Methods for Dating Very Old Groundwater: Eastern Great Artesian Basin Case Study. T. Torgersen

Transcription

1 Methods for Dating Very Old Groundwater: Eastern Great Artesian Basin Case Study T. Torgersen Introduction The Great Artesian Basin (GAB) of Australia occupies one fifth of the Australian continent and 1.7 x 10 6 km 2. The region typically receives order 100mm of rain annually with a maximum of 600mm near the eastern margin recharge areas. Average annual runoff is much less than 10mm and most commonly less than 5mm. As a result, the land surface is marked with few perennial rivers and many ephemeral rivers; rivers draining into Lake Eyre in central Australia (60m below sea level) often dry up or are lost to infiltration before reaching the terminus in Lake Eyre. This dry, harsh climate of interior Australia limited the ability of the early settlers to explore and exploit the resources of this region. Beginning in the 1880s, artesian water was discovered in this vast interior region and eventually over 4700 wells were drilled into the primary artesian lithology, the Cadna-owie- Hooray aquifer (Habermehl, 1980). In addition, the overlying non-artesian Cretaceous aquifer was tapped with 20,000 pumped wells. This degree of exploitation led to significant drawdown of the water and by the1970s only 3100 wells remained flowing under artesian pressure and many of the non-flowing artesian wells were augmented by wind-powered pumps (Habermehl, 1980; Radke et al 2000). Recent estimates have placed the storage of water in the artesian system at 8.7 x 10 9 Mliters making it one of the largest artesian aquifers in the world (Radke et al., 2000). The last 30 years have seen a significantly increased scientific effort to understand the capacity of this system and its recharge although some efforts considerably predate this time period. In the following sections, a case study is developed for how the Eastern GAB was studied, quantified and evaluated. The case study is guided by hindsight that may or may not accurately reflect scientific thinking at the time. However, it is hoped this case study will enable the structure and components of this scientific study of very old groundwater to be placed in a logical framework so the precedents and the lessons of the GAB can be applied most successfully to other systems. The GAB is comprised of multiple subsystems (Fig. 1) with each system subject to its own developmental pressures and often its own unique recharge flow and discharge conditions. Jack (1923) recognized the convergence of two separate flow systems near Lake Eyre characterized by the high alkalinity waters from the east and high sulfate waters from the west. Here, the focus will be on the primary, long time-scale flow within the Central Eromanga Basin. The primary water use in this Central Eromanga Basin system is agriculture with an emphasis on cattle for beef and sheep for wool. The driving economics of the Basin is utilization of a sustainable resource, with conservation, rehabilitation and/or removal of some wells. As water in this Central Eromanga Basin has some of the longest flowlines and some of the slowest flow rates, it provides an endmember for the application, analysis and evaluation for dating very old groundwater. Habermehl (1980) sets the stage for detailed scientific investigation of the GAB system and the individual studies discussed below. To a large extent, the discussion in the following sections is an overview of the information contained within the individual studies with a historical context to elucidate how dating methods and field sampling were guided by existing information. The reader is referred to these primary literature sources for greater detail. Radke et al. (2000) provide detailed interpretation of the combined structural, hydraulic, geochemical, radiochemical and isotopic analyses and is an example of how information can be combined, integrated and utilized for the fullest understanding of very old groundwater basins. 1

2 Deposition, Structure and Hydrogeology of the Eastern GAB The primary deposition of the sediments comprising the aquifers and the aquitards of the GAB began with an intracratonic downwarp (Veevers, 1984; Shaw, 1990) of the Australian continent. During Jurassic-Cretaceous ( Ma) time, fluvial and fluivo-lacustrine sediments were deposited in the basin as a result of drainage from both the east and the west. Rising sea levels in Early Cretaceous times deposited shallow marine sediments over much of the previous fluvio-lacustrine sequences with maximum marine inundation Ma (Struckmeyer and Brown, 1990). Major uplift of the continent commenced around 95Ma ago and reached 200m over a time span from Ma (Gurnis et al., 1998). These fluvio-lacustrine and marine sediments were converted to a hydrogeologic groundwater system by regional uplift in the eastern margins between 10-15Ma (Senior and Habermehl, 1980) and 5Ma ago (Toupin et al., 1997). The Eastern margin of the present GAB was marked by Jurassic and Cenozoic volcanism (Duncan and McDougal, 1986). This hydrogeologic history will become important in the subsequent evaluation because the residence time of groundwaters in the GAB (order 1Myr) may or may not have had time to flush connate waters from the Cadna-owie-Hooray sandstone and/or the overlying marine connate waters in the Cretaceous aquifer. Secondly, because Cenozoic volcanism was common along the eastern margin of the GAB, specific locales have been spiked with components representing contact with the volcanic centers (e.g., low 87 Sr/ 86 Sr groundwater, Collerson et al, 1988; and high 3 He/ 4 He ratios, Torgersen et al., 1987). After the uplift of the eastern margins (10-15Ma, Senior and Habermehl, 1980 or 5Ma ago, Toupin et al., 1997), it is likely that Australia was latitudinally positioned to benefit from coastal rains in summer along the northeast slopes of the Great Dividing Range that include the eastern recharge zone as well as some monsoonally driven rains that can penetrate deeper into the continental interior. The paleoclimatic records of Lake Eyre (Magee et al 2004) and the fluvial records of the interior (Nanson et al., 1992; Nanson and Price, 1999) suggest a clear trend to aridity from 130Ka to the present. The analysis of Torgersen et al. (1985, 1988) suggests that Holocene/Pleistocene climates in the Gulf of Carpentaria region (northern sections of the GAB) were similar to the present. Setting the Stage The synthesis of Habermehl (1980) summarized the structure of the GAB, its primary conductive aquifers and its fundamental chemical characteristics. Hydraulic conductivities were found to be generally in the range of 0.1 to 10 m d -1 with transmissivities of the order m 2 d -1. Pressure heads had been mapped and the 1970 potentiometric contours defined the loss of potentiometric head and storage relative to the reconstructed 1880 (predevelopment) potentiometric head (Fig. 2). This reconstruction and the bore measured hydraulic conductivities enabled the flow model of Seidel (1980) to define flow directions and estimates of flow velocities for broad sections of the GAB (Fig. 3). Regional scale applications of the model were developed (Habermehl and Seidel, 1979) and model application and calibration were reported in Seidel (1978a,b). Subsequent updates of flow models have preserved the basic characteristics while defining details. The basic bidirectional flow pattern that defines the eastern GAB and the western GAB was confirmed by the distribution of water chemistry types and confirmed the preliminary analysis of Jack (1923). The original hydrodynamic model of Seidel (1978b) appears to have been utilized for a considerable period of time to be slowly replaced by a MODFLOW model described by Welsh and Doherty (2005) and Welsh (2006). Stable Isotope and 14 C Measurements Airey et al (1979) reported stable isotope (D, 18 O) measurements from 69 wells within the GAB and showed that the composition was consistent with a rainwater source and minimal interaction with the host rock. The minimal variation in the stable isotope ratios along apparent flowlines indicated a minimal climatic signal over the timescale represented by the transect and 2

