Werner Aeschbach-Hertig

Interpretation of dissolved atmospheric noble gases in natural waters

Several studies have used the temperature dependence of gas solubilities in water to derive paleotemperatures from noble gases in groundwaters. We present a general method to infer environmental parameters from concentrations of dissolved atmospheric noble gases in water. Our approach incorporates statistical methods to quantify uncertainties of the deduced parameter values. The equilibration temperatures of water equilibrated with the atmosphere under controlled conditions could be inferred with a precision and accuracy of 60.28C. The equilibration temperatures of lake and river samples were determined with a similar precision. Most of these samples were in agreement with atmospheric equilibrium at the water temperature. In groundwaters either recharge temperature or altitude could be determined with high precision (60.38C and 660 m, respectively) despite the presence of excess air. However, typical errors increase to 638C and 6700 m if both temperature and altitude are determined at the same time, because the two parameters are correlated. In some groundwater samples the composition of the excess air deviated significantly from atmospheric air, which was modeled by partial reequilibration. In this case the achievable precision of noble gas temperatures was significantly diminished (typical errors of 618C).

A paleotemperature record derived from dissolved noble gases in groundwater of the Aquia Aquifer (Maryland, USA)

Low 14C activities in groundwater of the confined part of the Aquia aquifer in southeastern Maryland suggest that most of this water infiltrated at least 30,000 years ago. However, radiocarbon contents of the dissolved inorganic carbon seem to be affected by isotopic exchange, possibly with secondary calcite deposits in the formation, leading to overestimated 14C ages. Whereas the geochemistry of the Aquia aquifer complicates the application of the widely used 14C dating method, the accumulation of radiogenic He seems to provide a viable alternative for establishing a chronology. The quasi-linear increase of He concentrations with flow distance observed in the Aquia aquifer can be explained entirely by accumulation of in situ produced radiogenic He. U and Th concentrations in Aquia sand were measured in order to determine the accumulation rate of 4He with sufficient confidence to establish a He time scale. Concentrations of dissolved atmospheric noble gases were used to derive mean annual ground temperatures at the time of infiltration. These noble gas temperatures (NGTs) clearly show the presence of water that infiltrated under much cooler conditions than at present. NGTs are correlated with chloride concentrations, corroborating the hypothesis that chloride variations in this aquifer constitute a climate signal. In contrast, the stable isotope ratios δ180 and δD do not provide a clear record of past climatic changes in the Aquia aquifer and the correlation between NGTs and stable isotope ratios is weak. The NGT record suggests that mean annual temperatures in this midlatitude coastal site during the last glacial maximum (LGM) were (9.0 ± 0.6) °C colder than during the Holocene. This difference is slightly lower than estimates derived from pollen data for this region, but considerably larger than the rather uniform cooling of about 5°C indicated by noble gas studies in more southern locations of North America. The larger cooling is ascribed to the influence of the Laurentide ice sheet, which at its maximum extension came as close as 250 km to our study site.

Infiltration of river water to a shallow aquifer investigated with 3H/3He, noble gases and CFCs

Noble gas isotopes ( 3 He, 4 He, Ne, Ar, Kr, Xe), tritium ( 3 H), chlorofluorocarbons (CFCs) and dissolved oxygen (O 2) were seasonally measured in a small groundwater system recharged by infiltration of river water at Linsental, northeastern Switzerland. All Groundwater samples contained an excess of atmospheric noble gases (‘excess air’) usually with an elemental composition equal to air. The concentrations of atmospheric noble gases in the groundwater were used to calculate the excess air component and the water temperature at recharge. The noble gas temperatures (NGTs) in the boreholes close to the river vary seasonally, however, the average NGT of all samples lies close to the mean annual temperature of the river water. Groundwater ages were calculated using the tritium/helium-3 ( 3 H/ 3 He) dating method. The water ages of the samples obtained near the river depend on the amount of recently infiltrated river water and are young during times of active river discharge. In contrast, the mean water age of about 3 years of the deep aquifer remained nearly constant over the sampling period. The observed CFC-11 (CFCl3) and CFC-12 (CF2Cl2) concentrations are significantly higher than the atmospheric equilibrium concentrations and therefore CFCs do not provide any direct information on the residence time of the groundwater. Nevertheless, the CFC excess in the groundwater shows a linear increase with the 3 H/ 3 He age. Additionally, both accumulation of radiogenic He ( 4 Herad) and O2 consumption are strongly correlated with residence time. All these correlations can be interpreted either in terms of mixing of recently infiltrated river water with older groundwater or in terms of accumulation/consumption rates. q 1999 Elsevier Science B.V. All rights reserved.

