tracers are time dependent and generally only useful in a certain timescale. A multitracer approach is therefore applied to cover the whole range of possible groundwater ages in the investigated aquifers. Moreover, tracers with similar dating ranges are combined (e.g. 3H/3He and 85Kr) enabling other processes such as mixing of different groundwater bodies and dispersion within the aquifer to be investigated (Ekwurzel et al., 1994; Corcho et al., 2005).

The tracer measurements are used in conjunction with hydrochemical measurements for investigating other processes such as chemical evolutions along the flow direction and mineral precipitation and dissolution rates. The knowledge of the water chemistry is also essential for dating groundwater by the radiocarbon method. Furthermore, the noble gases (He, Ne, Ar, Kr and Xe) dissolved in groundwater and the stable isotopes of the water molecule 2H and 18O are applied to investigate the climatic conditions that prevailed at groundwater recharge. The noble gas composition of groundwater is a direct measure of the temperature during groundwater recharge, and has been used worldwide to reconstruct Holocene, glacial and older palaeotemperatures (Andrews, 1993; Beyerle et al., 1998; Zuber et al., 2004). In the other hand, the content of the stable isotopes 2H and 18O in precipitation for mid and high latitudes is strongly linked to the air temperature showing correlations on annual and interannual time scales (Rozanski, 1993). This allows the use of the content in groundwater of both isotopes as proxies for temperature. Additionally, the parameter deuterium excess (d) which is derived from the combination of both isotopes (d=δ2H-8 δ18O) have been used to determine moisture sources, palaeohumidity conditions and variations in the atmospheric circulation patterns (Rozanski, 1985). It should be kept in mind that underground processes such as hydrodynamic dispersion within the aquifer and mixing of different water bodies tend to reduce and smooth the climate signature stored in groundwater; therefore the interpretation of these data must be carried out keeping a special attention to these processes (Rozanski, 1985; Fontes et al., 1993).

2. Study area The Bohemian Cretaceous Basin is the largest sedimentary basin of the Bohemian Massif

in the Eastern part of the European Hercynides. The Basin covers the area of 14,600 km2, of

61

which 12,490 km2 lie within the territory of the Czech Republic (Fig. 1). It has a usual thickness varying between 200 and 400 meters, but in some regions the thickness reaches up to 1100 meters. The sediments accumulated from the early Cenomanian or even in the late Albian to Santonian period. The Bohemian massif was tectonically and vulcanologically reactivated during alpine orogeny. Previous studies in the area have reported a large flux of gases from the mantle (CO2, He, CH4) into the crustal formations (Weinlich et al., 1999; Braeuer et al., 2003; Paces and Smejkal, 2004).

Figure 1. Map of the Czech Republic with the location of the investigated area. The filled triangles

represent the wells sampled in the Turonian aquifer and the filled rhombus the ones in the Cenomanian aquifer. The samples were taken along the flow direction north-south. The numbers in circles are the 14C activities measured in groundwater. The dashed line A-B is a cross-section of the investigated area (see Figure 2).

The flow of groundwater occurs in the large pore and fracture space of the sandstone

layers (a cross section is presented in Figure 2). The sandstone layers are made of up to 99 % of quartz sands. The most important aquifers are the regional deeper artesian aquifer in Cenomanian sandstones and the regional aquifer with a free and partly confined water table in Middle Turonian sandstones. These aquifers are separated by low-permeability marlstone and claystone aquitards. The thickness of the semiconfined Turonian sands aquifer varies between less than 10m in the South and East parts, to about 190 m in the western part. This aquifer has transmissivity values ranging from 0.3-3318 m2/day (Herčík et al., 2003). The main source of recharge to the aquifer is precipitation. The confined Cenomanian sands aquifer is typically 30-80 m thick in the studied area, with an average transmissivity of 48 m2/day (with values ranging from 0.02-148 m2/day) and a mean porosity of about 0.20 (Herčík et al., 2003). The base of the aquifer is formed by impermeable Permian rocks and, in the region Mladá Boleslav by schists and granitic rocks. The Cenomanian aquifer is recharged by infiltrating

62

rainwater in a 1-2km wide zone along the Luzice fault where the sands outcrop (Fig. 1). Groundwater flows from north to South direction.

The wells for our investigation are located in the western section of the Basin, which is the most exploited for water supply. The Stráz tectonic block, where intensive uranium mining took place, is located at the north-western margin of the Bohemian Basin (Fig. 1) and has an area of 240 km2. In this block both aquifers are fully developed and contaminated due to uranium mining by underground acid leaching (Novak et al., 2000). Since the uranium mining finished in 1996, methods of recovery have been applied in order to restore acceptable groundwater conditions in the region (Andel and Pribán, 1996). One of the main goals of this investigation is to assess to which extent the contamination in this area could affect other parts of the aquifers.

Cenomanian sandstones

Turonian sandstones

Turonian calcareous claystones and marlstones

Turonian marlstones

Turonian calcareous claystones, marlstones

Cenomanian sandstones

Turonian sandstones

Turonian calcareous claystones and marlstones

Turonian marlstones

Turonian calcareous claystones, marlstones

A B

Cenomanian sandstones

Turonian sandstones

Turonian calcareous claystones and marlstones

Turonian marlstones

Turonian calcareous claystones, marlstones

Cenomanian sandstones

Turonian sandstones

Turonian calcareous claystones and marlstones

Turonian marlstones

Turonian calcareous claystones, marlstones

A B

Figure 2. Cross section of the investigated area with details of the aquifers. A and B come from the

figure 1 (Hercík, 2003).

3. Methods A total of four wells from the Turonian aquifer and seven wells from the Cenomanian

aquifer were sampled in April 2003. In the Cenomanian aquifer, the investigated wells were selected along appropriate groundwater flow directions, and are located in the region exploited for water supply to the southeast of the mining area. Groundwater samples were analysed for hydrochemistry; stable isotopes of the water (18O and 2H); 3H; carbon isotopes (13C and 14C) and dissolved stable (He, Ne, Ar, Kr, Xe) and radioactive (37Ar, 39Ar, 85Kr and 222Rn) noble gases. Wells were purged for more than one hour prior to sampling. Field measurements of water temperature, pH, dissolved oxygen and electrical conductivity were carried out. These measurements were used to determine when wells were sufficiently purged for sampling.

The main, minor and trace components in groundwater samples were measured by inductively coupled plasma mass spectrometry (ICP-MS) and atomic adsorption spectrometry (FAAS, ETASS) at the Czech Geological Survey (CGS). The 222Rn activities in groundwater were measured at the CGS.

A few thousands of liters of groundwater were degassed in the field, and the gases compressed in steel cylinders. The krypton and argon fractions of these samples were separated and the 37Ar, 39Ar and 85Kr activities counted by Low Level gas proportional counting at the Deep Laboratory of the Physics Institute, University of Bern (Switzerland) (Loosli,1983; Loosli et al.,1986; Foster et al., 1992).

Amber glass bottles of 250 ml were used to sample water for 3H and stable isotopes 18O and 2H. The stable isotope contents were determined by isotope ratio mass spectrometry in aliquots of the samples. An enrichment step was utilized for 3H, and the activity measured by liquid scintillation counting. Both determinations were done in the Physics Institute, University of Bern.

63

Copper tubes were used to sample 45ml of water for stable noble gas analyses. The samples were taken under recommended conditions: copper tubes with external clamps were connected to the point of water withdrawal by flexible plastic tubing secured with hose clamps (gas tight) and the water was flushed by several minutes (until no bubbles were detected in the plastic tube) through the copper tube at high pressures before the steel clamps were closed. The analyses were conducted at the EAWAG laboratories, Zurich, according to the procedure described by Beyerle et al. (2000). The interpretation of the data in order to estimate noble gas temperatures (NGT), excess air components (expressed as the neon excess - ΔNe) and radiogenic helium was achieved following the methods described by Aeschbach-Hertig et al. (1999, 2000).

The total dissolved inorganic carbon contained in 40-100 litres of groundwater was precipitated in the field as BaCO3, and the precipitate collected in 1L plastic bottles. Complete precipitation was assured increasing the pH to 10 or higher by the addition of NaOH. The precipitate was converted to CO2 in the laboratory and the 14C activity measured in CH4 by low level gas proportional counting at the Deep Laboratory of the Physics Institute, University of Bern. The 14C activities are reported as percent of modern carbon (pmC). Additionally, water samples were collected in plastic bottles of 1 liter, and the 13C content of CO2 derived from dissolve inorganic carbon (DIC) analysed by isotope ratio mass spectrometry. The 13C content is expressed as the per mil (‰) deviation from the VPDB standard.

4. Results and discussion 4.1. Hydrochemistry

Firstly we discuss the water chemistry considering its importance for the radiocarbon

dating of the groundwater samples. The results of the chemical determinations in the Turonian and Cenomanian groundwater samples are given in Table 1 listed in order of increasing distance from recharge. The saturation indices (SI=log IAP/K, where IAP is the ion activity product for the reaction and K is the equilibrium constant) of the most important minerals in the aquifer are presented in table 2.

Table 1. Field measurements and groundwater chemistry data. Well T pH Cond. O2 NH4+ Na+ Mg2+ K+ Ca2+ Mn2+ HCO3

- NO3- F- SiO2 SO4

2- Cl- Li+ TDS (oC) (μs/cm) (mg/l) mg/l mg/l mg/l mg/l mg/l μg/l mg/l mg/l mg/l mg/l mg/l mg/l μg/l mg/l

Turonian aquifer VP7523 9.6 7.33 524 1.7 <0.02 1.66 2.2 1.0 108.3 5 286.8 <0.3 0.1 7.0 26.9 5.8 2 443VP7512 11.3 7.41 448 2.7 0.04 2.92 6.8 1.1 82.2 10 247.1 5.7 0.1 6.5 15.1 5.6 2 377VP7524 9.9 7.23 740 5.8 <0.02 4.99 16.0 1.6 130.9 <5 311.2 22.2 0.1 9.2 77.4 18.8 4 600VP7520 11.7 7.21 928 0.2 0.13 5.38 17.8 2.6 165.7 15 311.2 6.4 0.2 6.9 151.0 44.3 23 716

Cenomanian aquifer VP7502 12.0 7.12 441 1.2 0.3 3.60 12.9 4.4 64.5 468 253.2 0.9 0.2 11.6 3.1 3.0 36 361VP7506 11.7 7.91 331 0.2 <0.02 1.80 5.0 1.0 59.3 90 201.4 <0.0 0.1 <2.0 0.3 1.1 5 271VP7500 12.0 7.63 362 0.2 0.1 6.10 10.2 4.6 48.1 74 210.5 <0.3 0.3 2.5 0.3 1.9 15 286VP7515 17.3 6.63 320 0.2 <0.02 14.4 5.9 2.4 35.0 104 125.1 <0.3 0.1 8.0 20.6 14.8 16 227VP7517 18.8 6.76 284 0.2 0.1 19.9 5.1 3.0 23.0 94 100.7 <0.3 0.1 8.0 21.5 15.9 23 198VP7519 14.7 6.85 345 0.2 0.2 29.9 8.0 5.2 22.1 54 119.0 <0.3 0.1 7.5 25.7 22.9 43 241Káraný B 11.6 6.48 509 0.2 <0.02 41.0 13.2 9.3 34.6 33 177.0 0.9 0.4 6.9 54.6 26.9 84 225

4.1.1. Turonian sandstone aquifer

The groundwaters present Ca-HCO3 hydrochemical facies, with total dissolved solids

varying between 370 and 720 mg l-. The water temperatures range from 9.6 to 11.7 oC, and

64

the pH from 7.2 to 7.4 (Table 1). The water chemistry is dominated by calcium and bicarbonates ions; however, elevated concentrations of nitrate, sulphate, chloride, sodium and potassium were measured in the two shallower wells (VP7524 and VP7520; Table 3). The water conductivities follow the same trend with higher values in the two shallower wells. The semiconfined Turonian aquifer is located in an area where the land is mainly used for agriculture and the input of water contaminated with fertilizers increases the concentrations of dissolved components. This result shows the vulnerability of the upper part of the aquifer to surface contamination. Also, the vertical distribution of dissolved substances indicates a possible stratification of the aquifer. Table 2. Mineral saturation indices (SI) for groundwaters of the Turonian and Cenomanian aquifers. SI were

calculated with WATEQ4F (Ball and Nordstrom, 1991). (SI=log IAP/K, where IAP is the ion activity product for the reaction and K is the equilibrium constant).

Well Calcite Dolomite Aragonite Gypsum Chalcedony SiO2 Quartz Turonian aquifer

VP7523 -0.07 -2.33 -0.22 -1.95 -0.19 -1.09 0.29 VP7512 0.07 -1.42 -0.09 -2.29 -0.25 -1.14 0.23 VP7524 -0.04 -1.48 -0.19 -1.49 -0.08 -0.97 0.40 VP7520 0.15 -1.13 -0.01 -1.16 -0.23 -1.11 0.25

Cenomanian aquifer VP7502 -0.30 -1.74 -0.45 -3.07 -0.01 -0.89 0.47 VP7506 -0.11 -1.74 -0.26 -4.07 VP7500 0.10 -0.92 -0.05 -4.17 -0.67 -1.56 -0.20 VP7515 -0.90 -2.91 -1.05 -2.44 -0.23 -1.10 0.22 VP7517 -1.39 -3.74 -1.54 -2.58 -0.25 -1.11 0.20 VP7519 -1.66 -4.15 -1.81 -2.54 -0.23 -1.10 0.23 Káraný B -0.86 -2.60 -1.02 -2.08 -0.23 -1.11 0.25

Calcite, CaCO3; Dolomite, CaMg(CO3)2; Aragonite, CaCO3; Gypsum, CaSO4; Chalcedony, SiO2; Quartz, SiO2.

4.1.2. Cenomanian sandstone aquifer

The chemistry analyses revealed that groundwaters are low mineralized, with TDS

varying between 190 and 370 mg l-. The pH varies between 7.91 and 6.48, and the water temperature between 11.6 and 18.8 oC. The chemical composition of groundwater depends on the duration of water-rock interaction. Hence, the water changes from Ca-HCO3 type in the recharge area, with apparent ages of a few hundreds of years, to a more saline (Na-Ca-Mg-HCO3-SO4-Cl) chemical type at the end of the flow path, with ages of a few thousands of years. Groundwater undergoes a chemical evolution as a result of a number of different geochemical processes.