3 the time interval allowed by the well spacing. Transects of Na, Cl and HCO 3 indicated some covariation down flow lines (Fig. 4) that may or may not have been defined by rainfall variations (Kershaw 1978; see also Kershaw 1994, Kershaw and Nanson, 1993). 13 C ratios were shown to co-vary with HCO 3 -. This variation was interpreted as a two component system where HCO 3 - introduced during recharge is increasingly reacted and diluted with dissolved CaCO 3 of marine origin along the flow path. The 1982 Field Work: Stannum-to-Innamincka and Bonna Vista-to-Thargomindah The above studies set the stage for the field work of 1982 which focused on sampling the Stannum-to-Innamincka and Bonna Vista-to-Thargomindah flowlines (Fig. 5) identified by the potentiometric head and previous chemistry and stable isotope profiles. This field work effort included sampling for 14 C, 36 Cl and 4 He that had the potential to quantify flow rates and groundwater ages within the eastern GAB in addition to chemistry and stable isotopes. Stable isotope measures from the 1982 field work (Airey et al., 1983, Calf and Habermehl, 1984) confirm flow lines defined from potentiometric heads (Fig. 6). Numerous 14 C samples collected minimally down the flowline from the recharge areas confirmed the presence of relative young waters (for the GAB) that contained measureable modern carbon (>1%). The transects of 13 C generally confirm flow directions inferred from the reconstructed potentiometric surface when considering the dominate down flow reaction includes dissolution of marine carbonates identified in earlier work (Airey et al., 1979). The deuterium contours show tongues of isotopically heavier water (Airey et al 1983) that are consistent with the flow directions inferred previously. Bentley et al (1986) evaluated the 36 Cl groundwater dating method and developed its potential. Extensive 14 C samples in the recharge area were used to define the initial condition for 36 Cl within the confined portion of the basin. Hydrologic ages were used to select wells to define the recharge condition for 36 Cl; 14 C was used to confirm the appropriateness of those wells. An initial groundwater 36 Cl is the result of (1) 36 Cl produced (a) in the atmosphere by spallation reactions on 40 Ar, (b) 36 Cl production in surface rocks by cosmic ray spallation of K and Ca, and (c) 35 Cl activation by cosmic ray neutrons. This initial 36 Cl content is diluted with Cl carried by rains and leached from surface soils. An initial 36 Cl/Cl ratio of 110 x was estimated from wells containing modern 14 C with an initial 36 Cl concentration of 12 x 10 7 atoms L -1. In addition, one sample was collected from the overlying Cretaceous aquifer to identify the role of Cretaceous water contamination of the primary Cadna-owie-Horray aquifer water. In the confined aquifer, 36 Cl would then decay (half life of 301kyrs) to radioequilbrium with 36 Cl produced in situ that gives an equilibrium 36 Cl/Cl ratio of 5.7 x or a 36 Cl concentration of 0.55 x 10 7 atoms L -1 (see Bentley et al or Phillips et al (this guidebook) for details). The Bentley et al (1986) evaluation of the Cl samples identified dissolution of dead Cl (presumably from ancient marine deposits) and /or addition of Cl by other means to be a significant diluent for 36 Cl. Consequently, 36 Cl ages are calculated by one of three endmember equations. The comparison in Fig. 7 generally confirms the validity of the apparent groundwater ages through the calculation of the apparent tracer ages. It is noted however, that the calculated apparent 36 Cl tracer ages do not account for mixing of waters via flow convergences and it is noted that the flow lines for 1982 expedition were selected to minimize flow convergence. The helium isotope collection of 1982 sought to verify the potential for 4 He dating of groundwater as proposed by e.g. Davis and Deweist (1966) and within the background of work that had already been conducted by e.g. Marine (1976, 1979), Andrews (1977) Andrews and Lee (1979), Heaton and Vogel (1979) and Torgersen (1980). Torgersen and Clarke (1985), working with knowledge that the apparent hydraulic groundwater ages had been confirmed by 36 Cl measures, showed that young groundwaters of the GAB increased their 4 He concentration 3