A comparison of groundwater dating with 81Kr, 36Cl and 4He in four wells of the Great Artesian Basin, Australia

The isotopic ratios 81 Kr/Kr and 36 Cl/Cl and the 4 He concentrations measured in groundwater from four artesian wells in the western part of the Great Artesian Basin (GAB) in Australia are discussed. Based on radioactive decay along a water flow path the 81 Kr/Kr ratios are directly converted to groundwater residence times. Results are in a range of 225^400 kyr with error bars in the order of 15% primarily due to counting statistics in the cyclotron accelerator mass spectrometer measurement. Additional uncertainties from subsurface production and/or exchange with stagnant porewaters in the confining shales appear to be of the same order of magnitude. These 81 Kr ages are then used to calibrate the 36 Cl and the 4 He dating methods. Based on elemental analyses of rock samples from the sandstone aquifer as well as from the confining Bulldog shale the in situ flux of thermal neutrons and the corresponding 3 He/ 4 He and 36 Cl/Cl ratios are calculated. From a comparison of: (i) the 3 He/ 4 He ratios measured in the groundwater samples with the calculated in situ ratios in rocks and (ii) the measured N 37 Cl ratios with the 4 He concentrations measured in groundwater it is concluded that both helium and chloride are most likely added to the aquifer from sources in the stagnant porewaters of the confining shale by diffusion and/or mixing. Based on this ‘working hypothesis’ the 36 Cl transport equation in groundwater is solved taking into account: (i) radioactive decay, (ii) subsurface production in the sandstone aquifer (with an in situ 36 Cl/Cl ratio of 6U10 315 ) and (iii) addition of chloride from a source in the confining shale (with a 36 Cl/Cl ratio of 13U10 315 ). Lacking better information it is assumed that the chloride concentration increased linearly with time from an (unknown) initial value Ci to its

Improving noble gas based paleoclimate reconstruction and groundwater dating using 20Ne/22Ne ratios

Abstract The interpretation of noble gas concentrations in groundwater with respect to recharge temperature and fractionated excess gas leads to different results on paleo-climatic conditions and on residence times depending on the choice of the gas partitioning model. Two fractionation models for the gas excess are in use, one assuming partial re-equilibration of groundwater supersaturated by excess air (PR-model, Stute et al., 1995) , the other assuming closed-system equilibration of groundwater with entrapped air (CE-model, Aeschbach-Hertig et al., 2000) . In the example of the Continental Terminal aquifers in Niger, PR- and CE- model are both consistent with the data on elemental noble gas concentrations (Ne, Ar, Kr, and Xe). Only by including the isotope ratio 20Ne/22Ne it can be demonstrated that the PR-model has to be rejected and the CE-model should be applied to the data. In dating applications 3He of atmospheric origin (3Heatm) required to calculate 3H-3He water ages is commonly estimated from the Ne excess presuming that gas excess is unfractionated air (UA-model). Including in addition to the Ne concentration the 20Ne/22Ne ratio and the concentration of Ar enables a rigorous distinction between PR-, CE- and UA-model and a reliable determination of 3Heatm and of 3H-3He water ages.

