Australia: The Land Where Time Began

A biography of the Australian continent 

Great Artesian Basin of Australia – connectivity with the Underlying Basins, and the Cainozoic Cover

For the understanding and the classification of groundwater systems a defined hydrological basin is convenient. Groundwater generally results, however, from the pressure and density gradients and may pass through boundaries that are defined by hydrostratigraphy. For the Great Artesian Basin (GAB), earlier conceptual models ascribed all artesian water in the region to the GAB. It formed an unwieldy and unpredictable boundary definition for the GAB, though it may have facilitated regulation and groundwater usage. As the appreciation of interconnectedness of this Jurassic-Cretaceous system with underlying sedimentary basins and overlying cover steadily grows, so does the understanding of hydrodynamics in the GAB. The definition of the GAB as a hydrogeological basin was modified in 2012 to align with the Eromanga, Surat and Carpentaria geological basins to simplify this issue for the GAB Water Resource Assessment (GABWRA). Since then this revised definition has been adopted broadly and it has remained in use; this definition has, however, necessitated an understanding that is more consistent of connectivity with deeper basins which have been of increasing interest and importance for the exploitation of petroleum and gas reserves. It is common for the extraction to produce groundwater as a by-product, modifying pressure in the host reservoir and potentially in adjoining aquifers, as well as causing possible contamination through fluid transfer. There are 27 sedimentary basins that are currently known to underlie the GAB and this number is growing steadily through regional seismic studies and exploration.

According to Radke & Ransley the objective of this discussion paper is to provide an overview of regional interconnectivity between the Great Artesian Basin (GAB) and its basement and overlying cover, and to document the regional hydraulic relationships between the GAB and the basins that underlie it. Understanding of the geological basement of the GAB has been increasing steadily since the 1950s, mainly as a result of petroleum exploration (Gravestock et al., 1986; Audibert, 1976; Habermehl, 1980; O’Neil, 1989; Habermehl & Lau, 1997; Radke et al., 1, 2000; Ransley & Smerden, 2012; Ransley, 2015).

The regional approach to intrabasinal connectivity that was proposed first in the GAB Water Resource Assessment (GABWRA; Ransley et al., 2012), and subsequently in the Hydrogeological Atlas of the Great Artesian Basin (Ransley et al., 2015), was to a large extent conceptual as there was very limited knowledge of the hydrogeology of many of the underlying basins. Assessment was based on indirect or variable criteria that were used to describe a broad hydraulic characterisation to individual formations, in these early summations. As it is proving to be difficult to parameterise the basins under standard criteria, current basin-wide or regional connectivity mapping has remained qualitative. Knowledge that is available of the many basins is extremely varied, as it depends largely on the intensity of exploration to date, if any at all. Regardless of these limitations, however, the approach has remained useful as it highlights areas that require further research in order to clarify and quantify hydrogeological understanding for any specific management issue.

Importance of conceptualising connectivity at the base of the GAB

According the Radke & Ransley, this discussion of connectivity is an extension of the concept of regional hydraulic continuity, as was proposed by Tóth (1995), in that the continuity and potential of groundwater migration between sedimentary basins is acknowledged, but attempts to quantitatively rank this connection spatially. It has not been possible to extend Tóth’s concept of hydraulic continuity to all basement – tectonised, metamorphosed and crystalline – through the probability of fracture pathways in these rocks because of the very limited knowledge of geological basement character other than of the immediate underlying sedimentary basins. Cross-formational flow occurs where there is a vertical pressure gradient between 2 formations and there is not enough permeability to permit flow between them. Therefore, the transfer rate between units is a function of the connectivity in terms of the permeability between the 2 units and the magnitude of the pressure gradient between them. Where good seals are formed by aquitards the transfer rate is extremely slow, even if there is a strong pressure gradient (Smerden et al., 2012).

Intrabasinal connectivity can be through the contact of permeable units at the boundaries of basin and along segments of faults that are permeable. Considerable attention has been paid to the latter (Smerden et al., 2012; Moya et al., 2014), while the sheer scale of hydrostratigraphic connectivity is a challenge to the means of assessing it. Assessment was based on indirect or variable criteria to ascribe hydraulic characteristics to hydrostratigraphic units, with the first iteration of connectivity mapping in GABWRA (Ransley et al., 2012). For the regional mapping of relative permeability 5 categories were used:

·       Aquifer,

·       Partial aquifer,

·       Leaky aquitard,

·       Aquitard,

·       Aquiclude.

The term ‘partial aquifer’ is a hydrostratigraphic category used in this paper, following the usage of Ransley & Smerden et al. (2012) and (Ransley et al., 2015) to denote formations that are heterogeneous in character – predominantly interfluvial aquitards, but which include channel sands as aquifers. This use of ‘partial aquifer’ has been necessitated by the broadscale state of current mapping across the GAB. The few regions of extreme minimal connectivity are highlighted as “aquiclude”, in correspondence with Ransley & Smerden et al. (2012) and Ransley et al., (2015), though it is argued that there is hydraulic connectivity across almost the entire GAB.