4 at a rate that was consistent with a 4 He source derived from local sources within the sandstone. However, further along flowlines in older groundwaters, the rate at which 4 He increased was 74x the rate of in situ production (Fig. 8). 222 Rn measures in the 1982 groundwater samples showed no significant increase in the 222 Rn activity in the older portion (high 4 He flux region) of the flow line and thus no evidence that a stronger, local, in situ source of 4 He was acting to increase the rate of 4 He introduction. These helium results indicated the predominance of an external source of 4 He that was consistent with a basal flux derived from U,Th series decay in the crust beneath the GAB. Most importantly, the inference of a large basal flux of 4 He was supported by the simple hydraulic flow model of Torgersen and Ivey (1985) that accounted for the delay time over which the enhanced rate of 4 He accumulation (the basal flux) was observed. As a result, the apparent 4 He tracer ages were never calculated from the GAB 4 He measurements. Helium isotope measures from the 1982 (and 1985) expedition (Torgersen et al. 1987) (Fig. 9) showed the isotope composition of helium accumulating along most flow paths was consistent with in situ production and indistinguishable from an expected deep crustal production. However, some paths showed 3 He/ 4 He ratios that were significantly elevated and required a mantle source of helium. These data suggest a point source input of a 3 He-rich component that is diluted down flow by mixing and addition of helium from the in situ and basal fluxes. This stable helium tracer thus defines a unique source region within the recharge area that can be tracked downstream to define a flow path. This 3 He-rich tracer thus further defines and supports the validity of the flow paths estimated by Habermehl (1980) from the potentiometirc surface. In addition, Collerson et al. (1988) showed that recharge area groundwaters were spiked with low 87 Sr/ 86 Sr ratios (and a higher Ca/Sr ratio) as a result of volcanic Cenozoic intrusions and Jurassic and younger volcanic activity that was active along the eastern margin of what would become the GAB (Fig. 10). The downstream dilution of this recharge area 87 Sr/ 86 Sr spike again serves to verify the validity of the flow lines identified from potentiometirc surfaces. Torgersen et al (1989) show that 40 Ar results appear to follow the same dynamics as 4 He in the Great Artesian Basin and are also consistent with in situ production followed by an external basal flux. That the 4 He and 40 Ar data are generally explained by similar processes but that the 4 He/ 40 Ar ratio remains quite variable suggest that these deep crustal fluxes are decoupled. Decoupling of 4 He and 40 Ar from their U-,Th-series and K sources is likely accomplished by processes at the grain scale including recoil, diffusive release and dissolution as well as large scale fracturing that will additional separate 4 He and 40 Ar (Torgersen and O Donnell, 1991). Given the relative signal strength, the 40 Ar results suggest that 4 He results are more useful for very old groundwater basins. The 1985 Field Work: Fairlight Trust-to-Clayton, Athol-to-Mutti Mutti and Mt. Crispe-to-Curdimurka Continuing to build a base knowledge of the flow paths and flow rates of the eastern GAB, 1985 field work addressed additional flow lines in the eastern GAB as well as a flow line in the western GAB to be discussed elsewhere. The flow lines sampled explore the northern area of the basin and the role of converging flowlines as a complication in the application and use of apparent tracer age calculations. Torgersen et al (1991) identify an increase in Cl and SO 4 down flowline for the Athol-to- Mutti Mutti group which also show 36 Cl ages on the high side of hydraulic ages (see Fig. 11). A discontinuous 36 Cl evolution is distinct near Meerabooka where the East-West transect should intersect the North-South transect. Distinct addition of Cl is also evident near the end of the North-South transect (Clayton) as flowing waters rise from depth and are subject to mixing. The impact of such mixing on the calculated apparent tracer age was estimated based on linear mixing along an exponential decay curve and for this flow line was suspected to cause minimal offset of the calculated apparent tracer age. 4

5 Torgersen et al (1992) combine the helium isotope data of 1982 and 1985 and show the area of highest He concentration to be coincident with the oldest water and the lowest (and near radioequilibrium) 36 Cl concentration (Fig. 12). The helium data also show the impact of the western recharge area and the limited recharge along the northwestern recharge. Modelling 36 Cl and 4 He The results of the 1982 and 1985 fieldwork for both 36 Cl and 4 He presented new data and new methods for the evaluation of very old groundwater. Discussion in the literature regarding the approach, assumptions and necessary conditions for the application of these methods (Bentley et al., 1986) is often the case. Andrews and Fontes (1992,1993; see also Mazor, 1992) presented an evaluation of GAB 36 Cl data that explained variations to be the result of mixing between flowline waters and deep, dead waters but presented no mechanisms by which such mixing could occur (near instantaneously) over such a broad region. Torgersen and Phillips (1993) refuted their argument showing that the conditions necessary to generate the mixing lines proposed by Andrews and Fontes (1993) necessitated unreasonable endmembers and required virtually no radiodecay of 36 Cl and thus no time to be transported. The exchange highlighted the need to evaluate the role of mixing in the interpretation of apparent 36 Cl ages. Bethke et al (2000), Park et al (2002) and Radke et al (2000) included such effects in their interpretations. Mazor (1995) and Mazor and Nativ (1992;1994) suggested that helium in the older portions of the GAB can be explained as the result of water trapped in stagnant zones. However, among a number of points raised by Torgersen (1994), such stagnation is inconsistent with the continuing evolution of 87 Sr/ 86 Sr (Collerson et al., 1988) and would require not only that flow stopped in stagnant zones but also that geochemical processes of mineral dissolution were halted in a unique increasing sequence. The discussion highlights the need to examine all the ancillary physical, chemical and isotopic data available including the progression in geochemical reactions, stable isotopes, and dating tracers, to establish a means to validate the basic assumptions used for calculating groundwater flow velocity and age. This synthetic analysis of data (all together) was adopted by Radke et al (2000) for their benchmark analysis. Zhao et al (1998) evaluate the role of hydrodynamic flow in a typical confined aquifer on the distribution and accumulation of 4 He in the flowing groundwaters. Their modeling suggests flow divergence and convergence and especially upflow in a discharge zone can induce significant mixing. The study thus begins the process of questioning the application of piston flow mechanics to either hydraulic ages or apparent groundwater dating methods. Bethke et al (1999, 2000) undertake a full scale modeling of the Great Artesian Basin flow system including the 4 He accumulation and conclude that the distribution of 4 He is controlled not only by lateral flow but also by upflow and downflow regions driven by the basin topology and structure. The modeling effort shows that 4 He and groundwater flow can be reconciled quantitatively within 3D modeling. Similar modeling of the 36 Cl dynamics (Park et al, 2002) again confirm the need to consider 3D flow in aquifer systems and the role of upflow, downflow, and converging flow in the reconciling of hydraulic ages (apparent groundwater agers and apparent tracer ages. The summary manuscript (Bethke and Johnson, 2008) reiterates these principles. Geochemical Modelling of groundwater reaction paths Herczeg et al. (1991) conducted a detailed analysis of the water geochemistry from the 1985 field work using PHREEQE (Parkhurst et al 1980). That study confirmed the increase in TDS along flowlines and the distinct chemical differences between the Eastern GAB (Na-HCO3 type) and the Western GAB (Na-SO4-Cl type) waters. Fig. 13,14 show clearly the strong increase in alkalinity, DIC and 13 C along eastern GAB flowlines as well as the slowly increasing Na, Cl and TDS and the decreasing Ca and Mg. Herczeg et al (1991) discuss the increase in TDS as a result of (1) mixing of flowline waters with saline waters within the deeper portions of 5