The calculation of the saturation indices (WATERQ4F, Ball and Nordstrom, 1991) show that (Table 2): a) all the samples are saturated with respect to quartz and chalcedony (SiO2 is the main mineral in the rock matrix with more than 98% of the total weight, Table 4); b) groundwaters are at or near equilibrium with respect to calcite in the recharge area, and tend to be more undersaturated in downstream wells; c) the gypsum saturation index increases with distance of flow but remains undersaturated; and d) the dolomite saturation index decreases along the flow path with values below saturation in all the samples. The main chemical reactions controlling the water chemistry are:

Calcite dissolution-precipitation

CaCO3 + H2O + CO2 <-> Ca2+ + 2HCO3-

65

Ion exchange Ca2+ (Mg2+) <-> 2Na+

Dolomite dissolution CaMg(CO3)2 + 2CO2 + 2H2O <-> Ca2+ + Mg2+ + 4HCO3

- Gypsum dissolution

CaSO4.2H2O <-> Ca2+ + SO42- + 2H2O

The dissolution of residual rock-forming minerals and secondary salts in sandstones

increases progressively along the flow the concentration of ions such as Na, K, SO4 and Cl (Fig. 3, Table 1). Dissolution of calcite and very small amounts of dolomite dominate the water chemistry in the recharge area. But dissolution of gypsum (CaSO4.2H2O) after approximately 12 km distance from recharge, where flow probably contacts gypsum minerals, forces calcite precipitation according to the common ion effect decreasing the concentration of HCO3 in groundwater. The trend towards higher amounts of Na and lower amounts of Ca indicates that ion exchange is another important geochemical process in the aquifer; although dissolution of sodium feldspars and other silicate minerals are responsible for a considerable increase in the Na concentration. The Ca-Na exchanges in combination with calcite precipitation, explain the trend with distance from recharge to groundwater more undersaturated with respect to calcite (Table 2).

0 10 20 30 40 50 60 700

1

2

3

4

5

0 10 20 30 40 50 60 700.0

0.5

1.0

1.5

2.0

Mixture

VP7517VP7519

Karany B

VP7515

VP7500

VP7506

VP7502 HCO3-

Ca2+

Mg2+

[mm

ol/L

]

Distance from recharge (km)

VP7517

VP7519

Na+

Cl-

SO42- Karany B

VP7515

VP7500

VP7506

VP7502

[mm

ol/L

]

Distance from recharge (km)

Figure 3. Evolution of the concentrations of the major dissolved ions along the flow direction in the Cenomanian aquifer.

Precipitation of calcite along the flow favours the dilution of heavier isotopes (13C and

14C) in DIC (Fig. 4). The dissolution in groundwater of mantle CO2 (dead in 14C and enriched in 13C) in the area of location of the well VP7502 is responsible for the shift observed in the contents of the isotopes 13C and 14C in DIC (Fig. 4). In Karany B, which is located at the end of the flow path, mixing of two different water components is the process that shifted the 13C and 14C contents in DIC from the expected location in the trend lines (Fig. 4). Both cases are examined in more detail in the groundwater dating section. The geochemical processes above discussed can explain most of the variations along the flow direction in the 14C activity (together with radioactive decay) and in the δ13C content in DIC.

66

50 100 150 200 250 3000

10

20

30

40

50

60

50 100 150 200 250 300-15

-14

-13

-12

-11

-10

-9

-8

Mixture

Input of mantle CO2

End of flowpath

RechargeareaVP7506

VP7500Karany B

VP7515

VP7517

VP7519

VP7502

14C

(pm

C)

[HCO3-] (mg/L)

Mixture

Input of mantle CO2

VP7500

VP7506

Karany B

VP7519

VP7517

VP7515

VP7502

End of flow path

Rechargearea

d13C

(‰)

[HCO3-] (mg/L)

Figure 4. a) 14C activity and b) δ13C content in DIC against the bicarbonate concentration measured in

groundwater samples from the Cenomanian aquifer. The input of mantle CO2 (dead in 14C and enriched in 13C) into groundwater shifts the sample VP7502 from the trend lines. Groundwater from Karany B is a mixture of two different water components, an old component with a lower concentration of HCO3 and a young component with a higher concentration of HCO3.

4.2. Groundwater dating The results of the tracer measurements are given in table 3. These data are used to

investigate the age of groundwater in both aquifers as discussed below. 4.2.1. Turonian sandstone aquifer

High activities of 3H, between 8.7 and 10.5 TU, and of 85Kr, between 12.0 and 39.6

dpm/cm3Kr, were measured in two (VP7524 and VP7520) of the four investigated wells (Fig. 5, Table 3). These high activities coincide with the presence of signs of anthropogenic contamination, as was already discussed in the hydrochemistry section. The tritiogenic 3HeTrit (produced by decay of bomb 3H) contents in groundwater from these wells, which are derived from the interpretation of the noble gas data according to the methods presented by Aeschbach-Hertig et al. (1999, 2000), vary between 1.5 and 11.0 TU (Table 4). The comparison of the 85Kr piston flow ages (calculated using as input function the atmospheric 85Kr concentrations measured at the Institute of Atmospheric Research in Freigurg, Germany) with the 3H/3He ages resulted in large disagreements. This is not surprising considering that other important processes like tracers transport through the thick unsaturated soil zone over the Turonian aquifer (thickness of 10-70 m, Table 3), degassing of 3HeTri in the unsaturated soil zone and mixing of waters in the large sampled screens (screen lengths of 15-70 m) were not included in the interpretation.

In order to appropriately interpret the tracer data (3H, 3HeTri and 85Kr) in both wells, the correct input functios of 3H and 85Kr at the water table are calculated with a one-dimensional advective-diffusive-decay transport model that models the tracers transport through the unsaturated zone. A similar approach was used successfully in previous studies (Rueedi et al., in press; Corcho Alvarado et al., submitted). For the modelling, homogeneous physical conditions throughout the unsaturated soil zone and the following soil parameters were assumed: a total porosity of 0.2, a water filled porosity of 0.1 and a recharge rate of 150 mm per year (Herčík et al., 2003); a dispersivity of 0.1 m (Cook and Solomon, 1995); a gaseous

67

phase tortuosity of 0.6 (Millington, 1959) and a liquid phase tortuosity of 0.25 (Barraclough and Tinker, 1982). The thickness of the unsatuared zone is constrained based on the depth of the water table at the well locations (Table 3). The 85Kr atmospheric activities at Freiburg (Institute of Atmospheric Research, Freiburg, Germany) and the 3H fallout in Cracow (constructed averaging the fallout data reported for the station located in Cracow, data taken from IAEA/WMO GNIP database, 2004) were used as atmospheric concentrations.

Finally, the calculated tracer concentrations above the water table are used as input functions for the lumped parameter models (Zuber and Maloszewski, 2001). The best parameters that fit these models to the measurements (3H, 3HeTrit and 85Kr) were determined by inverse modelling minimizing a χ2 function (Purtschert et al., 1999; Corcho et al., 2005). An admixture of 35±4 percent of old groundwater in recent infiltrated groundwater was estimated with the dispersion model for a a high dispersion parameter in the well VP7524. The calculated mean residence time of the younger groundwater component in this well is 4±1 years. The age of the old component was determined with the tracer 39Ar, resulting in a value of 280±50 years. In the well VP7520, the groundwater follow an exponential age distribution with a mean residence time of 18±4 years. This mean age is comparable within the error to the exponential age derived from the measured 39Ar activity of 92±5 %modern (exponential mean residence time: 34±21 years), which further confirms the calculated parameters.

In the wells VP7523 and VP7512, the groundwater ages were determined based on the 39Ar data, resulting in values up to 200 years if piston flow is assumed (Table 3). If an exponential age distribution is considered, mean residence times between 68 and 258 years are calculated.

Table 3. The results of the measurements of the activities of 3H, 85Kr, 39Ar, 14C and the stable isotopes (13C, 2H, 18O) composition in groundwater samples from the Turonian and Cenomanian aquifers are shown. 3H/3He, 85Kr and 39Ar ages were calculated assuming piston flow conditions, and the 14C ages with the NETPATH code.

Well Water table depth

Screen depth

3H 3H/3He age

85Kr 85Krage

39Ar 39Ar age

14C 14C age

δ 13C 222Rn δ 18O δ 2H d-excess

(mbgl) (mbgl) [TU] (yrs) [dpm/cm3 Kr] (yrs) [%modern] (yrs) [pmC] (yrs) ‰ SMOW [Bq/l] ‰ SMOW ‰ SMOW ‰ SMOWTuronian aquifer

VP7523 89 100-200 ≤DL >50 0.5±0.3 >50 85±5 63±23 73.4±0.2 -1761 -12.7±0.2 3.2 -10.04 -71.08 9.2 VP7512 6 70-200 ≤DL >50 0.9±0.2 >50 60±5 198±32 63.8±0.3 -1094 -13.0±0.2 3.1 -10.08 -72.59 8.1 VP7524 9 11-29 8.7±0.8 3 39.6±3.2 9 82±5 77±24 78.8±0.3 -2262 -12.9±0.2 23 -10.05 -73.07 7.3 VP7520 58 60-130 10.5±0.8 13 12.0±1.3 20 92±5 32±21 63.9±0.2 -762 -13.5±0.2 14 -9.54 -69.31 7.0

Cenomanian aquifer VP7502 109-200 ≤DL >50 0.8±0.2 >50 196±6 Modern 16.3±0.1 10630 -8.4±0.2 6.4 -10.00 -69.62 10.4 VP7506 240-275 ≤DL >50 1.2±0.3 >50 21±4 600 53.8±0.4 600 -12.0±0.2 9.8 -9.77 -66.74 11.4 VP7500 355-404 ≤DL >50 0.2±0.1 >50 <8 >1200 45.8±0.2 965 -10.6±0.2 2.2 -10.01 -68.75 11.3 VP7515 345-388 ≤DL >50 1.4±0.5 >50 10±5 >1200 24.0±0.3 7080 -14.1±0.2 3.9 -10.25 -71.15 10.8 VP7517 329-384 1.9 >40 NS NS 14.2±0.2 11565 -13.0±0.2 2.8 -10.69 -74.95 10.6 VP7519 207-241 ≤DL >50 NS NS 6.3±0.3 17725 -11.9±0.2 3.6 -11.48 -81.45 10.4

Karany B <150(?) ≤DL >50 0.1±0.1 >50 38±3 376 40.8±0.2 25600a) -12.7±0.2 5.1 -10.02 -72.14 8.0 a) Minimum 14C age of the old component after correction for mixing. mbgl: meters below ground level.

The 14C activities in DIC vary between 63 and 79 pmC. These activities were interpreted

in terms of 14C ages with the Fontes-Garnier model (Fontes and Garnier, 1979) assuming: a) an activity of 14C in soil CO2 of 100 pmC, b) an activity of 0 pmC for 14C in calcite from the rock matrix, c) a δ13C in soil CO2 of -25 ‰, and d) a δ13C in calcite from the rock matrix of 2 ‰ (measured value). The resulting 14C ages are given in table 3, and they are negative in all cases. This result suggests a very recent origin for the groundwaters, which may still contain

68

an input of 14C from thermonuclear weapon tests in some wells, in agreement with the interpretation of the other tracers (3H, 3HeTrit, 85Kr and 39Ar).

The groundwater mean residence times show a tendency to increase with an increase of the depth of the sampled screen below the water table (older ages are found in wells with deeper screens: VP7523 and VP7512). This result possibly shows age stratification in the aquifer. Table 4. Noble gas data. The results of the interpretation of these data with the unfractionated air model are also

shown. Well He

(10-8) 3He/4He (10-7)

Ne (10-7)

Ar (10-4)

Kr (10-8)

Xe (10-8)

χ2 P (%)

NGT (oC)

Excess air (ΔNe)

3HeTrit (TU)

4HeRad+Terrig (10-8)

22Ne/20Ne

40Ar/36Ar

Turonian Aquifer VP7524 5.14±0.03 14.4±0.1 2.20±0.02 4.02±0.02 9.5±0.1 1.37±0.02 0.3 85.9 8.94±0.08 7.0 1.5 0.04 0.102 296.1 VP7520 5.54±0.07 17.1±0.1 2.17±0.02 4.15±0.03 10.3±0.1 1.50±0.03 6.7 3.5 6.56±0.43 2.7 11.0 0.63 0.102 295.8

Cenomanian aquifer VP7502 16.9±0.2 7.8±0.1 3.06±0.02 4.87±0.03 11.1±0.1 1.55±0.03 7.6 2.3 5.19±0.49 43.5 11.0 10 0.102 296.0 VP7500 132±2 9.6±0.1 2.99±0.02 4.74±0.04 11.0±0.1 1.59±0.03 0.1 94.4 5.44±0.06 40.3 469 125 0.102 295.8 VP7515 335±4 10.6±0.1 3.10±0.02 4.68±0.04 10.7±0.1 1.61±0.03 4.8 9.0 6.29±0.39 46.4 1383 328 0.102 295.6 VP7519 934±12 75.6±0.1 4.89±0.04 6.25±0.05 13.8±0.2 1.94±0.03 3.3 19.6 0.79±0.33 117.5 28305 921 0.102 295.9 Karany B 1580±16 7.7±0.1 2.91±0.03 4.64±0.05 10.6±0.1 1.50±0.02 4.8 9.2 6.43±0.33 38.7 4836 1570 0.102 296.4

4.2.2. Cenomanian sandstone aquifer

Modern tracers (bomb 3H or 85Kr) were not detected in the investigated part of the Cenomanian aquifer, indicating that groundwater ages must be older than 50 years. 39Ar activities were detected only in the northern recharge area of the aquifer, and are equivalent to mean groundwater ages of up to 600 years. Hence, groundwater from the wells located farther downstream should have residence times older than 1,000 years. A decreasing trend along the flow direction is observed for the activities of 14C from 54 to about 6 pmC (Table 3, Fig. 6). This tendency indicates mean groundwater ages between a few hundreds and thousands of years.

For the radiocarbon dating of the groundwater samples it is important, first, to understand the geochemical processes that occur in the aquifer and that can affect the 14C activity in DIC; and second, to identify the initial 14C activity in groundwater used to derive the radiocarbon age. The first point was already discussed in the hydrochemistry section. For the second point, we propose the use of the measurements of the radioisotope 39Ar (half-life of 269 years) to identify the initial activity of 14C, and to calculate the age associated to this activity.