A very thick marine sequence, that is largely an aquitard, separates the upper and lower continental sequences of the GAB. Fluviolacustrine basins are included among many of the underlying basins. Numerical modelling was subsequently completed for the Arckaringa Basin below the Western Eromanga Basin region of the GAB (Purczel, 2015) and this was followed by tracer studies (Priestly et al., 2017). Basement connectivity in the eastern Eromanga, Surat and Carpentaria basins  were addressed by other studies (Klohn, Crippen & Berger, 2016a, b, c).

The same categories and presentation strategies as was used by Ransley et al. (2012) is used by the current regional connectivity mapping, though it has been revised to include basins that have been newly recognised as well as any hydrostratigraphic information. Across the basement of the GAB connectivity potential is portrayed by the use of this categorisation of units that are in direct contact across the boundary. This approach does not detract from the requirement of conceptual mapping to highlight areas of higher vulnerability where contamination or water loss issues may arise, impacting on GAB underground resources, though this approach has remained qualitative Many of the underlying basins were targets for coal and petroleum exploration, some with commercial extraction and growing extraction of hydrocarbons, in past decades. In the past co-production of groundwater with the resource extraction (petroleum, CSG, dewatering of coal and base metal mines) has already produced many observable drawdowns in pressure in the GAB (Great Artesian Basin Coordinating Committee, 2016). Detailed data acquisition is required to quantify the connectivity in such areas. Disequilibria effects that may develop over short periods of time are not identified or predicted by the current approach; however, pressure reduction, groundwater contamination from extraction processes and hydraulic fracturing. A prime concern of governments and local communities is the resource extraction in the future is achieved within acceptable impacts on existing users and the environment.

Connectivity with underlying basins and basement

Connectivity with underlying basins is offered by the overlapping of aquifers and leaky aquitards juxtaposed below and above the basal unconformity of the GAB. Approximations have previously been offered (Ransley et al., 2012).

Central Eromanga region

The Eromanga sequence thickens over several smaller, mainly nonmarine basins from the Permian-Triassic including the Bowen, Cooper and Galilee basins. There are extensive underlying crystalline basement provinces, and sedimentary basins range in age from Precambrian to Carboniferous. Included among these are the Drummond Basin, Millungera Basin, Arrowie Basin, Barka Basin, Warburton Basin, Warrabin Trough, Adavale Basin, Georgina Basin, and Darling Basin. Hydraulic      connection exists in the central Eromanga Basin with aquifers in several underlying basins and forms a patchwork over about 50% of the central Eromanga region. There are 4 areas where hydraulic connectivity is evident, described as follows.

·       The Galilee and exposed edges of underlying Adavale Basin adjoin with a predominance of leaky aquitards to the east of the Canaway Fault and approaching the recharge zone. The reason for the inclusion of these Triassic sequences in the GAB were the Warang and Clematis Sandstones in the Galilee and Bowen basins, and their pressure potential for the input of groundwater into the GAB.

·       The faulted western margin of the central Galilee Basin shown against the Hutton Rand Structure where the Joe Joe Group partial aquifer and Clematis Group aquifer are structurally upturned and covered directly by the Hutton Sandstone aquifer is an example of connectivity. Here Upper Permian coal measures lie directly over the Joe Joe Group partial aquifer and therefore have potential connectivity with the GBA through this partial aquifer as well as upfault.

·       The Georgina Basin on the Boulia Shelf, northwestern margin, has karstic carbonate and terrigenous aquifers in small areas underlying the Longsight Sandstone. In the Late Cretaceous, during the inversion of the central Eromanga Basin, it is probable that connate groundwater was expelled from Eromanga sequences into this basin (Toupin et al., 1987), In the Cainozoic the interaction reversed, however, as indicated by elevations of the water table (Kellett et al., 2012).

·       On the southern margin of the Central Eromanga Basin in the Lake Frome Embayment, the uppermost Grindstone Sandstone in the remnants of the Arrowie Basin offer connection at the basal unconformity.

·       The Cooper Basin, and the adjoining Warrabin Trough below the Central Eromanga Depocentre, are blanketed to a large extent by less permeable Poolowanna and Evergreen Formations which reduce connectivity; but along the southeastern margin where the overlap of the Poolowanna Formation is not complete, partial aquifers within the Cooper sequence underlie aquifers of the Eromanga Basin.

·       Oil fields in the Cooper area of the Central Eromanga sequence result largely from hydrocarbon that migrated up from the Cooper Basin, though a smaller proportion was generated within the Eromanga. Vertical migrating of hydrocarbon signals groundwater migration as well, though the evidence of whether the migration continues at the present is equivocal.