6 the basin, (2) ion filtration through shale membranes, and (3) dissolution of evaporates carbonate minerals or other minerals in situ. They infer that mixing with saline waters would be most likely from the overlying Cretaceous marine sediments; but given the leakage of water from the lower (and artesian) Cada-owie-Hooray aquifer to the upper (nonartesian) Cretaceous aquifer, that mixing is likely to be slight. They further infer that ion filtration is likely to be minimal based on the 36 Cl studies of Bentley et al (1986) and conclude that evapotranspiration in the recharge areas in combination with mineral dissolution along flowlines represents the most plausible mechanism for the chemical evolution observed in Fig. 13,14. The eastern GAB flow line trends of increasing Na alkalinity and DIC together with the decreasing Ca and Mg were evaluated in terms of a model proposed by Blake (1989) which include (1) dissolution of carbonate resulting in increased Ca, Mg and HCO3, (2) cation exchange of Ca and Mg for Na in clay minerals, followed by (3) a reaction of Na with kaolinite to produce Na-smectite. Reasoning that the downflow increase in alkalinity can only be accomplished by carbonate dissolution in the presence of high pco 2, they conclude that CO 2 respiration by plants in the recharge zone can drive these reactions to an isotopic composition of DIC that reaches from an initial -15%o to approximately -12%o. However, to enable the reactions to continue requires the addition of CO 2 as a result of anaerobic fermentation and reduction of CO 2 to methane. The 2000 Benchmark and Synthesis Following the seminal paper of Habermehl (1980), multiple Australian agencies cooperated in the assembly and synthesis of extant knowledge with the publication of Radke et al (2000). This benchmark synthesizes flow models, geochemical reaction and transport models and the apparent groundwater dating of the Cadna-owie-Hooray aquifer waters to infer the timescale of recharge and the in situ flow rates within the Great Artesian Basin. It further utilizes distinctive hydrogeochemical anomalies to infer details of flow in this aquifer. In the few paragraphs that follow, their findings are summarized especially as they relate to the case study presented above and the additional detail they deduce through the use of modern hydrodynamic codes and hydrogeochemical codes. The mapped occurrences of modern 14 C and 36 Cl (Fig. 15,16) are consistent with previous work. The location of very old groundwater (near equilibrium values of 36 Cl) is consistent with the mapped regions of very high 4 He (Fig. 12). Recharge, flow direction and discharge zones (Fig. 17,18) reported by Radke et al. (2000) are largely similar to the original diagrams of Habermehl (1980). Recharge is estimated to be of the order 10 6 ML yr -1 with an unknown proportion distributed between rainfall and river infiltration. This current recharge rate may be enhanced as a result of the previous century of drawdown. Natural discharge is order 0.45 x 10 6 ML yr -1 with most discharge via upward leakage and minimal discharge occurring via outlet springs (see Table 1). However, it is noted that crude estimates of residence time for water within the Cadna-owie-Hooray Aquifer of the GAB based on the rates of natural discharge and the estimated storage of the order 8.7 x 10 9 ML yield residence times ages of the order 17,400years in stark disagreement with the 36 Cl and hydraulic estimates. The climate reconstructions of e.g. Magee, et al (2004) suggest a decline in rainfall over Pleistocene time and discharge springs have experienced a significantly reduced head over the past century. Thus, it is not anticipated that discharge rates contribute significantly to this residence time discrepancy. However, the range of estimates for GAB storage is significant (GAB Information Pamphlet, storage is 8.7 x 10 9 ML (Habermehl, pers. comm. 2009); QDNRW, storage is 6.5 x ML, Hillier and Foster, 2002; Habermehl and Seidel (1979) - storage is ML). Using the Habermehl and Seidel (1979) estimated storage (10 12 ML) would yield discharge-based residence times of 2Myr. The precision and the precise meanings of these differences in storage are not clear but it may be the difference between in situ storage and recoverable storage. 6