For the selection of the initial 14C activity, the wells located in the beginning of the flow path are analysed. The trend in the 14C activity suggests that the composition in the well VP7502 (Table 3, Fig. 6a) is not representative of the input water for wells located further downstream. From the trend line observed in figure 6a, a 14C activity of about 55-60 pmC is guessed for groundwater in VP7502; however the measured activity is as low as 16.3 pmC. This well is either not located in the main recharge area or/and as a result of one or more events in the last 600 years (timescale derived from the 39Ar age in the well VP7506, which is located further down stream; see Table 3) considerable amounts of dead carbon were introduced in the form of CO2 by upward flow through the Luzice fault. The last hypothesis is in agreement with results of previous studies which confirmed an upward flux of large amounts of mantle CO2

(enriched in 13C) through fault zones in the Bohemian massif (Weinlich et al., 1999; Braeuer et al., 2003; Paces and Smejkal, 2004). This flux would also explain the enriched 13C value measured in DIC of this sample (Table 3, Fig. 4).

69

0 50 100 150 200 250 3000

10

20

30

40

50

60

0 10 20 30 40 50 60 700

10

20

30

40

50

60

Assumed end member (Water Age > 20 000 yrs)

Final water with mixed gases

Composition of gases from deeper formations ([39Ar]=230-300 %Modern, and dead CO

2)

Decay?

Assumed initial water(Age ~200 yrs)

VP7502

Karany B

VP7500

VP7515

VP7506

Mixing lines

Decay curve

14C

(pm

C)

39Ar (%Modern)

Input of dead CO

2

Guessed initial activity for VP7502 (ca.55-60 pmC)

Mixture

VP7506

VP7500

VP7515

VP7517

VP7519

Karany B

VP7502

14C

in D

IC (p

mC

)

Distance from recharge (km) Figure 6. a) 14C activities versus the distance of the well from the recharge area. The probable shift of

the 14C activity in the well VP7502 produced by the input of dead CO2 is represented (An initial activity of 55-60 pmC is guessed from the trend observed in the rest of the wells). b) 14C activities versus 39Ar activities measured in groundwater from the Cenomanian aquifer. The dashed lines represent different mixing lines between different end-members. The solid line represents a decay curve from an assumed initial groundwater with an age of about 200 years (39Ar activity of 59 %modern, and a 14C activity of about 58 pmC).

Furthermore, the upward flux of gases would as well explain the high 39Ar activity (196

%modern) measured in groundwater from this well, which cannot be explained by underground production within the Cenomanian sands aquifer. A maximum contribution of the in situ produced 39Ar to the total activity of 39Ar in groundwater of 1 %modern is calculated from the aquifer data (U=7 ppm, Th=19 ppm, K=1600 ppm, Porosity=20 % and an assumed Ar fractionation loss coefficient of 1 %) (Table 5 and 6, Fig. 6b). A detailed description of the methodology to calculate the underground production of 39Ar can be found in Loosli and Lehmann (1991) and Loosli et al. (1999). The calculated low in situ production of 39Ar within the aquifer is supported by the non detection of this radioactive noble gas in wells located downstream and by the negligible activities of 37Ar (Table 6). Additionally, the relatively low 222Rn activities in groundwater (< 10 Bq/L) reflect the low content of radioisotopes of the natural decay series in the aquifer (Table 3). The in situ production of 39Ar within the aquifer has been acknowledged as one of the main limitations of the 39Ar dating method (Foster et al., 1992); however this process can be neglected in the Cenomanian Sands. In deeper formations much higher potassium concentrations (ca. 36,000 ppm) were measured (Herčík et al., 2003). An 39Ar equilibrium activity of 300%modern is estimated assuming a similar rock composition as in the aquifer. Considering that the half life of 39Ar is 269 yrs, a relatively fast upward transport of gas through the Luzice fault is required to explain the high enrichment in 39Ar in VP7502. A range of 39Ar activities in groundwater of this well, after mixed with underground produced 39Ar, is presented in figure 6b. Table 5. Main, minor and trace rock composition of the Cenomanian and Turonian aquifers.

Aquifer SiO2 TiO2 Al2O3 Fe2O3 MnO MgO CaO SrO Na2O K2O P2O5 CO2 Ua Tha B Cr Pb Sr Ti V Zr % % % % % % % % % % % % ppm ppm ppm ppm ppm ppm ppm

Cenomanian Sandstone

98.28 0.14 0.9 0.21 0.001 0.03 0.03 0.002 0.04 0.19 0.05 0.01 4 (7) 12(19) 35 29 9 9 850 9 100

Turonian Sandstone

1 10

a Values in parenthesis are the maximum concentrations measured.

70

The well VP7506, which is located further downstream from VP7502, is then taken as the initial well (up gradient well) for the calculation of the radiocarbon ages of the DIC. This well shows an 39Ar activity of 21 %modern which corresponds, assuming piston flow, to a mean groundwater age of 600 years.

We use the computer model NETPATH (Plummer et al., 1991) to perform the geochemical modellings between the selected up gradient well VP7506, representative of the initial conditions, and the down gradient wells. The computer code corrects for the geochemical reactions occurring along the flow path that affect the 14C activity in DIC and provides the travel time from the initial to the final well. In the model, the geochemical reactions are constrained to take place between realistic reactant and product minerals, which are consistent with the observed chemical and isotopic data of the groundwater (Tables 1, 2 and 3). The isotopic data from the water samples used in the modelling are summarised in Table 3. For the solid phases the following values are used: a) 2.0±1.1‰ for the δ13C in Cenomanian carbonates (an averaged value obtained from several determinations), b) 0 pmC is assumed for the 14C activity in the aquifer carbonates. All reactions observed were constrained by mass balance on Ca, Mg, Na, C, S, Cl and K. The validity of the models was checked with the observed δ13C of DIC.

Table 6. Results of the direct measurements of 37Ar and 39Ar in two groundwater samples. The 37Ar and 39Ar equilibrium concentrations in rocks (calculated from neutron production rates and rock chemical composition), the correspondent concentrations in water and the activity in water derived are shown.

Calculated values a)

Well Isotope

Measured values

Equilibrium concentration in rock [atoms/cm3 rock]

Concentration in water[atoms/g water]

Concentration of argon [cm3 STP/g water]

Activity in water

37Ar (8±3) 10-3 dpm/L Ar 7 10-6

4 10-7 1 10-5

dpm/L Ar VP7500 39Ar < 8

%modern 1.9 0.096

(4.7 ± 0.1) · 10-4 0.9

%modern 37Ar (17±3) 10-3

dpm/l Ar 7 10-6 4 10-7

1 10-5 dpm/l Ar Karany B

39Ar 38 ± 3 %modern 1.9 0.096

(4.6 ± 0.1) · 10-4 0.9

%modern a) Calculations were made assuming the following data: uranium and thorium concentrations of 7 and 19 ppm,

respectively; an escape factor from rocks to the water phase of 1% (Loosli et al., 1991); and a saturated porosity of 20%.

The calculated relative ages between the initial well and the downstream wells are then

converted to absolute ages by calibrating the age of groundwater in the initial well (VP7506) with the radioisotope 39Ar. Finally, the adjusted 14C ages vary from 600 years in the recharge area to more than 20,000 years in the end of the flow path (Table 3).

The well Karany B has a screen interval several hundreds of meters shallower than the rest of the wells (Table 3). An increased portion of the younger water components is then very probable in this well. This is also supported by the elevated 39Ar and 14C concentrations compared to the wells located further upstream. The range of ages of the water components in the mixture can be constrained based on the tracer data. According to the location of the well in the flow path we can exclude any 39Ar in the older water component; therefore the total activity of 39Ar measured in groundwater belongs to the younger water component. Based on this fact, an age for the younger component between 50 (pre-modern recharge because no modern tracers were detected in the well) and 400 years (maximum age calculated when we assume 100% of young water) is calculated. This would mean that the percent of the younger water component in the mixture could vary from 40 to 100%. But, if we take into account the 14C data, this percent range can be further constrained. Assuming for the younger water component a chemical composition similar to the ones measured in the Turonian samples and

71

using the Fontes-Garnier model (Fontes and Garnier, 1979) under the same conditions assumed for the Turonian samples, an initial 14C activity of 66 pmC is estimated. Using this value for the younger end member and assuming a 14C activity of 0 pmC for the older end member (a value that can be derived from the trend seen in Figure 6a), the percent of the younger water component is then calculated to be between 60 and 65. A similar mixing ratio (0.6-0.7) can be derived from the mixing line presented in Figure 6b which connect two assumed end groundwater members: a) an older water component tracer free (no 14C and 39Ar), and b) a younger component with an age of about 200 years (39Ar=60 %modern, and 14C=58 pmC). Based on the above calculations we predict an age for the older groundwater component of minimum 20,000 years. 4.3. Investigating underground processes and recharge conditions 4.3.1. Groundwater flow velocity in the Cenomanian aquifer

The wells in the confined Cenomanian sands aquifer are located along the same flow line.

This makes possible an estimation of the groundwater flow velocity base on the tracer dating. A good linear correlation is obtained between the 14C age and the distance from the recharge area (Fig. 7), which shows that piston flow is a good approximation for the type of flow in this confined aquifer. An average groundwater flow velocity of 2.3 m/y is calculated from the linear correlation. This velocity stays nearly constant along the entire flow path, which indicates that the sands have, in average, homogeneous hydraulic properties. As a first application of this information, we attempt to estimate the age of the old groundwater component in the well Karany B, which is used as a water supply for Prague and is located at 64.3 km from the recharge area. An age of approximately 25,000 years is predicted based on the relationship presented in Figure 7. This value agrees remarkably well with the range calculated based on the tracer data.

0 5000 10000 15000 20000 25000 300000

10

20

30

40

50

60

70

Karany B

VP7519

VP7517VP7515

VP7500

VP7506

(km) = 0.0023 * (Age) + 7.4r=0.95

Dis

tanc

e fro

m re

char

ge (k

m)

14C age (yrs.) Figure 7. Distance from the recharge area as a function of the 14C ages in the Cenomanian aquifer. The

slope of the linear fit corresponds to a flow velocity of 2.3 m/year. The dashed circle represents the location of the well Karany B deduced from the correlation line and the distance of the well from recharge.

72

The groundwater flow velocity was as well calculated based on reported hydraulic data to be compared with the velocities estimated from the tracer dating. A mean transmissivity value of 48 m2/d (transmissivity varies between 0.02 and 181 m2/d), a mean aquifer thickness of 60 m (thickness varies between 30 and 80 m), a mean hydraulic gradient of 0.002 and a mean porosity of 20 % were used for the calculation (data taken from Hercík et al., 2003). The estimated mean groundwater flow velocity is 2.7 m/y, but it can vary in a large range if we consider the large range of variation of the hydraulic parameters. This mean velocity is in relative good agreement with our estimated velocity based on the tracer dating (2.3 m/y).

4.3.2. Mineral dissolution rates. Chemical tracers as time indicators The chemical signatures in groundwater may be used to provide information about the

timescales of mineral dissolutions and precipitations. For example, if we assume that the dissolved ions behave conservatively, then the mineral dissolution rates along segments of the flow path can be estimated plotting the amount of ion dissolved in groundwater against the groundwater age (Fig. 8). Hence, in the Cenomanian sands aquifer the dissolution rate for chloride bearing minerals is 3.5 10-5 moles per kg of water per year and for the sulphate bearing minerals the rate is 1.6 10-5 moles per kg of water per year. The intercept of the correlation lines in figure 8 reflects most probably the concentrations of Cl and SO4 at groundwater recharge.

As was introduced in an early work by Edmunds and Smedley (2000), some dissolved elements in groundwater can also be used to derive concentration-age relationships, which allow extending the range of the groundwater dating beyond the 14C method up to several hundreds of thousands of years. This application is limited to reactions that do not reach either equilibrium or partial equilibrium states in the aquifer, and to sites where other processes that could affect the interpretation like mixing can be excluded. The relationships obtained in figure 8 can then be used to predict the age of groundwater in other parts of this aquifer.

0 10000 20000 300000.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

VP7506

VP7500

VP7515

VP7515

VP7517

VP7517

VP7519

VP7519

[SO4] = 1.6E-5 * (Age) + 0.01

[Cl] = 3.5E-5 * (Age) + 0.04

Amou

nt o

f ion

in g

roun

dwat

er [m

mol

/l]

14C age (yrs) Figure 8. Concentrations of Cl (■) and SO4 (▲) in groundwater from the Cenomanian aquifer as a

function of the 14C age. The well Karany B was excluded from the correlation lines considering the occurrence of mixing.

73

4.3.3. 3H and 4He in groundwater from the Cenomanian aquifer The 4He and 3He concentrations increase linearly with the age of the water in the

Cenomanian aquifer (Table 4, Fig. 9). The concentrations are enriched by a factor of up to 300 compared to water in solubility equilibrium with the atmosphere (Table 4). 4He accumulates in groundwater at a rate of 4.8 10-10 cm3STP/cm3y, and 3He at a rate of 3.8 10-15 cm3STP/cm3y (3.8 10-16 cm3STP/cm3y if the sample VP7519 is excluded from the correlation) (Fig. 9).

0 5000 10000 15000 20000 25000 30000 35000

2x10-11

4x10-11

6x10-11

8x10-11

0 5000 10000 15000 20000 25000

2.5x10-6

5.0x10-6

7.5x10-6

1.0x10-5

[3He] = 3.8E-16 * (Age) - 4.2E-13VP7500VP7515

VP7519

[3He] = 3.8E-15 * (Age) + 7.7E-12

[3 He]

(cm

3 STP/

gwat

er)

14C age (yrs)

VP7502

In situ accum. rate (calculated) 2E-11 cm3HeSTP/cm3water/yr

[4He] = 4.8E-10 * (Age) + 1.6E-7

VP7500

VP7515

VP7519

VP7502

[4 He]

(cm

3 STP/

cm3 w

ater

)

14C age (yrs.) Figure 9. Linear correlation between the 4H concentrations and the 14C ages (dashed line) measured in

groundwater samples from the Cenomanian aquifer. The solid line represents the accumulation of the in situ produced 4H (Cenomanian aquifer). b) Linear correlations between the 3H concentrations and the 14C ages in the Cenomanian aquifer. Two correlations are shown, the first one where the sample VP7519 is taken into account and the second one where it is neglected.