Injection of mantle type helium into Lake Van (Turkey): the clue for quantifying deep water renewal

Helium, neon and tritium concentrations have been measured to study long-term vertical mixing and deep water renewal in Lake Van (Eastern Turkey), the largest soda lake on Earth. Helium excesses were found in the water column of Lake Van, although neon concentrations were close to air saturation. The excesses of both isotopes are strictly correlated and increase with depth. In the bottom water, 4He supersaturation is about 20% and the corresponding 3He concentration is 2.5 times the air saturation value. The mean excess 3He/4He ratio, of 1 1.2 · 10−5, is slightly higher than the MORB ratio found in Lake Nemrut, a neighbouring volcanic crater lake, in which a large input of mantle helium was detected. Mantle helium accounts for the majority of the helium excesses in Lake Van, although part of the 3He excess is attributed to the presence of tritium. A one-dimensional lake model was used, which describes tritium input, vertical mixing and gas exchange, to reconstruct the evolution of the helium isotopes and the tritium. As a conclusion, the model, based on the measured 3He and 4He profiles, shows that: (1) the vertical exchange of deep water occurs within 1–2 yr; (2) the flux of mantle helium into the lake, averaged over the total cross-section of Lake Van, is 0.23–0.35 cm3 STP · m−2 yr−1 (2–3 · 1011 atoms · m−2 s−1). The present estimate of the global mantle helium flux averaged over the total surface area of the Earth is < 3.9 · 109 atoms · m−2 s−1; the flux into Lake Van would account for at least 0.04–0.06% of this.

Quantification of gas fluxes from the subcontinental mantle: The example of Laacher See, a maar lake in Germany

Abstract Vertical and horizontal distributions of helium and neon isotopes were measured in the water of Laacher See, a maar lake in the East Eifel volcanic field in Germany. Neon is in approximate atmospheric equilibrium throughout the lake, but the concentrations of both helium isotopes increase with depth and reach values of up to 10 ( 4 He) and 50 ( 3 He) times the atmospheric equilibrium concentration. The isotopic ratio of the helium excess, R ex = 3 He ex / 4 He ex , is (7.42 ± 0.03) · 10 −6 , i.e., R ex / R a = 5.36 ± 0.02, where R a = 1.384·10 −6 is the isotopic ratio in air. This indicates the presence of mantle-derived helium. Laacher See is thermally stratified during the summer. It acts as an almost perfect trap for injected gases. The helium concentration in the hypolimnion (deep water) approximately doubled from May to September 1991 yielding a mean 4 He-flux of (10 ± 2) · 10 12 atoms m −2 s −1 . This value is 30 times smaller than in Lake Nyos, Cameroon, but 20 times larger than in Crater Lake, USA. Release of bubbles of nearly pure CO 2 (>99%) observable in shallow waters at the eastern shore, as well as detected by divers in about 30 m depth, was identified as an important source of the helium excess in the lake. If carefully sampled by divers to avoid air contamination, the bubbles exhibit mantle influenced isotopic signatures not only for helium, but also for neon and argon ( 20 Ne/ 22 Ne = 9.92 ± 0.03; 40 Ar/ 36 Ar = 391 ± 4). Based on the elemental ratios of carbon, neon, and argon to 3 He, the fluxes of volatiles from the mantle into Laacher See are determined as 7.4·10 7 atoms m −2 s −1 for 3 He, 13·10 7 atoms m −2 s −1 for 20 Ne, 26·10 7 atoms m −2 s −1 for 361 Ar, and 6.4·10 17 molecules m −2 s −1 for CO 2 . Non-atmospheric neon was also found in the well “Wallenborn” in the West Eifel volcanic field with a 20 Ne/ 22 Ne-ratio as high as 11.21 ± 0.06.

ACCUMULATION OF MANTLE GASES IN A PERMANENTLY STRATIFIED VOLCANIC LAKE (LAC PAVIN, FRANCE)

Abstract Lac Pavin is a volcanic crater lake in the Massif Central (France), characterized by a permanent vertical density stratification resulting from a strong and persistent chemocline between about 60 and 70 m depth. The deep water below the chemocline forms the monimolimnion, in which most dissolved ions as well as helium, carbon dioxide, and methane are strongly enriched. The 3He/4He isotope ratio of the excess helium is (9.09 ± 0.01) · 10−6, or (6.57 ± 0.01) Ra. These findings clearly indicate a flux of mantle-derived magmatic gases into the monimolimnion. In order to derive the fluxes of magmatic volatiles into Lac Pavin, it is essential to understand the hydrologic characteristics of the lake. Previously published two-box models have assumed groundwater input at the lake bottom, a short residence time in the monimolimnion, and biogenic origin of the CO2. We propose an alternative model with a flux of magmatic gases, but not of water, into the monimolimnion, and a weak diffusive coupling between the monimolimnion and the overlying mixolimnion which leads to a long deep-water residence time (≈ 70 yr). We reassess the carbon budget of the lake and conclude that the major part of the accumulated CO2 in the monimolimnion is of magmatic origin. From the model-derived water exchange rates, we calculated a mantle 4He flux of (6 ± 2) · 1011 atoms m−2 s−1. This value lies near the lower end of the range found in comparable volcanic lakes. The flux of magmatic CO2 is estimated as (1.2 ± 0.4) · 10−7 mol m−2 s−1, which is also comparatively low. The monimolimnion appears to be in steady state with respect to these fluxes, therefore no further, potentially hazardous, accumulation of CO2 takes place.