7 Table 1: Water Balance for the Great Artesian Basin after Radke et al (2000) NATURAL RECHARGE rainfall plus rivers 1.0 million ML yr-1 DISCHARGE upward leakage 0.45 million ML yr-1 STORAGE natural springs anthropogenic million ML yr million ML yr million ML The Eastern GAB is characterized by increasing TDS and alkalinity along flow lines. Chemical anomalies are noted along the Birdsville-Diamantina Track (BTR) with additional anomalies occurring as high chloride along the Eulo-Nabine Ridge area. Groundwater temperatures increase with depth of the aquifer and the Central Eromanga Basin shows a distinct increase in Na especially when expressed as sodium excess (e.g., [Na]-[Cl]). High fluoride concentrations occur over basement highs where the aquifer is possibly in contact with the basement. The stable isotopic signature of the groundwater suggests major summer rainfall in the eastern recharge zone is responsible for a majority of recharge although evapotranspiration before recharge may be significant. The lack of significant variation in stable isotopes throughout the basin precludes identification of a paleoclimatic signature that may be the result of little climatic variation and/or dispersive mixing. Hydrodynamic models of flow calibrated against 14 C in the recharge area suggest flow rates ranging from 1.4 to 2.9 m yr -1. Flow rates calculated by hydrodynamic models and estimates of the 36 Cl controls indicate several areas with flow rates as low as m yr -1 (see Fig. 17). Very slow flow regions are areas of predominate water loss via upward leakage although the hydrochemical signatures in slow flow areas also indicate the possibility of downward diffusion of some ions (Fig. 18). It would appear that discharge at springs (11%) is small compared to discharge via upward leakage (89%) for a total discharge of 450,000 ML yr -1. Overall, the benchmark paper of Radke et al (2000) is in general agreement with earlier studies including the flow rate and age of groundwater inferred from 14 C, 36 Cl and 4 He studies. However, their detailed analysis appears to identify high flow regimes with lower salinities in shallow depth of burial regions. NETPATH (Plummer et al. 1994) modeling of GAB geochemistry agrees with the overall net reactions postulated by Herczeg et al (1991) but suggests that carbonate dissolution reactions are the primary driving mechanism rather than smectite-kaolinite transformations. Increased TDS with increasing depth is best explained by diffusive overprints where diffusion rates and chemical reactions increase with increasing temperature. Thermal convection (Pestov, 1999) is not substantiated. Continuing Work in the GAB As a result of the large data base available for the GAB, several anomalies have been identified within the recharge areas (e.g. Torgersen et al 1987 ; Collerson et al 1988), the TACEM anomalies (Radke et al 2000) as well as along flow paths. In particular, the groundwaters along the Birdsville Track are characterized by higher temperatures and higher total dissolved solids. This trend is inferred to be the result of mixing induced by the rapid rise in basement level based on geothermometry studies of Pirlo (2004). Evidence for this mixing of waters is provided by the discord among silica geothermometers and cation geothermometers. Because silca geothermometers re-equilibrate relatively quickly compared to cation geothermometers, the low Si temperatures and the high cation temperatures indicate mixing from deeper, hotter (faster reactions) regions and shallow regions. 7

8 O Shea and Jankowski (2007) have performed hierarchical cluster analysis on the chemical compositions in the Coonamble embayment (southeastern GAB and an area not discussed above) and compared them with traditional Piper diagrams (Piper, 1944), Durov diagrams (Chilkigar, 1956) and a Chada (1999) modification of a Piper analysis. Their cluster analysis indicates three main geochemical water types that can be associated with processes of ion exchange, precipitation, and mixing from different sources. They were also able to identify an anomalous sample that is suspected to be influenced by magmatic sources of CO 2. This indication of magmatic CO 2 is consistent with the location of other anomalies observed in 87 Sr/ 86 Sr (Collerson et al 1988) and 3 He/ 4 He Torgersen et al (1987) that are geographically associated with the eastern recharge zone that was subject to considerable Cenozoic and pre- Cenozoic volcanism (Duncan and McDougal, 1986). References Airey, P.L., Calf, G.E., Campbell, B.L., Habermehl, M.A., Hartley, P.E., & Roman, D., Aspects of the isotope hydrology of the Great Artesian Basin, Australia. In: Isotope Hydrology 1978, 1, P Proceedings International Symposium on Isotope Hydrology - International Atomic Energy Agency and United Nations Educational, Scientific and Cultural Organisation, Neuherberg, Fed. Rep. Germany, June International Atomic Energy Agency, Vienna, Airey, P.L., Bentley, H., Calf, G.E., Davis, S.N., Elmore, D., Gove, H., Habermehl, M.A., Phillips, F., Smith, J., & Torgersen, T., Isotope hydrology of the Great Artesian Basin, Australia. In: Papers of the International Conference on Groundwater and Man, Sydney, 5 9 December Australian Water Resources Council Conference Series, Australian Government Publishing Service, Canberra, (8): Andrews, J.N Radiogenic and Inert gases in groundwater. In, Proc. Second Intl. Symposium on Watewr Rock interaction, (Ed. H. Paquet and Y. Tzrdy) Strasbourg, France p[p., Insitute de Geologie, Universite Louis Pastuer. Andrews, J.N. and P.J. Lee Inert Gases in groundwater from the Bunter Sandstone of England as indictors of age and paleoclimate trends. J., Hydrol. 41: Andrews, J.N., & Fontes, J.C., Importance of in-situ production of chlorine-36, argon-36 and carbon-14 in hydrology and hydrochemistry. P In: IAEA (eds), Isotope Techniques in Water Resources Development, Vienna Andrews, J.N., & Fontes, J.C., Comment on Chlorine 36 Dating of very old groundwater, Further results on the Great Artesian Basin, Australia by T. Torgersen et al. Water Resources Research, 29 (6): Bentley, H.W., Phillips, F.M., & Davis, S.N., Chlorine-36 in the terrestrial environment. P In: Fritz, P., & Fontes, J.C. (eds). Handbook of Environmental Isotope Geochemistry, Vol. 2. Amsterdam, Elsevier. Bentley, H.W., Phillips, F.M., Davis, S.N., Habermehl, M.A., Airey, P.L., Calf, G.E., Elmore, D., Gove, H.E., & Torgersen, T., Chlorine 36 dating of very old groundwater. The Great Artesian Basin, Australia. Water Resources Research, 22 (13): Bethke, C.M., Xiang Zhao, & Torgersen, T., Groundwater flow and the 4 He distribution in the Great Artesian Basin of Australia. Journal of Geophysical Research, 104, B6: Bethke, C.M.,T. Torgersen, J Park, The age of very old groundwater: Insights from reactive transport models. Journal of Geochemical Exploration, Jour Geochem Explor, 69-70,1-4. Bethke, C. and T.M. Johnson Groundwater Age and Groundwater Age Dating. Annual Reviews Earth Planet. Sci. 36:

9 Blake, R The origin of high sodium bicarbonate waters in the Otway Basin, Victoria, Australia. In, D.L. Miles (Ed.) Water-Rock Interactions. Balkema, Rotterdam, pp Calf, G.E., & Habermehl, M.A., Isotope hydrology and hydrochemistry of the Great Artesian Basin, Australia. In: Isotope Hydrology 1983, P Proceedings International Symposium on Isotope Hydrology in Water Resources Development, International Atomic Energy Agency and United Nations Educational, Scientific and Cultural Organisation, Vienna, Austria, September International Atomic Energy Agency, Vienna, Chada, D.K A proposed new diagram for geochemical classification of natural waters and interpretation of chemical data. Hydrogeology Jour. 7: Chiligar, G.V Durov s classification of natural waters and chemical composition of atmospheric precipitation in USSR; a review. Trans. Amer. Geophys. Un. 37(2): Collerson, K.D., Ullman, W.J., & Torgersen, T., Ground waters with unradiogenic 87Sr/86Sr ratios in the Great Artesian Basin, Australia. Geology, 16, p Davis, S.N. and Deweist, R.J.M Hydrology. Wiley, 463pp. Duncan, R.A. and I. McDougal Time-space relationships for Cainozoic Intraplate Volcanism in Eastern Australia, the Tasman Sea and New Zealand. In, J.W. Johnson (ed.), Intraplate Volcanic Activity in Australia and New Zealand. Cambridge University Press, Cambridge. pp Fontes, J.-C., & Andrews, J.N., Comment on Reinterpretation of 36Cl data: physical processes, hydraulic interconnections and age estimations in groundwater systems by E. Mazor. Applied Geochemistry, 8: Gurnis, M., Muller, R.D., & Moresi, L., Cretaceous vertical motion of Australia and the Australian-Antarctic Discordance. Science, 279: Habermehl, M.A., The Great Artesian Basin, Australia. BMR Journal of Australian Geology & Geophysics, 5: Habermehl, M.A., & Seidel, G.E., Groundwater resources of the Great Artesian Basin. In: Hallsworth, E.G., & Woodcock, J.T., (eds). Proceedings of the Second Invitation Symposium Land and Water Resources of Australia Dynamics of utilisation, Australian Academy of Technological Sciences, Sydney, 30 October-1 November Australian Academy of Technological Sciences, Melbourne, p Heaton, T.H.E. and J.C. Vogel Gas concentrations and ages of groundwaters in Beaufort group sediments, South Africa. Water S.A. 5(4): Herczeg, A.L., Torgersen, T., Chivas, A.R., & Habermehl, M.A., Geochemistry of ground waters from the Great Artesian Basin, Australia. Journal of Hydrology, 126: Hillier, J.R. and L. Foster THE GREAT ARTESIAN BASIN - IS THERE A SUSTAINABLE YIELD? International Association of Hydrologists Conference, Darwin, Abstacts with program Jack, R.L., The composition of the waters of the Great Australian Artesian Basin in South Australia and its significance. Transactions and Proceedings of the Royal Society of South Australia, 47: Kershaw, A.P Record of the last interglacial-fglacial cycle from northeastern Queensland. Nature 272: Kershaw, A.P., Pleistocene vegetation of the humid tropics of northeastern Queensland, Australia. Palaeogeography, Palaeoclimatology, Palaeoecology, 109: Kershaw, A.P., & Nanson, G.C., The last full glacial cycle in the Australian region. Global and Planetary Change, 7: 1 9. Magee, J., Miller, G.H., Spooner, N.A., and Questiaux, D Continuous 150kyr monsoon record from Lake Eyre, Australia: insolation forcing implications and unexpected Holocene failure. Geology 32(10): doi /G

10 Marine, I.W Geochemistry of groundwater at the Savannah River Plnt. DP1356. E.I. DuPont de Nemours Co. Savannah River Laboratory NTIS. Marine, I.W The use of naturally occurring helium to estimate groundwater velocities for studies of geological storage of radioactive water. Water Resources Res. 15(5) Mazor, E., Reinterpretation of Cl-36 data: physical processes, hydraulic interconnections, and age estimates in groundwater systems. Applied Geochemistry, 7: Mazor, E., Stagnant aquifer concept Part 1. Large-scale artesian systems Great Artesian Basin, Australia. Journal of Hydrology, 173: Mazor, E., & Nativ, R., Hydraulic calculations of groundwater flow, velocity and age: Examination of the basic premises. Journal of Hydrology, 173: Mazor, E., & Nativ, R., Stagnant groundwater stored in isolated aquifers: implications related to hydraulic calculations and isotope dating Reply. Journal of Hydrology, 154: Nanson, G.C., & Price, D.M., The drying of Australia during the Middle and Late Quaternary. Quaternary Long Records Workshop, Geology Department, Australian National University, Canberra, Program and Abstracts. Nanson, G.C., Price, D.M., & Short, S.A., Wetting and drying of Australia over the past 300 Ka. Geology, 20: O Shea, B. and J. Jankowski Detecting subtle hydrochemical anomalies with multivariate statistics: an example from homogeneous groundwaters in the Great Artesian Basin, Australia. Hydrological Processes 20: Park, J., C. Bethke, T. Torgersen, T. Johnson Transport Modeling applied to the interpretation of groundwater 36 Cl age. Water Resources Res. 38: /2001 WR Parkhurst, D.L., L.N. Plummer, and D.C. Thorstenson PHREEQE a computer program for geoch=emcial reactions. U.S. Geological Survey Water Resources Investigation 80-96, 210pp. Pestov, I., Thermal convection in the Great Artesian basin: existence, importance and modelling requirements. Unpublished draft report to the GAB Technical Working Group. Land & Water Sciences Division, Bureau of Rural Sciences, Canberra, 16 p. Phillips, F.M., Comment on Reinterpretation of 36Cl data: physical processes, hydraulic interconnections and age estimations in groundwater systems by E. Mazor. Applied Geochemistry, 8: Piper, A.M A Graphic procedure in the geochemical interpretation of water analyses. Trans. Amer. Geophys. Un. 25: Pirlo, M.C Hydrogeochemistry and geothermometry of thermal groundwaters form the Birdsville Track Ridge, Great Artesian Basin, South Australia. Geothermics 33: Plummer, L.N., Prestemon, E.C., & Parkhurst, D.L., An interactive code (NETPATH) for modeling NET geochemical reactions along a flow PATH. Version 2.0. U.S. Geological Survey, Water Resources Investigations Report , 139 p. Radke, B.M., J. Ferguson, R.G. Cresswell, T.R. Ransley, M.A. Habermehl Hydrochemistry and implied hydrodynamics of the Cada-owie-Hooray Aquifer Great Artesian Basin. Bureau of Rural Sciences, Australia. ISBN Seidel, G.E.,1978a. Hydraulic calibration of the GABHYD model for the Great Artesian Basin. Bureau of Mineral Resources, Australia, Record 1978/12. Seidel, G.E., 1978b. Operating manual for the GABHYD model. Bureau of Mineral Resources, Australia, Record 1978/70. Seidel, G.E., Application of the GABHYD groundwater model of the Great Artesian Basin, Australia. BMR Journal of Australian Geology & Geophysics, 5:

11 Shaw, R.D The seismo-tectonics of southeastern Queensland. In, Finlayson, D.M. (ed.) The Eromanga-Brisbane transect: a guide to basin development across Phanerozoic Australia in Southern Queensland. Bureau of Mineral Resources. Geology and Geophysics, Australia Bulletin 232. Struckmeyer, H.I.M., & Brown, P.J., Australian sealevel curves. Part 1, Australian inundation curves. Bureau of Mineral Resources, Australia, Record 1990/11, 67 p. Torgersen, T Controls on porefluid concentration of 4He and 222Rn and the calculation of 4He/222Rn ages. J. Geochem. Explor. 13: Torgersen, T., Hydraulic calculation of groundwater flow velocity and age: Examination of the basic premises Comment. Journal of Hydrology, 154: Torgersen, T., & Clarke, W.B., Helium accumulation in groundwater, I: An evaluation of sources and the continental flux of crustal 4He in the Great Artesian Basin, Australia. Geochimica et Cosmochimica Acta, 49: Torgersen, T., & Ivey, G.N., Helium accumulation in groundwater. II: A model for the accumulation of the crustal 4He degassing flux. Geochimica et Cosmochimica Acta, 49: Torgersen, T. and J. O'Donnell The degassing flux from the solid Earth: release by fracturing. Geophys. Res. Lett. 18: Torgersen, T. and F.M. Phillips Reply to "Comment on 'Chlorine-36 Dating of very old groundwater 3. Further results on the Great Artesian Basin, Australia' by T. Torgersen, et al." by J.N. Andrews and J.-C. Fontes. Water Resources Res. 29: Torgersen, T., M.R. Jones, A.W. Stephens, D.E. Searle and W.J. Ullman Late Quaternary hydrologic changes in the Gulf of Carpentaria. Nature 313: Torgersen, T., Habermehl, M.A., & Clarke, W.B., Helium isotopic evidence for recent subcrustal volcanism in eastern Australia. Geophysical Research Letters, 14: Torgersen, T., Luly, J., De Deckker, P., Jones, M.R., Searle, D.E., Chivas, A.R., & Ullman, W.J., Late Quaternary environments of the Carpentaria Basin, Australia. Palaeogeography, Palaeoclimatology, Palaeoecology, 67: Torgersen, T., Kennedy, B.M., Hiyagon, H., Chiou, K.Y., Reynolds, J.H., and Clarke, W.B., Argon accumulation and the crustal degassing flux of 40Ar in the Great Artesian Basin, Australia. Earth and Planetary Science Letters, 92: Torgersen, T., Habermehl, M.A., Phillips, F.M., Elmore, D., Kubik, P., Jones, B.G., Hemmick, T., & Gove, H.E., Chlorine-36 dating of very old groundwater 3. Further studies in the Great Artesian Basin, Australia. Water Resources Research, 27 (12): Torgersen, T., Habermehl, M.A., & Clarke, W.B., Crustal helium fluxes and heat flow in the Great Artesian Basin, Australia. Chemical Geology (Isotope Geoscience Section), 102: Toupin, D., Eadington, P.J., Person, M., Morin, P., Wieck, J., & Warner, D., Petroleum hydrogeology of the Cooper and Eromanga Basins, Australia: some insights from mathematical modeling and fluid inclusion data. AAPG Bulletin, 81 (4): Welsh, W.D Great Artesian Basin transient groundwater model. Australia, Bureau of Rural Sciences. ISBN Available from: Welsh, W.D. and J. Doherty Great Artesian Basin groundwater mdelling. Engineers Australia, 29 th Hydrology and Water Resources Symposium February, 2005, Canberra ACT. Veevers, J.J. (ed.), Phanerozoic Earth History of Australia. Oxford University Press, New York, 418 p. Zhao, X., T.L. Fritzel, H. Quinodoz, C. Bethke, and T. Torgersen Controls on the distribution and isotopic composition of helium in deep groundwater flows. Geology 26:

12 Fig. 1: The Great Artesian Basin of Australia from Radke et al (2000). The basin is composed of several sub-basins that are both hydrologically and structurally identifiable. The recharge areas, the reconstructed 1880 potentiometric surface and the inferred flow directions are also shown.

13 Fig. 2: Regional drawdown (meters) of the potentiometric surface of the Cadna-owie- Hooray aquifer following development during the period (from Habermehl, 1980)

14 Fig. 3: Recharge, natural spring discharge areas and directions of regional groundwater flow in the primary Cadna-owie-Hooray aquifer of the Great Artesian Basin based on 1980 data sources (from Habermehl, 1980). These inferred flowlines were used to guide the sampling locations for the 1982 and 1985 field work. Compare to Fig. 1.