In order to investigate whether the helium excess observed in the samples results from in

situ production or from an external source, the in situ production rates of 4He and 3He were calculated as follows:

( ) [ ] [ ]( )ThUHeP ⋅⋅+⋅⋅= −− 14134 109.2102.1 (cm3 STP per g of rock per year)

( ) [ ] [ ]( ) [ ]LiThUHeP ⋅⋅+⋅⋅= − 4.14.610 233 (cm3 STP per g of rock per year)

where P(3,4He) are the production rates in the rock. [U], [Th] and [Li] are the concentrations of uranium, thorium and lithium, respectively (in ppm, Table 4). Then, the accumulation rates were calculated according to the expressions:

( ) ( ) ( )4

44 1

Λ⋅⋅−⋅

=water

rockHePHeAρρ

θθ (cm3 STP per cm3 of water per year)

( ) ( ) ( )3

33 1

Λ⋅⋅−⋅

=water

rockHePHeAρρ

θθ (cm3 STP per cm3 of water per year)

where A(3,4He) are the accumulation rates in water; θ is the aquifer porosity which has a value of 0.2; ρ are the rock (2.6 g/cm3) and water (1 g/cm3) densities; and Λ3,4 are the transfer efficiencies of 3H and 4He from the rock matrix to the water. For sedimentary formations, it

74

can be assumed that Λ=1 (Torgersen, 1980). The in situ accumulation rates deduced from the above formulae are 2 10-11 and 10-21 cm3/cm3yr for 4He and 3H, respectively. These values of accumulation rates cannot account for the high values of accumulation rates observed in the aquifer. Thus, external sources of 3He and 4He (e.g. crust, mantle) have to be considered to explain the observed rates. This would mean that a vertical flux of 3He and 4He from deeper layers is present in the investigated area.

Typical crustal 3He/4He ratios range from 2 to 7 10-8, while mantle 3He/4He ratio is about 10-5. Using this information, we attempt to distinguish and quatify the different sources of helium in the Cenomanian groundwaters. As concluded previously in this section, the atmospheric components of He in all the groundwater samples are negligible compared to the total measured He concentrations. Therefore, the measured ratio must be dominated by the crustal and mantle components. Considering the typical ratios found in the crust and in the mantle, the measured ratio in groundwater can then be approximated by,

meas

crustmantle

meas

mantlecrust

measmeas He

HeHeHe

HeHeHeHeR 4

4845

4

33

4

3 10210 ⋅⋅+⋅=

+=⎟⎟

⎠

⎞⎜⎜⎝

⎛=

−−

where the subscript meas refers to measured, and crust and mantle to the He components derived from the crust and mantle, respectively. Then, considering that the mantle component is (a) percent of the total measured concentration, the above equation can be modified to,

( )meas

measmeasmeas He

HeaHeaR 4

4845 110210 ⋅−⋅⋅+⋅⋅=

−−

This equation can finally be simplified and resolved for (a), which results in,

5

8

1010

−

−−= measRa

Based on this equation, the He component from mantle origin in groundwater was estimated. It can be concluded that the main external source of helium are crustal rocks where helium is produced by radioactive decay of U and Th; but the results indicate a substantial contribution of mantle helium. The drill holes are located in an area with volcanic activity in Tertiary period. This volcanic event is known to be precursor of an increased recent flux of CO2 with mantle helium in the Bohemian Massif (Weinlich et al., 1999; Braeuer et al., 2003 and Paces and Smejkal, 2004). The contribution of mantle helium (3He/4He ratios ~10-5) is estimated to be less than 10% except in the well VP7519.

Different models have been proposed to estimate the magnitude of the external flux of helium into aquifer. In our particular case, we used a simple model presented by Balderer and Lehmann (1991) which relate the flux of crustal 4He (F, cm3 STP He per cm2 per year, [cm3/cm2yr]) into the aquifer to the mean groundwater residence time (tR) and the measured 4He concentration [4He] by,

[ ]pdtF

He R

⋅⋅

=4

where d is the thickness of the aquifer (30-80 m) and p is the total porosity (0.2). This model considers the aquifer as a closed system for helium, where it is gradually accumulated along the flow. The in situ production inside the aquifer is neglected. Based on this relationship, a helium flux in the range 0.3-0.8 10-6 cm3/cm2yr into the Cenomanian aquifer is calculated. The magnitude of this range is close to the range 0.07-0.5 10-6 cm3/cm2yr reported by Stute et al. (1992) for the Great Hungarian Basin, and lower than the value of 3.1 10-6 cm3/cm2yr, which represents a flux characteristic of the degassing of the whole continental crust

75

(Torgersen and Clarke, 1985; Ballentine et al, 2002). The estimated He flux can be used to derive groundwater ages in other sections of the Cenomanian aquifer and to extent the dating range to groundwater ages older than 30 000 years (range of the 14C method). 4.3.4. Recharge conditions and palaeoclimate

The stable isotope ratios δ2H and δ 18O in groundwater from the Turonian aquifer plot

along the local meteoric water line (LMWL) calculated for the region (data taken from the Vienna, Cracow, Leipzig and Neuherberg stations of the GNIP database, 2004) (Fig. 10). The deuterium excess (d-excess) in the samples, defined by d=δ2H-8δ18O, varies between 7.0 and 9.2 ‰ (Table 3). In the Cenomanian aquifer, the stable isotopes ratios δ2H and δ18O plot above the LMWL calculated for the region (Fig. 10). Most of the samples plot along a different water line (δ2H=8 δ18O+10.8) which is similar to the global meteoric water line proposed by Rozanski et al. (1993) (δ2H=8.1 δ18O+10.8). The significant isotope depletion of the sample VP7519 in the southern part of the aquifer may be due to a lower recharge temperature at the time of infiltration. The d-excess in the samples from this aquifer varies between 10.4 and 11.4 ‰ (Table 3). Only the well Karany B shows a lower d-excess and plot in the LMWL. Since water from this well is a mixture of different water components we conclude that young water, probably from the upper Turonian aquifer, is mixed with old water from the Cenomanian aquifer.

-12.5 -12.0 -11.5 -11.0 -10.5 -10.0 -9.5 -9.0 -8.5-85

-80

-75

-70

-65

VP7517

VP7515

VP7500VP7502

VP7506

VP7519

Cenomanian Aquiferd2H = 8.0 (d18O) + 10.8(r=1.0)

Karany BMixture

Turonian Sands Aquifer Cenomanian Sands Aquifer

LMWL (GNIP stations, IAEA, 2004)d2H = 7.5 . d18O + 3.1 (r=0.99)Composite of Cracow, Vienna, Leipzig and Neuherberg

d2 H (‰

vs

SMO

W)

d18O (‰ vs SMOW)

Figure 10. 18O versus 2H content in groundwater samples from the Turonian and Cenomanian aquifers. The noble gas data were interpreted with different models that account for the dissolution

of the noble gases in groundwater (Aeschbach-Hertig et al., 2000). The best fits were obtained with the closed-system equilibration (CE) model. This model assumes that the noble gas concentrations in recharging water are in equilibrium with the atmosphere at atmospheric pressure and soil temperature, and that bubbles of atmospheric air are trapped. The bubbles are not completely dissolved in water creating a reservoir of entrapped gas in the quasi-saturated zone. The parameters that best fitted the CE model to the data are shown in Table 4.

76

A mean noble gas temperature (NGT) of 7.8±1.9 oC was calculated for the groundwater samples of the Turonian aquifer. This value is comparable to the present mean annual air temperature in the region (6-8 oC, by the Czech Hydrometeorological Institute).

The NGTs calculated for waters which recharged the Cenomanian aquifer during the Holocene are similar to the present mean annual air temperature of the region (Table 4). The well VP7519 that recharged during the end of the Pleistocene shows a lower NGT of 0.79±0.33 oC (Table 4). Old groundwater from well VP7519 is depleted by 1.5‰ in δ18O and 10‰ by δ2H relative to the Holocene samples (Fig.10a, 11b). Depleted δ18O and δ 2H values and low NGT indicate that groundwater from this well probably recharged during colder climatic conditions than the present atmospheric temperature (Table 3 and 4). The NGTs reveal a temperature difference between Holocene and the end of the Pleistocene of about 5 °C (Fig. 11a). This figure is in agreement with findings in other European aquifers (Andrews, 1993; Aeschbach-Hertig et al., 1999 and 2000; Beyerle et al., 1999; Zuber et al., 2004).

0 10000 20000 300000

2

4

6

8

0 10000 20000 30000-12

-11

-10

-9

Holocene mean NGT (oC)

VP7520VP7524

VP7500

VP7515

VP7519

Karany B(Mixture)

VP7502

recharge 5 oC colder NG

T (o C

)

14C ages (yrs.)

2H

d18O

(‰)

14C ages (yrs.)

0 10000 20000 30000

-85

-80

-75

-70

-65

-60 18O

VP7506VP7500

VP7515

VP7517

VP7519

Karany B(Mixture)

VP7502

Colder recharge conditions

d2 H

(‰)

Figure 11. a) Noble gas temperatures (NGT) versus the 14C ages for the Turonian and Cenomanian

aquifers. An atmospheric cooling by approximately 5 oC has been derived from the data of the well VP7519. A minimum 14C age of 25000 years is used to represent the old groundwater component in the well Karany B. b) Stable isotopes 18O and 2H contents in groundwater versus the 14C age. A shift to more depleted 18O and 2H is observed for the sample VP7519, which reflects colder atmospheric temperatures.

The d-excess values of precipitation are strongly influenced by water vapor source and can therefore be used as a tool for reconstructing atmospheric circulation patterns. In the Cenomanian aquifer, the d-excess does not show any strong variation from the Pleistocene to the Holocene (Table 3), which imply that relative humidity over the subtropical regions of the North Atlantic Ocean and circulation patterns of the atmosphere over Europe have not changed considerably during this period. This result agrees with the observations made in other aquifers over Europe (Rozanski, 1985, Huneau et al., 2001). However, the d-excess in this aquifer is characterized by a slight increasing trend during the Holocene (Fig. 12). Somewhat similar trends have been observed in Greenland records and have been attributed to an increase of low-altitude annual mean insolation (warming ocean temperatures) and a decrease of high-latitude annual mean insolation (cooling ocean temperatures) in response to the progressive Holocene decrease in obliquity (Masson-Delmotte et al., 2005b). This phenomenon should enhance the relative contribution of low-latitude moisture sources.

A large difference is observed between the d-excess values of the Turonian aquifer containing young water (up to 200 yrs) and the values of the Cenomanian aquifer containing

77

old water (older than 600 years) suggesting either a significant change in the atmospheric circulation patterns over Europe (a different moisture source component with lower d-excess) or an increase of the moisture content in the location of the vapour formation (this would increase the kinetic fractionation and consequently would lower the d-excess) in the last 200 to 600 years. A similar pattern was only observed in the Malm limestone and the Tertiary sands aquifers in Cracow (data taken from Zuber et al., 2004), which is located a few hundreds of kilometres north-east from our investigated area, if only the glacial and modern waters are plotted (Fig. 12).

0 10 20 30 40 50 60 70 80 90 1006

7

8

9

10

11

12

13

Decrease of the d-exceess

Old Groundwaters

Modern Groundwaters

Cenomanian and Turonian sands aquifers Jurasic Limestone and Tertiary sands aquifers, Cracow

d-ex

cess

(per

mil)

14C (pMC) Figure 12. d-excess as a function of the 14C activity measured in groundwater samples from the

Turonian sands and the Cenomanian sands aquifers (this study), and the glacial and modern samples from the Jurassic Limestone and the Tertiary sands aquifers (Data taken from Zuber et al., 2004).

The stable isotope ratios δ2H and δ18O in groundwater from the Cenomanian aquifer (Holocene and Pleistocene waters) follow the same continental gradient observed by Rozanski (1985) in recent infiltrated waters and in precipitation; therefore a similar atmospheric circulation pattern may be deduced. A slight increase in the moisture contents (<5%) at the ocean surface remain then as the most probable reason for explaining the decrease of the d-excess at present days. No evidences have been observed in any other investigation. We should keep in mind that a small number of samples were studied in both aquifers. Therefore, further studies, with larger number of samples, are required to fully understand the origin of this tendency. 5. Conclusions

The groundwater residence times in the semiconfined Turonian sands aquifer vary from

modern to 200 years. This aquifer shows chemical and age stratifications. The presence of modern recharged groundwater coincides with the detection of elevated concentrations of nitrate and other ions, indicating the vulnerability to surface pollution.

78

In the confined Cenomanian sands aquifer, the measurements of 39Ar are proposed to identify the initial activity of 14C and date the groundwater in this initial well, both very important points for radiocarbon dating. The calculated 14C ages vary between a few hundreds of years and more than 20,000 years, showing a linear increase with the distance from the recharge area. This result probably indicates that piston flow is a good approximation for the flow in the aquifer. The groundwater undergoes a chemical evolution along the flow path as a result of a number of different geochemical processes. This evolution was used to investigate processes such the dissolution rates of some minerals and to extend the range of the groundwater dating.

The deep reaching Luzice fault may facilitate an upward transport of gases from underlying formations into the Cenomanian aquifer. This is suggested by a large content of subsurface produced 39Ar and by the input of carbon in the form of CO2 (dead in 14C and enriched in 13C) in groundwater from the well VP7502.

A vertical flux of helium from deeper layers (crust and mantle) is present in the investigated area. This flux has a value approximately one order of magnitude lower that that predicted for the degassing of the whole continental crust. Most of helium results from radioactive decay of U and Th in the sandstones of the sampled area. Larger contribution of mantle helium was found in the well VP7519.

The noble gas temperatures and the stable isotope signature show that water in the well VP7519 recharged at cooler temperatures than the present day annual average temperature. This result agrees with the dating, which predicts that the water recharged during the end of the Pleistocene when climate with lower air temperatures prevailed.

Finally, we can conclude that if the contamination from the Stràz block where uranium was leached by sulphuric acid is transported through the aquifers with the calculated groundwater velocity, it will take between several hundreds and thousands of years to reach the area most exploited for water supply.