The physical structure and dynamics of a deep, meromictic crater lake (Lac Pavin, France)

The characteristic feature of the physical structure of Lac Pavin is a distinct and permanent chemically induced density increase between about 60 and 70 m depth. This chemocline separates the seasonally mixed mixolimnion from the monimolimnion, which is characterized by elevated temperature and salinity as well as complete anoxia. Previously published box-models of the lake postulated substantial groundwater input at the lake bottom, and consequently a short water residence time in the monimolimnion and high fluxes of dissolved constituents across the chemocline. We present a new view of the physical structure and dynamics of Lac Pavin, which is based on the results of high-resolution CTD profiles, transient and geochemical tracers (tritium, CFCs, helium), and numerical modeling. The CTD profiles indicate the existence of a sublacustrine spring above rather than below the chemocline. A stability analysis of the water column suggests that vertical turbulent mixing in the chemocline is very weak. A numerical one-dimensional lake model is used to reconstruct the evolution of transient tracer distributions over the past 50 years. Model parameters such as vertical diffusivity and size of sublacustrine springs are constrained by comparison with observed tracer data. Whereas the presence of a significant water input to the monimolimnion can clearly be excluded, the input to the mixolimnion – both at the surface and from the indicated sublacustrine spring – cannot be accurately determined. The vertical turbulent diffusivity in the chemocline is well constrained to K ≈ 5×10-8 m2 s-1, about a factor of three below the molecular diffusivity for heat. Assuming thus purely molecular heat transport, the heat flow through the chemocline can be estimated to between 30 and 40 mW m-2. With respect to dissolved constituents, the very weak turbulent diffusive exchange is equivalent to a stagnant monimolimnion with a residence time of nearly 100 years. Based on these results and vertical concentration profiles of dissolved species, diffusive fluxes between monimolimnion and mixolimnion can be calculated. A large excess of helium with a 3He/4He ratio of (9.09 ± 0.01)×10-6 (6.57 Ra) is present in the monimolimnion, clearly indicating a flux of magmatic gases into the monimolimnion. We calculate a flux of 1.0×10-12 mol m-2 s-1 for mantle helium and infer a flux of 1.2×10-7 mol m-2 s-1 (72 t year-1) for magmatic CO2. The monimolimnion appears to be in steady state with respect to these fluxes.

Noble Gas Thermometry in Groundwater Hydrology

Concentrations of dissolved atmospheric noble gases in water constitute a thermometer, whose application to the groundwater archive provides a method of paleoclimate reconstruction. In addition, noble gases have found wide application as tracers in hydrogeology. This chapter reviews the historical development, the theoretical foundations, the sampling and analytical techniques, as well as the spectrum of applications of this important tool of tracer hydrology. A detailed account of currently available sampling techniques is given, as this information is of great practical importance but not fully available in the scientific literature. The analytical methods are better documented in the literature, although the many lab-specific details and constant development make it hard to provide an authoritative overview, so that this part is kept comparatively short. The focus of the chapter lies on the methods for data reduction and interpretation, which have undergone rapid and important development in the recent past. Nevertheless, in this respect still substantial research needs exist. Finally, this chapter provides an overview of applications of noble gases in groundwater hydrology, which range from the classical paleothermometry and the determination of other paleoclimate parameters such as humidity to various hydrological investigations, such as groundwater dating or the study of water origin and recharge conditions in hydrothermal, glaciated, alluvial, coastal, managed, and mountainous aquifer systems.