15 (a) (c) (b) (d) Fig.4: Chemical profiles along multiple flow lines within the Central Eromanga Basin of the eastern GAB. (a) Chloride, (b) sodium, (c) bicarbonate, (d) initial bicarbonate based on a 13 C balance. The solid lines indicate 8-point moving averages of all wells except those in parentheses. The insert in part (a) shows the variation of rainfall for the region over the last 120kyrs. These data (from Airey et al., 1979) provided guidance for 1982 field work.

16 Fig. 5: Sampling locations and bores numbers for both the 1982 and the 1985 field work. Discussion of flow line evolution is typically discussed in terms of first and last sample (from Torgersen et al., 1992).

17 Fig. 6: Contours of deuterium isotopes (top) show a tongue of isotopically heavy water following the trends in flow directions seen in Fig. 3. Contours of 13 C isotopes (bottom) reflect the trend in flow directions and the dominate geochemical dissolution of marine carbonates (from Airey et al., 1983)

18 Fig. 7: 36 Cl/Cl vs. 36 Cl concentration for the Great Artesian Basin (top) and the comparison of apparent 36 Cl tracer ages to apparent groundwater ages calculated from hydraulic properties (from Bentley et al., 1986)

19 Fig. 8: Helium concentrations vs. groundwater hydraulic age (as supported by 36 Cl dates) for the Great Artesian Basin, Australia as deduced from the 1982 field work. For groundwaters younger than 40kyrs, the accumulation rate is 4.6x10-12 ccstp 4 He cm -3 H2O yr -1 which agrees with the calculated rate (Λ He =1) of 3.95 x10-12 ccstp 4 He cm -3 H2O yr -1. For groundwaters greater than 100kyrs, the rate of accumulation is 2.91x10-10 ccstp 4 He cm -3 H2O yr -1 or 74x in situ production. An external bottom boundary flux provides the best explanation for the source and the timescale over which it can be seen (from Torgersen and Clarke, 1985).

20 Fig. 9: Mantle 3 He is identified in portions of the two flow lines of the GAB, (Fairlight Trust to Wilfred Downs and Juanbong to Bonna Vista; left). The occurence of Cenozoic volcanism (top left) and the groundwater flow lines (top right) show how this signal is carried into the interior of the GAB (from Torgersen et al., 1987)

21 Sr/ 86 Sr S recharge S flowline N recharge N flowline linear distance, km Fig.10: The evolution of dissolved 87 Sr/ 86 Sr within two flowlines of the GAB (north is Stannum to Innamincka and south is BonaVista to Thargomindah). The low 87 Sr/ 86 Sr ratio is acquired in the recharge zone likely via contact with Cenozoic volcanics (see Fig. 9 and note also the acquired 3 He signature). As the flowing groundwaters dissolve the host rock (0.7111) and marine carbonates ( ) in situ, that endmember is steadily acquired in the dissolved phase (after the measurements of Collerson et al. 1988).

22 Fig. 11: 36 Cl measures along Fairlight Trust-to-Clayton flow line show a distinct discontinuity at Meerabooka possibly indicating flow line mixing as deduced by both the 36 Cl/Cl vs. 36 Cl plot (Bottom) and the apparent 36 Cl age vs the apparent hydraulic groundwater age(top) (from Torgersen et al 1991)

23 Fig. 12: Contours of 4 He in the Great Artesian Basin compiled from the 1982 and 1985 fieldwork. High 4 He concentrations clearly delineate the regions of oldest water within the southern Central Eromanga Basin (from Torgersen et al., 1992) and concur with the 36 Cl contours

24 Na+, mmoles kg Athol-to-Mutti Mutti Fairlight-to-Clayton MtCrispe-to-Curdimurka TDS, ppm 1000 Athol-to-Mutti Mutti Fairlight-to-Clayton 10 MtCrispe-to-Curdimurka linear distance, km linear distance, km Athol-to-Mutti Mutti Fairlight-to-Clayton MtCrispe-to-Curdimurka K+, mmoles kg Athol-to-Mutti Mutti Fairlight-to-Clayton MtCrispe-to-Curdimurka [Na-Cl], mmoles kg linear distance, km linear distance, km Fig. 13: Evolution of Na and K chemistry and TDS along the 1985 sampling lines as reported by Herczeg et al (1991). Note that the Western GAB (MtCrispe-to-Curdimurka) is distinct from the Eastern GAB and that mixing with the Eastern GAB flow is apparent at the end of the western flow line.

25 Athol-to-Mutti Mutti Fairlight-to-Clayton MtCrispe-to-Curdimurka 5.00 Athol-to-Mutti Mutti Fairlight-to-Clayton MtCrispe-to-Curdimurka Alk, meq kg Ca2+, mmoles kg linear distance, km 0.0 Athol-to-Mutti Mutti -2.0 Fairlight-to-Clayton MtCrispe-to-Curdimurka linear distance, km Athol-to-Mutti Mutti Fairlight-to-Clayton MtCrispe-to-Curdimurka 1000 delta 13C, per mil Mg2+, mmoles kg linear distance, km linear distance, km Fig. 14: Evolution of alkalinity, 13C, Ca+2 and Mg+2 chemistry along the 1985 sampling lines as reported by Herczeg et al (1991). note that the Western GAB (MtCrispe-to-Curdimurka) is distinct from the Eastern GAB and that mixing with the Eastern GAB flow is apparent at the end of the western flow line.

26 Fig 15: Contours of 14 C (in per cent modern) from Radke et al. (2000). The primary recharge areas are clearly delineated and the loss of 14 C in the older regions of the GAB is clear.

27 Fig 16: Contours of 36 Cl from Radke et al. (2000). The primary recharge areas are clearly delineated as is the locale for very old groundwater. The agreement with the age inferred from 4He contours (Fig. 12) is clear.

28 Fig. 17: Benchmark synthesis of flow rates and chemical evolution of the Great Artesian Basin (from Radke et al. 2000)

29 Fig. 18: Benchmark synthesis of the Great Artesian Basin showing flow rates, upward leakage and chemical evolution (increasing salinity; blue to orange) displayed over the basal structure of the Cadna-owie-Hoorray aquifer (from Radke et al. 2000).