Acknowledgement This project was financed by the programme BASELINE, European Commission,

amendment 1 to contract EVK1-CT-1999-00006. We thanks to Dr. D.Remenarova, V. Kodes (Czech Hyddrometeorological Institute), Z. Herrmann (AQUATEST, Prague) and J. Urban for offering us hydrological and hydrogeological data and help during sampling of ground water. References Aeschbach-Hertig W., Peeters F., Beyerle U. and Kipfer R. (1999). Interpretation of dissolved atmospheric noble

gases in natural waters. Water Resour. Res., 35(9): 2779-2792.

Aeschbach-Hertig W., Peeters F., Beyerle U. and Kipfer R. (2000) Palaeotemperature reconstruction from noble gases in ground water taking into account equilibration with entrapped air. Nature 405, 1040-1044.

Andel P. and PribánV. (1996). Environmental restoration of uranium mines and mills in the Czech Republic. In: Planning for environmental restoration of radioactively contaminated sites in Central and Eastern Europe, Vol. 1: Identification and characterization of contaminated sites, IAEA-TECDOC-865, Vienna, 113-135.

Andrews J.N. (1993). Isotopic composition of groundwaters and palaeoclimate at aquifer recharge. In Proceedings of the Int. Symp. On applications of Isotop. Tech. In studying past and current changes in the hydrosphere and atmosphere, IAEA, Vienna.

79

Balderer W., Lehmann B.E. (1991) Isotopes formed by underground production: 3He and 4He. In Applied Isotope Hydrogeology, a Case Study in Northern Switzerland (eds. F.J. Pearson et al.). Elsevier, Amsterdam, 239-296.

Ball, J.W. and Nordstrom D.K. (1991). User's manual for WATEQ4F, with revised thermodynamic data base and test cases for calculating speciation of major, trace, and redox elements in natural waters: U.S. Geological Survey Open-File Report 91-183, 189 pp. (revised and reprinted August 1992).

Ballentine C. J., Burgess R., and Marty B. (2002). Tracing fluid origin, transport and interaction in the crust. Rev. Min. Geochem. 47, 539-614.

Barraclough P.B. and P.B. Tinker (1982). The determination of ionic diffusion coefficients in field soils. II. Diffusion of bromide ions in undisturbed soil cores. J. Soil Sci. 33:13–24.

Beyerle U., Aeschbach-Hertig W., Imboden D.M., Baur H., Graf T., Kipfer R. (2000). A mass spectrometry system for the analysis of noble gases and tritium from water samples. Environ. Sci. Technol. 34, 2042-2050.

Beyerle, U., R. Purtschert, W. Aeschbach-Hertig, D.M. Imboden, H.H. Loosli, R. Wieler, and R. Kipfer (1998). Climate and groundwater recharge during the last glaciation in an ice-covered region. Science, 282, 731-734.

Braeuer K., Kaempf H., Strauch G., Wieise S.M. (2003). Isotopic evidence (3He/4He, 13CCO2) of fluid intraplate seismicity. J. Geoph. Res., V.108, No.B2.

Cook P.G. and Solomon D.K. (1995) Transport of atmospheric trace gases to the water table: Implications for groundwater dating with chlorofluorocarbons and krypton-85. Water Resour. Res. 31(2), 263-270.

Corcho Alvarado J.A., Purtschert R., Hinsby K., Troldborg L., Hofer M., Kipfer R., Aeschbach-Hertig W., Arno-Synal H. (2005). 36Cl in modern groundwater dated by a multi tracer approach (3H/3He, SF6, CFC-12 and 85Kr): A case study in Quaternary sand aquifers in the Odense Pilot River Basin, Denmark. Applied Geochem., V. 30, 3, 599-609.

Corcho Alvarado J.A., Purtschert R., Barbecot F., Chabault C., Rueedi J., Schneider V., Aeschbach-Hertig W., Kipfer R. and Loosli H.H. Constraining groundwater age distribution using 39Ar: a multiple environmental tracer (3H/3He, 85Kr, 39Ar and 14C) study in the semi-confined Fontainebleau Sands aquifer (France). submitted to Journal of Hydrology.

Edmunds W.M., Smedley P.L. (2000). Residence time indicators in groundwater: the East Midlands Triassic sandstone aquifer. Applied Geochem. 15, 737-752.

Ekwurzel, B. Schlosser, P., Smetthie, W.M., Jr., Plummer, L.N., Busenberg, E., Michel, R.L., Weppernig, R. and Stute, M. (1994). Dating of shallow groundwater-Comparison of the transient tracers 3H/3He, chlorofluorocarbons, and 85Kr. Water Resour. Res., 30: 1693-1708.

Fontes J.C., Garnier J.M. (1979). Determination of the initial 14C activity of the total dissolved carbon: a review of the existing models and a new approach. Water Resour. Res. 15, 399-413.

Fontes, J.Ch., M. Stute, P. Schlosser, and W.S. Broecker, 1993. Aquifers as archives of paleoclimate. Eos Trans AGU, Vol. 74, pp. 21-22.

Forster M., Ramm K. and Maier P. (1992). Argon-39 dating of groundwater and its limiting conditions. In Isotope techniques in water resource development 1991 (ed. IAEA), IAEA, Vienna, 203-214.

Herčík F., Herrmann Z. and Velečka J. (2003). Hydrogeology of the Bohemian Cretaceous Basin. Edited by AQUATEST a.s. and KAP Ltd., Czech Geological Survey, Prague.

Huneau, F., B. Blavoux, W. Aeschbach-Hertig & R. Kipfer, 2001. Palaeogroundwaters of the Valréas Miocene aquifer (Southeastern France) as archives of the LGM/Holocene climatic transition in the Western Mediterranean region. In International Conference on the Study of Environmental Change Using Isotope Techniques, IAEA, Vienna. IAEA-CN-80/24: 27-28.

Loosli H.H. (1983) A dating method with 39Ar. Earth Planet. Lett. 63, 51.

Loosli, H.H.; Möll, M.; Oeschger, H., Schotterer, U., 1986. Ten years low-level counting in the underground laboratory in Bern, Switzerland. Nuclear Instruments and Methods in Physics Research Section B 17, 402-405.

Loosli H.H., Lehmann B.E. (1991) Isotopes formed by underground production: 39Ar and 37Ar. In Applied Isotope Hydrogeology, a Case Study in Northern Switzerland (eds. F.J. Pearson et al.). Elsevier, Amsterdam, 239-296.

80

Loosli, H.H., Lehmann, B.E., Smethie, W.M., 1999. Noble gas radioisotopes: 37Ar, 85Kr, 39Ar, 81 Kr. In : Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 379-396.

Masson-Delmotte V., Jouzel J., Landais A., Stievenard M., Johnsen S. J., White J. W. C., Werner M., Sveinbjornsdottir A., Fuhrer K. (2005). GRIP Deuterium Excess Reveals Rapid and Orbital-Scale Changes in Greenland Moisture Origin. Science, Vol 309, Issue 5731, 118-121.

Masson-Delmotte, V., Landais A., Stievenard M., Cattani O., Falourd S., Jouzel J., Johnsen S. J., Dahl-Jensen D., Sveinsbjornsdottir A., White J. W. C., Popp T., and Fischer H. (2005). Holocene climatic changes in Greenland: Different deuterium excess signals at Greenland Ice Core Project (GRIP) and NorthGRIP, J. Geophys. Res., 110, D14102.

Millington R.J. (1959). Gas Diffusion in Porous Media. Science 130: 100-102.

Novak J., Smetana R., Strof P., Ira P., Emmer J., Paces T. (2000). Mining of uranium by acid leaching and its environmental consequences. In: A Geochemical and Mineralogical Approach to Environmental Protectio (I. Memmi, J.C. Hunziker and C. Panichi, eds.) 215-240, Int. School Earth and Planetary Sciences, Universita degli Studi Siena, Siena, ISSN 1122-8830.

Paces T., Smejkal V. (2004). Magmatic and fossile components of thermal and mineral waters in the Eger River continental rift. Pages 167 – 172 in R. B. Wanty and R.R. Seal II, editors. Water-Rock Interaction 11, A.A. Balkema Pub., Leiden.

Plummer L.N., Prestemon E.C., Parkhurst D.L. (1991). An interactive code (NETPATH) for modelling net geochemical reactions along a flow path. USGS Water-Resources Investigations Report 91-4078, Reston, Virginia, USA.

Purtschert, R., Loosli, H.H., Beyerle, U., Aeschbach-Hertig, W., Imboden, D., Kipfer, R., Wieler, R. (1999). Dating of young water components by combined applications of 3H/3He and 85Kr measurements. In: International Symposium on Isotope Techniques in Water Resources Development and Management, Vienna, Austria, 59-60.

Rozanski, K., 1985. Deuterium and oxygen-18 in European groundwaters - links to atmospheric circulation in the past. Chem. Geol. (Isotope Geoscience Section), 52: 349-363.

Rozanski, K., Araguas-Araguas, L., and Gonfiantini, R. (1992) Relation between long-term trends of oxygen-18 isotope composition of Precipitation and Climate. Science, 258, 981.

Rozanski K., Araguás-Araguás L., Gonfiantini R. (1993). Isotopic patters in modern global precipitation, in Climate Change in Continental Isotopic Records, Geophys. Monogr. Ser., 78, ed. by P.K. Swart, et al, pp. 1-36, AGU, Washington, DC.

Rueedi J., Brennwald M.S., Purtschert R., Beyerle U., Hofer M., Kipfer R.. (in press) Estimating amount and spatial distribution of groundwater recharge in the Iullemmenden Basin (Niger) based on 3H, 3He and CFC-11 measurements. IAEA Conference on Isotope Hydrology, Vienna, 2003. Accepted by Hydrological Process – Special Edition.

Schlosser, P., 1989. Tritium/3He dating of waters in natural systems. In: Isotopes of noble gases as tracers in environmental studies, IAEA, Vienna, 123-145.

Schrag D.P., Hampt G., Murray D.W. (1996). Pore Fluid Constraints on the Temperature and Oxygen Isotopic Composition of the Glacial Ocean. Science 272: 1930-1932

Šilar J. (1990). Radioacarbon dating of groundwater in Czechoslovakia and palaeoclimatic problems of its origin in Central Europe. –Zesz. Nauk Politech. Slaskiej, Mat. Fiz. 61, 6, 133-141. Gliwice.

Šilar J. (1983). Kdy vznikly podzemní vody Českého masivu? – Věst. Ústř. Úst. Geol. 58, 1, 39-47. Praha.

Šilar J. (1977). Radiocarbon dating of ground water in the platform sediments of the Bohemian Massif. – Low-radioactivity Measurements and Applications, Slov. Pedagog. Nakl., 379-381. Bratislava.

Šilar J. (1976). Radiocarbon ground-water dating in Czechoslovakia – first results. Véstník Ústředního ústavu geologického 51, 209-220, Praha.

Solomon D.K. (1999). 4He in groundwater. In : Cook P.G. and Herczeg A.L. (eds.), Environmental Tracers in Subsurface Hydrology. Kluwer Academic Publishers, Boston, 425-439.

81

Stute, M., C. Sonntag, J. Deak, and P. Schlosser (1992) Helium in deep circulating groundwater in the Great Hungarian Plain: Flow dynamics and crustal and mantle He fluxes. Geochim. Cosmochim. Acta, 56, 2051-2067.

Torgersen, T. (1980). "Controls on pore fluid concentrations of 4He, 222Rn and the calculation of 4He/222Rn ages." J. Geochem. Explor., 13, 57-75.

Torgersen, T., and Clarke, W.B. (1985). Helium accumulation in groundwater, I: An evaluation of sources and the continental flux of crustal 4He in the Great Artesian Basin, Australia. Geochim. Cosmochim. Acta, v. 49, p. 1211-1218.

Weinlich F.H., Braeuer K., Kaempf H., Strauch G., Tesar J., Weise S.M. (1999). An active subcontinental mantle volatile system in the western Eger rift, Central Europe: Gas flux, isotopic composition (He, C, and N) and compositional fingerprints. Geochim. Cosmoch. Acta, Vol. 63. No.21, 3653-3671.

Zuber A., Maloszewski P. (2001). Lumped-parameter models. Environmental Isotopes in the hydrological cycle. UNESCO/IAEA.

Zuber, A., Weise, S.M., Motyka, J., Osenbrück, K., Rozanski, K. (2004). Age and flow patter of groundwater in a Jurassic limestone aquifer and related Tertiary sands derived from combined isotope, noble gas and chemical data. J. Hydrology, 286, p. 87-112.

82

83

Chapter 5 The age and origin of the Bath Thermal waters. W M Edmunds1, W G Darling1, R. Purtschert2 and J. Corcho2 1 British Geological Survey, Wallingford, UK 2 Climate and Environmental Physics, Physic Institute, University of Bern, Switzerland (Manuscript in preparation)

1. Introduction

As part of the development of the new spa, the opportunity exists to update the base of knowledge on the composition of the thermal springs using state-of-the-art geochemical tools, including detailed chemistry, radioactive and stable isotope studies on the water and the solutes, and investigation of dissolved gases. These investigations will provide the scientific basis, alongside hydrogeological and geophysical studies for explaining where the waters are likely to have come from, how old they are, as well as explaining many of their unique properties.

Scientific investigations of the Bath springs have been undertaken for around 175 years, although speculation on their origins have been made over a much longer period; the first scientific paper on their properties being by Glanville (1669). No fewer than 10 analyses of the springs have been published (although with varying degrees of detail and accuracy) since 1823. Detailed studies of the springs have been carried out by BGS and co-workers for the past 25 years with two detailed studies similar in many ways to the present investigations. The most recent geochemical studies have suggested that the water could be around 6000±2000 years; hydrogeological modelling has suggested that the age might be much lower than this, possibly as young as 500 years (Edmunds and Miles 1991).

Interpretation of the properties of the springs is made more difficult since there is only one point of emergence (although with bifurcation). Earlier studies have also suggested that there may be a degree of mixing of younger near-surface water with a much older and deeper source.

The objectives of the present investigations are therefore: to refine the estimates of age; to estimate the maximum temperature and depth of circulation; to determine the sources and amount of any modern groundwater contamination; to determine the source area(s) and

84

recharge conditions of the deeper sources as well as the reactions that have taken place during its circulation; and to explain several of the properties of the springs such as the degassing, the iron staining, the taste and odour, as well as other questions asked by the general public.

To achieve this a fully comprehensive suite of diagnostic indicators is being applied, representing the most complete look at the composition of the waters to date. The range of indicators and what they might show is summarised in Table 1. Several new tools are being used, notably the dissolved gas compositions. The highlight of this investigation is the analysis of 39Ar which should help resolve whether the water has any component with an age limit of 1000 yr. Other dissolved gas components such as 85Kr and CFCs (chlorofluorocarbons) can provide sensitive indications of recent additions to the water. Table 1. Geochemical indicators and their possible interpretation applied to the Bath thermal waters. Sampling

details and measurements carried out by different laboratories are given.

PARAMETER SCIENTIFIC ROLE LAB SAMPLING DETAILS AND NOTES

FIELD Temperature Field pH In situ properties Wellhead Dissolved Oxygen (DO) In situ Wellhead (anaerobic) Bucket Redox Potential (Eh) In situ Wellhead (anaerobic) SEC In situ Wellhead Alkalinity (HCO3) In situ Field titration INORGANIC WL Major anions (Cl, NO3-N), Fingerprinting and WRI F/UA (60 ml) Major cations (Na, K, Ca, Mg, SO4

F/A (60 ml)

Trace elements Anions: Br, F, I, NH4, NO2

F/UA (60 ml)

Trace elements (ICP-AES) Sr, Ba, Li, Si, B, Fe, Mn, Zn, P, V

Si geothermometry F/A (60 ml)

Trace elements (ICP-MS) Li, Be, B, Al, Cr, Mn, Co, NI, Cu, Zn, Ga, Ge, As, Se, Rb, Y, Mo, Ag, Cd, Sb, Cs, Ba, La,Tl, Pb, Bi, Th, U, REE (Rare Earth elements)

Fingerprinting, source rocks and WRI; residence time

F/A (60 ml) Plus 1 litre sample See working detection limits

AsIII/AsV F/A (60 ml) HNO3 ORGANIC WL DOC Natural vs contaminant source of C 50 ml (glass,Al seal) TOC 50 ml (glass Al seal) STABLE AND RADIOISOTOPES

WL,KY

δ 18O, δ 2H Origins and recharge characteristics of the water

WL Glass (30 ml)

δ 13C Source of inorganic carbon; correction of 14C

WL Glass (50 ml)

14C Groundwater age (5730 yr half life) GLA Amber Glass (2x500ml) δ 87Sr/ δ 86Sr Source of Sr KY DISSOLVED GASES (AND ISOTOPES)

CH4, N2, O2, N2/Ar WL Stainless steel “bomb” CFCs Fingerprint of modern water/age WL Glass 500ml (anaerobic) 39Ar/37Ar Groundwater age (269 yr half life) Bern Cylinders: process 5 m3 water 3He/4He Crustal component and high precision

groundwater age (as tritium) ZUR Copper tubes

Ne, Ar , Kr, Xe Recharge temperature ZUR Copper tubes 22Ne/21Ne ZUR Copper tubes 3H (3He) Modern component (half life 12.32 yr) ZUR Copper tubes 85Kr Modern component (10.7 yr half life) Bern Steel cylinders 222Rn Radioactivity of source rocks KY Amber glass (250 ml)

85

2. Methods

Arrangements were made for sampling of the inclined Kings’ Spring borehole and the Cross Bath spring on 11 August 2000. This involved cordoning off part of Stall Street above the borehole for a period of 24 hours to provide access for the gas sampling equipment. In addition, it was necessary to arrange a sampling line above the Cross Bath spring chamber via a shallow pipe inserted rather precariously over the bath.

A comprehensive series of samples were taken in various containers as indicated in Table 1. In situ measurements were made for a number of parameters that were unstable or could not be measured in the laboratory such as T, SEC, pH and redox potential (Eh).

A highlight of the present work was the analysis of 39Ar which was incorporated to help resolve whether water has any component with an age less than 1000 yr. The overall logistics for this sampling were considerable since approx 5 tonnes of water had to be processed to recover enough gas for analysis, taking most of one working day.

Analytical procedures

The chemical and isotopic analyses were carried out in four laboratories: two in the UK (BGS Wallingford and Keyworth) and two in Switzerland (University of Bern and ETH Zurich). Radiocarbon analysis was conducted through a research grant application with the NERC Radiocarbon Laboratory (East Kilbride) who processed the water, converting the inorganic carbon in HCO3 to graphite, the activity of which was then analysed by University of Arizona. The main techniques used are also summarised in Table 1. Problems were met with the ICP-MS analysis at Wallingford and so samples were also sent to a commercial laboratory in Vancouver. Full details of the analytical methods used can be obtained by contacting one of us (WGD) at BGS in Wallingford.

3. Results

Some of the areas for which the various measurements can be used for interpretation are summarised in Table 1. Full results of the 2000 sampling of the Kings Spring Borehole and the Cross Bath are listed in Table 2a-c and the main characteristics of the waters (their major and minor inorganic solutes, stable and radioisotope compositions, and dissolved gases) are briefly discussed under separate headings:

Field observations

The discharge temperature of the borehole is 44.8oC (table 2a), slightly higher than the Cross Bath, consistent with slight cooling of the latter. The pH of both sources is similar. The redox potential of the Kings Spring (-2 mv) indicates reducing conditions at depth and differs from the Cross Bath, which is less reducing and points to a possible difference in conditions between the deeper and shallower levels of emergence.

Major ion characteristics Seven major ions comprise the bulk of the dissolved solids in the Bath thermal waters

(Table 2a). The total mineral content of 2290 mg/l is dominated by sulphate and calcium. Sulphate comprises 1080 mg/l of the total. During its circulation the water has also picked up almost equivalent amounts of Na and Cl (218 and 344 mg l-1 respectively).

86

Table 2a. Site data, field and laboratory measurements of major and minor species for the August 2000 sampling of the Stall Street B/H and the Cross Bath. Site T pH Eh SEC Diss. O2 Ca Mg Na K Cl SO4 HCO3 NO3-N NH4-N TOC DOC F Br I Si

oC mv mS cm-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 mg l-1

Stall Street B/H

44.8 6.71 -2 2560 <0.1 385 50.9 218 19.9 344 1080 187 < 0.5 0.35 1.1 4.3 2.28 1.86 0.04 21.78

Cross Bath

44.5 6.70 158 2830 <0.1 407 52.5 221 20.4 335 1120 149 < 0.5 0.37 5.6 19.2 2.23 1.75 0.04 22.48

Table 2b. Trace element data for the August 2000 sampling of the Stall Street B/H and Cross Bath. All results expressed as μg l-1. Site Ag Al AsT As3 B Ba Be Bi Cd Ce Cs Cu Dy Er Eu Fe Gd Ge Hg Ir La Li Lu Mn Mo

Stall Street B/H < 0.05 3 7.1 5.92 516 32.2 0.15 < 0.05 0.13 0.06 3.33 1.1 0.01 0.01 < 0.01 774 0.03 0.72 < 0.1 < 0.05 0.05 0.39 < 0.01 45.6 1

Cross Bath < 0.05 1 7.1 5.55 520 30.9 0.09 < 0.05 0.09 0.06 3.32 1.2 0.01 0.01 < 0.01 722 0.03 0.76 < 0.1 < 0.05 0.05 0.41 < 0.01 48.3 1

Nb Nd Ni Os Pd Pr Pt Rb Re Rh Sc Se Sm Sn Sr Ta Te Th Tl V W Y Yb Zn Zr

Stall Street B/H 0.01 0.08 < .2 < 0.05 < 0.2 0.01 < 0.01 22.6 < 0.01 < 0.01 5.41 8.5 <0.05 0.07 6601 < 0.05 0.14 < 0.05 0.1 2 < 0.1 0.12 0.01 4.4 < 0.5

Cross Bath 0.01 0.07 < .2 < 0.05 < 0.2 0.01 < 0.01 22.0 < 0.01 < 0.01 5.31 8.8 <0.05 0.08 6395 < 0.05 0.09 < 0.05 0.11 1 < 0.1 0.12 0.01 6.5 < 0.5

Table 2c Isotope and gas data for the August 2000 sampling of the Stall Street B/H and Cross Bath.

δ2H δ 18O δ 13C-DIC A14C 87Sr/86Sr N2 O2 Ar CO2 CH4 N2/Ar CFC-11 CFC-12 CFC-113 39Ar 37Ar 85Kr 4He 3He/4He Ne 222Rn ‰ ‰ ‰ pmC ratio mg l-1 mg l-1 mg l-1 mg l-1 mg l-1 molar pmol l-1 pmol l-1 pmol l-1 pmA dpm l-1 Ar dpm l-1 Kr cm3 l-1 STP ratio cm3 l-1 STP Bq l-1

Stall St B/H -45.0 -7.14 -1.54 2.16 0.710551 23.5 < 0.7 0.52 36.3 0.053 45.3 0.17 < 0.01 0.02 15 0.14 600 3.59E-02 3.36E-08 3.29E-04 92.9

Cross Bath -45.6 -7.21 -1.86 3.27 0.710550 29.3 < 0.7 0.65 46.9 0.071 0.26 < 0.01 0.04 137.4

87

Dissolved organic carbon

The content of organic carbon in the Cross Bath is relatively high (Table 2a) and greater than that in the Kings Spring borehole. This may be a possible indicator of shallow groundwater influx. Minor and trace elements

This study provides the most comprehensive trace element analysis of the water carried out to date with 70 elements having been determined, 48 of which are present above their limits of detection (Table 2b). These elements are derived through reactions with the geological formations through which the water has passed, thereby providing a fingerprint of the thermal water that can be used to restrict some of the interpretations of its origin. Similarities between the Kings Spring borehole and Cross Bath also confirm the same primary source of the water. Stable isotopes (water and solutes)

Stable isotope ratios (δ18O, δ 2H) provide an indication of the source of the water, especially whether it can be related to modern rainfall or not. Two other isotope ratios δ 13C and 87Sr/86Sr provide additional important tracer information. These data are provided in Table 2c. Radio-isotopes

Radiocarbon (14C) and tritium (3H) provide the traditional indicators of absolute age for the waters. In the present study these measurements have been made but difficulties with radiocarbon interpretation exist, since most may have been lost through reactions with the aquifer rocks rather than simply by decay. Relative differences in the two sources (Table 2c) suggest that the Cross Bath (with higher radiocarbon) may have a component of more radioactive modern groundwater. In the present study however, the dissolved gas isotopes 85Kr and 39Ar provide additional indicators of the absolute age of the groundwaters (Table 2c). Dissolved gases and gas isotopes

Some gases dissolved in the groundwater at the time of recharge retain their compositions and ratios to each other during groundwater circulation, while others may undergo chemical reactions. Gas compositions in the discharge may also be modified by addition from subsurface sources. The concentrations of gases in the thermal water reveal something of the balance of processes in the reservoir.

Of particular importance in the present investigation are the noble gases (Ar, Kr, Ne, Xe), which record the groundwater temperatures at the time of recharge, and CFCs which can indicate if there is any modern contribution. The noble gases can also provide information via their isotopic content. The radioactive isotopes such as those of Ar and Kr provide quantitative residence time information, while those of helium are more qualitative but can nevertheless provide indications of contributions of very old water.

88

Table 3. Comparison of the current analysis for the Stall Street B/H with earlier data for Stall street and the King’s Spring.

Unit Stall St B/H Stall St B/H Kings Spring Year - 2000 1986 1979 T oC 44.8 44.4 45.3 pH - 6.71 6.70 6.65 Eh mV -2 variable variable Dissolved Oxygen mg l-1 <0.1 <0.1 <0.2 Ca mg l-1 385 390 382 Mg mg l-1 50.9 58 53 Na mg l-1 218 228 183 K mg l-1 19.9 18.1 17.4 Cl mg l-1 344 335 287 SO4 mg l-1 1080 1030 1032 HCO3 mg l-1 187 187 192 NO3-N mg l-1 < 0.5 <0.05 <1 F mg l-1 2.28 2.0 2.08 Br mg l-1 1.863 2.0 2.02 I mg l-1 0.04 0.042 0.043 Si mg l-1 21.78 19.5 20.6 B μg l-1 516 590 590 Ba μg l-1 32 30 24 Fe μg l-1 774 1000 880 Li μg l-1 0.394 230 242 Mn μg l-1 45.6 50 68 Sr μg l-1 6601 6300 5920 δ2H ‰ -45.0 -47 -47 δ 18O ‰ -7.14 -7.4 -7.4 δ 13C-DIC ‰ -1.54 -1.5 A14C pmC 2.16 4.5 87Sr/86Sr ratio 0.710551 0.71075

4. Comparison with earlier analyses

In this work a comparison is made (Table 3) between the current analysis and those made during two earlier detailed analyses in 1979 and 1986; comparisons with earlier analyses prior to 1979 are found in Edmunds and Miles (1991). The results are given only for those elements where a direct comparison is relevant.

A decrease in the temperature of the groundwater derived from the Stall Street borehole was observed compared with the Kings Spring and this slightly lower temperature is confirmed. Some seasonal fluctuation, possibly related to flow, may occur and monitoring information is required to resolve this. The recent temperature is shown in relation to historical temperature measurements in Figure 1.

89

Dec

-76

Mar

-77

Jun-

77

Oct

-77

Jan-

78

Apr

-78

Jul-7

8

Oct

-83

Jan-

84

Apr

-84

Aug

-84

Nov

-84

Feb-

85

Jun-

85

Sep

-85

Dec

-85

Mar

-86

King's Spring

Stall St B/H

Hetling Spring

B/H ave 44.4°C

Dec

-99

Mar

-00

Jun-

00

Sep

-00

Dec

-00

Ave 45.3°C

1961

1936

1873

42

43

44

45

46

47

48

49

Tem

p °C

Figure 1. Temperature of the Kings Spring and Cross Bath in relation to historical temperature

measurements. NB the change in monitoring from the Kings Spring to the Stall Street (Kings Spring) borehole in 1986.

The chemistry of the carbonate system (pH, HCO3, Ca) has been constant over the 20-year period. Sulphate and chloride have been used as indicators of the compositional trends in the past and the present results (Figure 2) are compared with present data; both species are slightly higher than in previous analyses.

850

900

950

1000

1050

1100

SO

4 mgl

-1

1961

1936

1873

260

280

300

320

340

360

Cl m

gl-1

Dec

-76

Mar

-77

Jun-

77

Sep-

77

Dec

-77

Mar

-78

Oct

-83

Jan-

84

Apr-

84

Jul-8

4

Oct

-84

Jan-

85

Apr-

85

Jul-8

5

Sep-

85

Dec

-85

Mar

-86

Jun-

86

King's Spring

Stall St B/H

Hetling Spring

Figure 2. Sulphate and chloride from the present study compared with historical measurements.

Other changes that may be significant are a slight increase in silica (indicating a slightly deeper circulation pathway, discussed below), higher F, Li, Sr – all trace elements which remain conservative in the system and may be interpreted in terms of a longer residence time. In contrast, the Br has decreased (in parallel with Cl) and the possible significance of this is discussed below.

90

Slightly more positive values for δ18O and δ2H are observed but these are within the analytical precision of the measurement and therefore probably not significant. Of more significance may be the decrease in 14C activity (possibly indicating an increase in relative age or, more likely, change in mixing ratios).

5. Interpretation of the data Age of the groundwater and extent of mixing with modern water

Two persistent problems in determining the groundwater age have been a) the difficulties

in the modelling of a radiocarbon age due to reaction with the reservoir rocks which is likely at elevated temperatures, and b) the degree of mixing with any shallow water with a residence time of less than 1000 yr. In the present study additional tracers (37Ar, 39Ar and 85Kr) allow the mixing problem to be properly addressed before the age of the water is considered.

Being noble gases, 85Kr and 39Ar undergo no chemical reactions and are, in ideal cases, relatively easy to interpret as residence time indicators. Over the last 50 years 85Kr has been released from nuclear fuel processing plants leading to steadily increasing concentrations in the atmosphere. With a half-live of 10.7 years 85Kr is suitable for the identification and dating of water components with residence times of decades. It has been shown that subsurface production of 85Kr can be neglected (Lehmann et al. 1993).

The atmospheric 39Ar concentration originates mainly from the interaction of neutrons derived from cosmic ray activity with 40Ar and does not vary over time (Loosli 1983). With a half-life of 269 yr, groundwaters with ages up to 1000 yr can be identified. The usefulness of 39Ar as a dating tool may be limited because 39Ar atoms can be produced in underground by the 39K(n,p)39Ar reaction. Subsurface neutron fluxes and production rates can be roughly estimated by theoretical calculations based on measured rock compositions and by comparison with other isotope concentrations that are produced in the subsurface. A suitable indicator is 37Ar because (i) chemically it behaves identically to 39Ar and (ii) because a half-life of only 35 days excludes an atmospheric origin in older groundwaters. 37Ar is produced by the reaction of neutrons with 40Ca.

Nuclide concentrations in water Nw are connected with concentrations in the bulk rock matrix NR:

peNN Rw ⋅= (1)

where e denotes the escape probability from the rock into the water and p the rock porosity; e is mainly a function of the rock geometry and porosity and the spatial distribution of the target elements K and Ca.

Estimations of the production rates in the rock as well as escape probabilities are subject to considerable uncertainties. Therefore calculated activities can only give the order of magnitude of the expected measured values in the water.

The underground production rates of 37Ar and 39Ar were estimated from the chemical composition of the rocks in the reservoir and the cross section data for the energy of the neutrons (Andrews et al. 1991). It has been assumed for this that the water has equilibrated in the Carboniferous Limestone.

The results of the calculations are compared with the measured gas concentrations in the water in Table 4.

91

Table 4. Measured Ar and Kr concentrations and calculated values for in-situ production.

Isotope Measured values Calculated values Isotope activity

Gas concentration

cm3 STP/cm3 water Isotope concentration

atoms/cm3 water Isotope concentration

atoms/cm3 rock 39Ar 15 ± 2.4

%modern Ar (4.81 ± 0.03) 10-4

total argon 1.65 ± 0.28 Range 0.3 – 1.0

37Ar 0.14 ± 0.01 dpm/L argon

(4.81 ± 0.03) 10-4 total argon

(4.9 ± 0.2) 10-3 ~5 10-3

85Kr 0.6 ± 0.1 dpm/cc Kr

(1.07 ± 0.01) 10-7 total krypton

0.52 ± 0.09 ~1 10-4

Adopting equation (1) and assuming an escape probability and a porosity of the limestone (Andrews et al. 1992) of 1% for both, the numbers in the last two columns can directly be compared. The measured and calculated 37Ar concentrations in water do approximately agree, indicating the correctness of the selected parameters used for the calculations. For the other two gases, however, there is a significant amount of atmospheric input denoting a modern component. In the case of 85Kr, the in situ production is insignificant and the presence of young water is the only explanation. This is also supported by the presence of a small amount of tritium (1±2 TU) measured in previous samples (Andrews et al. 1982) and confirmed again in the present investigation.

The large relative errors of the 85Kr and 3H measurements allow only a rough estimate of age and portion of the recent water component. The interpretation of the data assuming a dispersion flow model (Maloszewski and Zuber 1996) with a Peclet number in the range of 10 to 100 (Gelhar et al. 1992), yields a mean residence time for the young water component of one to three decades and making up 2-8 % in the mixture. This is similar to chemical estimates of a young component (5%) from earlier studies (Edmunds 1991).

This result has implications for the 14C interpretation. Assuming an activity of the young component of 85% modern carbon (pmC) (Vogel 1967) and a mixing ratio of 5% explains already all of the measured activity of 2.2 pmC. This must mean that the age of the main component cannot be recent water. The very high δ13C value of -1.5 ‰ which indicates a pronounced dilution of the atmospheric 14C due to isotope exchange (Fontes and Garnier 1979) with the aquifer material impedes any further conclusion about the age of the old component.

The confirmation of mixing also has implications for the 39Ar, some of which may be coming from the young component. The 39Ar concentration in modern water, according to the solubility at the recharge temperature (4.813 10-4 cm3 STP/g water at 10oC), is 11 atoms/cm3 water. After mixing, the concentration of 39Ar (5% of the young component) is 0.5 atoms/cm3 water. Consequently, 1.3 atoms per cm3 water (of the total of 1.7 atoms) are from the old component. The range of in situ production is variable (0.3-1), due to uncertainties in a number of the parameters used (Andrews et al. 1991; Loosli et al. 1991). This means that the residual atmospheric 39Ar that is relevant to dating is very similar to the in situ production value and that the minimum groundwater age from this isotope must be 1000 years and most probably much older.

An independent constraint on the groundwater age comes from the dissolved noble gas concentrations (not reported). These indicate the temperature at which the groundwater was recharged at source. The mean ratio from duplicate measurements of the Stall Street B/H water is 10.3 ± 0.5 oC, identical within error to modern. This must mean that the water entered the aquifer since the last ice age or, much less likely, during the last interglacial maximum

92

(more than 100,000 yr ago). This would put a likely upper limit on the groundwater age of 12,000 yr.

CFCs (chlorofluorocarbons) are extremely sensitive indicators of the presence of ‘modern’ (up to ~50 years old) waters. The existence of three CFC species allows a degree of cross-calibration. In the Stall Street BH, CFC-11, CFC-12 and CFC-113 give amounts of modern water as 3, <1 and 4 % respectively, whereas at the Cross Bath the figures are 5, <1 and 7 %. The CFC-11 and CFC-113 values are similar both to each other and to those obtained from consideration of noble gas and hydrochemical data (see above). On the other hand CFC-12, normally the most ‘reliable’ CFC, gives a below-detection value. However, little is yet known about how conservatively CFCs behave in thermal waters, and it may be that CFC-12 is more susceptible to thermal degradation than the other two species. The combined evidence is interpreted to indicate that a modern component no more than one or two decades old is present up to a maximum of 5%.

Unlike the noble gases and CFCs, the gas content of the thermal water can give only qualitative indications of residence time. The compositions measured at Bath are fairly typical of those from hot springs of moderate temperature. Water starts off at recharge with its gas content in equilibrium with the atmosphere. Subsequent processes may add to or subtract from individual gases. By the time the thermal water is intercepted in Bath various changes have occurred, principally the total disappearance of O2 and the elevation of CO2 to around one-third of the gas phase by volume. These effects are typical of long-residence thermal waters in carbonate aquifers: O2 has been consumed by water-rock interaction, while the steep rise in CO2 is due to re-adjustment of the carbonate system.

Nitrogen on the other hand tends to behave inertly in thermal systems unless there is some extra source of N2 gas such as denitrification. One check on this is to measure the N2/Ar ratio of the water: in the case of the Stall Street BH the value was 45.3, which is typical of groundwaters recharged at temperatures of around 10°C and which have been unaffected by denitrification.

Radon, with a half-life of only 3.8 days, cannot be regarded as a residence time indicator and is primarily used in calculations of fracture apertures, a subject beyond the scope of the present report. The present measurement is similar to those previously obtained and discussed by Andrews (1991).

The maximum temperature and depth of circulation

The measured temperature of the borehole discharge has remained the same (within measurement error) at 44.8°C, although as noted earlier, there may be seasonal fluctuations, probably related to the few percent of mixing with a shallow cooler source. Water-rock reactions are temperature dependent and certain dissolved constituents may be used to calculate the maximum temperatures to which the groundwater has been subjected. The reaction kinetics of dissolution may be quite rapid, but once in solution the kinetics of re-precipitation are slow and thus the water retains a memory of the maximum temperature. Geothermometers are most reliable in high temperature geothermal systems and for the Bath system only silica is likely to be applicable (Edmunds and Miles 1991).

The groundwaters are supersaturated with quartz (Table 5), but are close to saturation with chalcedony, which is more likely to control silica solubility at the relatively low temperatures involved. For an Si concentration of 21.8 mg l-1 (corresponding to SiO2 of 46.7 mg l-1) temperatures of 68.5°C (chalcedony) and 98.6°C (quartz) are obtained. Even if the lower temperature is regarded as being more likely, a minimum drop in temperature of more than

93

20°C is being exhibited by the thermal water between the reservoir and emergence at Bath. For the reasonably high output of 13 l s-1 this represents a significant thermal loss and suggests that the water is not being brought rapidly to the surface by a single major fracture as (for example) is suggested in the schematic cross-section of Andrews et al. (1982). It seems more likely that the water is travelling relatively slowly to the surface along a more complicated set of fractures while retaining its thermal signature.

Bath is in a low heat flow area of the UK with an estimated heat flow of about 45 mW/m2, corresponding to a geothermal gradient of around 20°C km-1 (Downing and Gray 1986). With a surface temperature of 11°C a maximum depth of circulation of 2.9- 4.4 km is indicated; however since the chalcedony control is more likely in this low temperature environment, a maximum depth of 3km is considered more probable. The silica concentrations are slightly higher in 2000 than in 1986 but, as with the change in radiocarbon activity noted earlier, this is likely to be the result of a small change in the amount of modern water mixing.

The source of the water, circulation pathways The isotopic evidence provides indications for the source of the water, while the solute concentrations (major and trace elements) provide clues as to the types of rocks through which the water has passed, with qualitative indications also of residence times.

-60

-55

-50

-45

-40

-9.0 -8.5 -8.0 -7.5 -7.0 -6.5 -6.0 -5.5

δ18O ‰

δ2 H ‰

Spa watersOther B/HCSO watersWML

Range of speleothemδ2H values (Holocene)

Sport Centre B/H

Weston Maltings B/H

Spa waters

Figure 3. Stable isotope diagram comparing present analyses with (i) Holocene speleothem δ2H values

(Dennis et al., 1998), (ii) presumed Pleistocene palaeowaters from the CSO investigation B/Hs in western Bath (Darling and Edmunds, 2001), and (iii) pumped B/Hs apparently producing a mixture of old and modern waters. WML – world meteoric line.

The stable isotope ratios (Figure 3) lie well within the range of modern waters measured in the region (Dennis et al, 1998) and distinct from palaeowater compositions (Darling et al, 1997). This supports the conclusion from the noble gas recharge temperatures that the groundwater is of Holocene age. Further support comes from deuterium data obtained from fluid inclusions from Mendip cave stalagmites dated at mid-Holocene age (Dennis et al, 1998): the values are identical within measurement error to those of the Bath thermal waters.

94

Hydrogeochemical evidence from waters confined by the Mesozoic fill to the west of the city centre suggests that the sediments have been invaded by thermal outflows from the Kings Spring area (Edmunds and Miles 1991; BGS, unpublished data). However, the stable isotope values of these waters are depleted and it appears that they must have been introduced in the late Pleistocene. If this is indeed the case, the implication is that the thermal circulation has been in existence significantly longer than the bulk of the water currently passing through it.

The Br/Cl ratio is a diagnostic indicator of the source of Cl and it was concluded in the 1986 study that since this ratio was slightly enriched in Br relative to sea water, the Cl was derived from organic rich formation waters rather than an evaporite source. This is consistent with an evolution in the Carboniferous Limestone sequence. The slight increase in Cl since 1986 is matched by a slightly lower Br/Cl ratio which is consistent with small amounts of mixing from higher Cl waters in either the Carboniferous or Mercia Mudstone (BGS, unpublished data).

Strong evidence for a prolonged groundwater evolution in the Carboniferous Limestone (or indeed any marine limestone) is provided by the δ13C value. As discussed in Edmunds and Miles (1991), the value of –1.54‰ (identical in both studies) indicates that isotopic equilibrium is approached with the rock – near to 0‰ - as compared with a young groundwater from limestone, which would have a value of near to -13‰, reflecting the initial reactions in the soil. This process is likely to have taken thousands rather than hundreds of years even at the higher temperatures.

Figure 4. Trilinear plot for the major cations (taken from Edmunds and Miles 1991).

The relatively low Mg/Ca ratio (see Figure taken from Edmunds and Miles 1992) also suggests that the groundwater has evolved in a relatively pure limestone formation. If this were a Mesozoic Limestone or a dolomitic limestone then it is likely that an Mg/Ca ratio after prolonged water rock interaction would be roughly equimolar (1:1). The ratio is slightly

95

enriched compared with younger waters in the Carboniferous Limestone, as would be expected following incongruent solution where traces of Mg are added to the groundwater over time. The similarity with the Hotwells (Bristol) thermal water which issues from the Carboniferous Limestone, is also apparent. There seems no doubt therefore that the Carboniferous Limestone is the host rock for the thermal water evolution.

The reducing conditions in the thermal waters restrict the mobility of most trace metals in the groundwater, with the exception of iron (Fe2+). Very few metals exceed 5μg/l. Arsenic and selenium are exceptions with concentrations respectively of 7.1 and 8.5μg/l respectively; the reduced form, arsenic (3) (5.9μg/l) predominates.

Based on the conclusions of Andrews (1991), the required karst storage for a thermal water age of 10,000 yr would be about 8 % of the Carboniferous Limestone volume within a 30° segment from Bath to the Mendip Hills (the traditionally accepted source of the Bath thermal water). Even allowing for a degree of karstic porosity over and above the 1 % porosity allocated to the Carboniferous Limestone largely on account of fracturing, 8 % is a highly improbable value. This implies that if the thermal water is of early Holocene age, storage must involve the Carboniferous Limestone throughout the Bath-Bristol basin rather than simply a relatively small segment.

Contributions from formations other than the Carboniferous Limestone

The possibility exists that the thermal water could contain small contributions from the overlying Coal Measures or the underlying Old Red Sandstone. There is no direct evidence of the former. Methane is present only at very low concentrations in the thermal waters (Table 2c). At Stall Street ethane was just detectable. The low CH4 concentration and the high CH4/C2H6 ratio of ~2000 are typical of bacterially-produced reduced carbon gases found in waters of comparable redox status in low-organic-content formations. There is no evidence from these results that the thermal waters have been exposed to Coal Measures or other rocks with an elevated organic content; higher amounts of CH4, C2H6 and C3H8 would be expected in such circumstances.

There is better evidence for a minor contribution from the Old Red Sandstone. Although the N2/Ar ratio measured for the Stall Street water suggested no perturbation of the dissolved N2 load (see above), the uncertainties associated with the measurement could mask small additions of Ar in the subsurface. This is derived from the radioactive breakdown of 40K to 40Ar. The extent of such additions can be revealed by measuring the ratio 40Ar/36Ar. As part of the present investigations a ratio of 297.0 was measured which exceeds the atmospheric solubility equilibrium value of 295.5. Owing to the low potassium concentration of the Carboniferous Limestone it is unlikely that the radiogenic production within this formation is solely responsible for this excess of 40Ar. The time to produce the observed excess would be over 20 million years assuming an escape of 10% for argon which is most probably an overestimation. The only explanation for this excess is an external source whether a very old water component from the underlying Old Red sandstone or a diffusive influx of argon gas. This finding is in agreement with the anomalously high 4He concentrations and low 3He/4He ratio found in the thermal waters at Bath (Table 2c). It is considered therefore that upward leakage of gases from underlying formations is necessary to explain these features.

96

6. Conclusions

Analyses of water, solutes, isotopes and dissolved gases carried out in 2000 provide the most comprehensive interpretation to date of the origins, age and circulation history of the Bath Thermal Springs. It is shown that the discharge contains a percentage, up to 5% of modern water, less than 50 years old as shown by 85Kr and CFCs, and is probably derived from Mesozoic strata some tens of metres below the point of emergence. This modern water probably contains oxygen and is responsible for the precipitation of iron colouring the waters on emergence. The modern water was intercepted formerly by the spring itself, but probably slightly more of this water has been intercepted since the drilling of the Kings Spring (Stall Street) borehole.

Confirmation of mixing with modern water helps an improved explanation of the residence time and the source of the deeper, thermal component of the spring. The age of this water must be in excess of 1000 years, as indicated by the argon-39 dating. However it must be less than 12000 years old on the basis of its dissolved noble gas and stable isotope contents. Qualitative evidence (for example enriched 13C and a likely negligible, residual 14C) suggests the water to be nearer the upper age limit, and a range of 6-10 kyr is proposed.

From geothermometry the water is likely to have reached a maximum temperature of between 69 to 99°C, probably nearer the lower figure, and indicating a most probable maximum circulation depth of around 3 km. The rise to the surface of the water is sufficiently indirect that a temperature loss of more than 20°C is sustained. There is overwhelming evidence that the water has evolved within the Carboniferous Limestone formation, although the chemistry alone cannot pinpoint the geometry of the recharge area or circulation route. Volumetric calculations imply a large storage volume and circulation pathway if typical porosities of the limestone at depth are used.

Residual thermal water is recognised in the surrounding area to a distance of a few km, and from its isotopic composition it is probable that recharge and discharge of thermal water has been taking place at least since the cooler climates of the Late Pleistocene.

References Andrews J N, Burgess W G, Edmunds W M, Kay R L F and Lee D J (1982) The thermal springs of Bath.

Nature 298, 339-343.

Andrews J N, Lehmann B E and Thalmann (1991) NSPEC_AR: software for calculations of underground production of 39Ar and 37Ar.

Darling W G and Edmunds W M (2001). The hydrogeochemistry of waters from the Bath CSO Project investigation boreholes. BGS Technical Report CR/01/126C.

Dennis P F, Rowe P J and Atkinson T C (1998) Stable isotope composition of palaeoprecipitation and palaeogroundwaters from speleothem fluid inclusions. In Isotope Techniques in the Study of Environmental Change, IAEA, Vienna, 663-671.

Downing R A and Gray D A (1986) Geothermal energy. The potential in the United Kingdom. HMSO, London. 187pp.

Edmunds W M and Miles D L (1991) The geochemistry of the Bath thermal waters. In Hot Springs of Bath (ed G A Kellaway), Bath City Council 143-156.

Edmunds W M (1996). Bromide geochemistry in British groundwaters. Mineralogical Magazine 60, 275-284.

Fontes, J.-C., and J.-M. Garnier. 1979. Determination of the initial 14C activity of the total dissolved carbon: A review of the existing models and a new approach. Water Resour. Res. 15:399-413.

97

Gelhar, L. W., C. Welty, and K. R. Rehfeldt. 1992. A critical review of data on field-scale dispersion in aquifers. Water Resour. Res. 28:1955-1974.

Kellaway G A (1991) Investigation of the Bath hot springs (1977-1987). In Hot Springs of Bath (ed G A Kellaway), Bath City Council 97-126.vc

Lehmann, B. E., S. N. Davis, and J. T. Fabryka Martin. 1993. Atmospheric and subsurface sources of stable and radioactive nuclides used for groundwater dating. Water Resources Research 29:2027-2040.

Loosli, H. H. 1983. A dating method with 39Ar. Earth and Planetary Science Letters 63:51-62.

Loosli H H and Lehmann B E (1991). Isotopes formed by underground production. In Applied Isotope Hydrogeology: A Case Study in Northern Switzerland (Eds F J Pearson et al), Elsevier.

Loosli H H, Lehmann B E and Smethie W M (1999). Noble Gas Radioisotopes 37Ar, 85Kr, 39Ar and 81Kr. In Environmental Tracers in Subsurface Hydrology (Eds P Cook and A L Herczeg), Kluwer, 379-396.

Maloszewski, P., and A. Zuber. 1996. Lumped parameter models for the interpretation of invironmental tracer data. Pages 9-58 in Manual on Mathematical Models in Isotope Hydrology. IAEA, Vienna.

Smethie W M, Solomon D K, Schiff S L and Mathieu G (1992). Tracing groundwater flow in the Borden aquifer using Krypton-85. Journal of Hydrology 130, 279-297.

Vogel, J. C. 1967. Investigation of groundwater flow with radiocarbon. Pages 355-369 in IAEA, editor. Isotopes in Hydrology. IAEA, Vienna.

98

99

PUBLICATIONS, CONFERENCE ABSTRACTS and REPORTS

Articles published in reviewed journals

Corcho Alvarado J.A., Purtschert R., Hinsby K., Troldborg L., Hofer M., Kipfer R., Aeschbach-Hertig W., Arno-Synal H. (2005). 36Cl in modern groundwater dated by a multi tracer approach (3H/3He, SF6, CFC-12 and 85Kr): A case study in Quaternary sand aquifers in the Odense Pilot River Basin, Denmark. Applied Geochemistry, V. 30, 3, 599-609.

Purtschert R. and Corcho Alvarado J.A. (2005). Anwendung von Edelgasisotopen in der Grundwasserhydrologie . In Zbl. Geol. Palaeont. Teil I, H.1/2, Stuttgart, Pages 141-165.

Articles submitted to reviewed journals

Corcho Alvarado J.A., Purtschert R., Barbecot F., Chabault C., Rueedi J., Schneider V., Aeschbach-Hertig W., Kipfer R. and Loosli H.H. Constraining groundwater age distribution using 39Ar: a multiple environmental tracer (3H/3He, 85Kr, 39Ar and 14C) study in the semi-confined Fontainebleau Sands aquifer (France). submitted to Journal of Hydrology.

Articles in preparation

Corcho Alvarado J.A., Purtschert R, Pačes T., Kipfer R. and Leuenberger M. Groundwater dating in the Turonian and Cenomanian aquifers of the Bohemian Cretaceous Basin: A first step in getting insights on underground processes and recharge conditions. In preparation.

Edmunds W.M., Darling W.G., Purtschert R.and Corcho Alvarado J.A.. The origin and age of the Bath thermal waters. in preparation.

Technical reports

Hinsby K., Troldborg L., Purtschert R. and Corcho Alvarado J.A. (2003). Integrated transient hydrological modelling of tracer transport and long-term groundwater/surface water interaction using four 30 year 3 H time series and groundwater dating for evaluation of groundwater flow dynamics and hydrochemical trends in groundwater and surface water. Report to the International Atomic Energy Agency (TECDOC in press by the IAEA ).

Edmunds W.M.; Darling W.G.; Purtschert R.; Corcho Alvarado J.A., (2001). The age and Origin of the Bath Thermal Waters: New Geochemical investigations as part of the Bath Spa Project. British Geological Survey, Commissioned Report, CR/01/263.

Conference abstracts

Corcho Alvarado J.A., Purtschert R., Hofer M., Aeschbach-Hertig W., Kipfer R., Troldborg L., Hinsby K. (2002). Comparison of residence time indicators (3H/3He, SF6, CFC-12 and 85Kr) in shallow groundwater: a case study in the Odense aquifer, Denmark. Goldschmidt conference, Davos. Geochim. Cosmochim. Acta 66: A152.

Corcho Alvarado J.A.; Purtschert R.; Chabault C.; Barbecot F.; Rueedi J.; Schneider V., Aeschbach-Hertig W.; Kipfer R.; Loosli H.H.; Dever L. (2003). Origin and temporal

100

evolution of the groundwater in the Fontainebleau Sands Aquifer (France) investigated using 3H, 85Kr, 39Ar, 14C and stable noble gases. Int. Symposium on Iso. Hydrol. and Integrated Water Res. Manag., IAEA, Austria.

Corcho Alvarado J.A., Purtschert R., Barbecot F., Chabault C., Rüedi J., Schneider V., Aeschbach-Hertig W., Kipfer R., Loosli H.H. (2004). Tracer transport in the unsaturated zone of the Fontainebleau sands aquifer. International Workshop on the Application of Isotope Techniques in Hydrological and Environmental Studies, IAEA/UNESCO, Paris, France.

Corcho Alvarado J.A., Purtschert R. and Pačes T. (2004). Establishing timescales of groundwater residence times based on environmental tracer data: a study of the Turonian and Cenomanian aquifers of the Bohemian Cretaceous Basin, Czech Republic. In: 32nd International Geological Congress, Florence, Italy.

Hinsby K., Troldborg L., Purtschert R., Corcho Alvarado J.A., Hofer M. and Kipfer R. (2004). Radionuclides (3H, 85 Kr,) for evaluation of flow dynamics and temporal chemical trends in groundwater and surface water. Goldschmidt conference, Copenhagen, Denmark. Geochimica et Cosmochimica Acta, A493 ( abstr. ).

101

ACKNOWLEDGEMENTS I would like to thank all those who have supported me in different ways in this work throughout the years: family, friends and colleagues. I am deeply indebted to my supervisor Roland Purtschert, for giving me the opportunity to work in very interesting research projects and for all his support, guidance and help throughout this work. Many thanks to Thomas Stocker, head of the Division of Climate and Environmental Physics and advisor of my thesis, for offering me to work in the exciting field of environmental physics. Many thanks to all colleagues and ex-colleagues of the group: Joerg Rueedi, Manuel Kaegi, Joachim Elsig, Rolf Althaus, Hugo H. Loosli, Bernhard Lehmann, Veronika Obertaufer, Rudiger Schanda, Kurt Grossenbacher, for their friendly help. Thanks to all colleagues at the Division of Climate and Environmental Physics, for creating a nice and friendly atmosphere in the everyday life. I gratefully acknowledge financial support from the Swiss National Science Foundation.

102

103

Curriculum Vitae José Antonio Corcho Alvarado, born in November 28, 1971, in Fomento, Sancti Spiritus, Cuba Education 1976-1982 Primary school in Fomento, Sancti Spiritus, Cuba. 1983-1986 Secondary school in Fomento, Sancti Spiritus, Cuba. 1986-1989 High school in Santa Clara (2 years) and in Havana (1 Year), Cuba. 1989-1994 Study of Radiochemistry at the Institute for Nuclear Sciences and

Technology, Havana, Cuba. 2001-2005 Ph.D. Student at the research division of Climate and Environmental

Physics, Physics Institute, University of Bern, Bern Professional Appoinments 1995-2000 Research Assistant at the Department of Environmental Radiation

Protection, Center for Radiation Protection and Hygiene, Havana, Cuba. 2000-2001 Research Assistant at the Environmental Radioactivity Research Center,

Mathematical Department, University of Liverpool, United Kingdom. 2001-2005 Ph.D. Student at the Uuniversity of Bern Conferences and courses attendances 2004 International Workshop on the Application of Isotope Techniques in

Hydrological and Environmental Studies, IAEA/UNESCO, Paris, France. 2004 32nd International Geological Congress, Florence, Italy 2003 International Symposium on Isotope Hydrology and Integrated Water

Resources Management, International Atomic Energy Agency, Vienna, Austria

2002 Goldschmidt Conference 2002, Davos, Switzerland. May-June 2000 Summer Colloquium on the Physics of Weather and Climate:

“Chemistry-Climate Interactions”, Trieste, Italy. March 2000 Regional Training Course of the International Atomic Energy Agency:

“Preparedness and response in radiological emergency”, National Center for Nuclear Safety, Havana, Cuba.

May-July 1998 International Atomic Energy Agency Group Fellowship Training Course: “Advanced radiochemical techniques for the separation and determination of alpha and beta emitters in environmental and biological samples”, Chemistry Unit, IAEA Seibersdorf Laboratories, Vienna, Austria.

1998 IV Regional Congress of Radilogical and Nuclear Safety, Havana, Cuba