Western Eromanga region

 In the western Eromanga region, it was interpreted that there is hydraulic connection          with aquifers in several underlying basins in 2 distinct areas, which can cover an area of 50% of the Eromanga region. This coverage is similar to that of the neighbouring Central Eromanga region, though it contrasts with the Surat and Clarence-Norton basins where there is less than 10% of the area of the basement. The 2 areas where connectivity is in evidence include:

·       North of the Denison-Willouran Divide, the Simpson and Pedirka basins. These were interpreted to be adjoined to the GAB by predominantly partial aquifers. Included among these partial aquifers are the fluvial and paludial sand, silt and coal sequences of the Purni Formation within the Pedirka Basin and the fine sandstones of the Peera Peera Formation within the Simpson Basin. It is of note that hydrochemical evidence from petroleum bores and springs in the Dalhousie Springs region suggests that groundwater from the Purni Formation within the Pedirka Basin is contributing to the spring water supply at Dalhousie Springs (Wolaver et al., 2013).

·       The Arckaringa Basin to the south of the Denison-Willouran Divide and in the vicinity of the southwest margin. These offer potential connection at the basal unconformity. Over much of its northeastern portion terrigenous aquifer units occur that are related to the sands and coal sequences in the Mount Toondina Formation. Over the southwestern portion of the Arckaringa Basin, interconnectivity with GAB sequences is by partial aquifers that are associated with the minor sandstone with the Stuart Range Formation, as well as the coarser-grained glaciogenic and marine clastics within the Boorthanna Formation. Though it had been suggested (Belperio, 2005) that the thicker units of siltstone and shale within the Stuart Range Formation may potentially act as a leaky aquitard between GAB aquifers above and the underlying Boorthanna Formation where the Mount Toondina Formation is not present, hydrochemical data were used to argue for a separation between GAB groundwater and water that had been extracted from the Boorthanna Formation aquifer (Howe et al., 2008). It was demonstrated by numerical modelling of the Arckaringa Basin that the outputs were most sensitive to the connectivity between these basement aquifers and the GAB (Purczel, 2015). It was estimated based on the results of environmental tracer studies that interaquifer leakage between basins, which diffuse across a thick aquitard, <1 mm/year, but where the aquitard is not present has yet to be reliably established (Priestly et al., 2017).

Surat and Clarence-Moreton region

Hydraulic connectivity with basement in this region is much more limited, only approaching 10% of the region, which contrasts with the central Eromanga Basin where the area of connectivity approaches 50%. There is, however, notable aquifer interconnection across the basal unconformity of the Surat Basin:

·       On the mid flanks of the Mimosa Syncline at its northern end where some leaky aquitards and partial aquifers are exposed by a synclinal Bowen sequence in the underlying Taroom Trough (Korsch & Totterdell, 2009a, b)

·       Below the St George Bollon Slope, where small, isolated flat-lying remnants of Reids Dome Beds of the Bowen Basin lie at the basal unconformity

·       In the Mulgildie Outlier where aquifers of the Bowen Basin sequence make direct contact with the basal Precipice Sandstone of the GAB

·       Less definitively, below the Laidley Sub Basin of the Clarence-Moreton Basin, where Triassic coal measures below the base are seen as leaky aquitards

·       Below the eastern side of the Coonamble Embayment, where patches of some upper aquifers of the Gunnedah Basin sequence offer connection

Carpentaria region

Not a lot is known about the connection between the Carpentaria region of the GAB and the underlying basins, though it appears there is an onshore areal extent of 5%, reducing to 3% for the combined area (offshore and onshore), that is over the Burketown and Canobe Depressions. Underlying Basins along the eastern flank of this region the Pascoe River, Oliver River and Gamboola Basins (Bain & Draper, 1997) have a small areal extent. The McArthur Basin from the Proterozoic offshore, however, covers almost 50% of basement. To the south and west of the Burketown Depression, in the basement, a subgroup of the Isa Superbasin sequence includes aquifers in the Loretta and Termite Supersequences (Bradshaw & Scott, 1999), and over a subgroup of the South Nicholson Basin there is a widespread region of leaky aquitard. The Lady Loretta Formation, of Palaeoproterozoic age, within the Loretta Supersequence has cavernous and fractured dolostones that have an extremely high transmissivity of 4,542 m²/day (Perryman, 1964). The Cadna-owie-Hooray aquifer of the GAB overlies this excellent aquifer. Nearby, an uppermost thin sequence has been interpreted to be the  leaky aquitards of the South Nicholson Group, of only a fair porosity (5-10%), and slight permeability is included in basement rocks from the Proterozoic (Dunster et al., 1989).

Connectivity of the GAB with overlying cover

Across the onshore GAB there is a considerable extent of Cainozoic cover. Cover from the Palaeogene-Neogene is extensive across the GAB, predominantly in the broad depocentres of the Karumba Basin (Doutch, 1976) and Lake Eyre Basin (Callen et al., 1995) as well as in smaller basins and former drainage systems (Langford et al., 1995). In part, this cover is overlain by drainage-controlled Quaternary alluvium that follows the main drainage tracts, adjoining colluvium flanks to areas of outcrop across the basin.

There is an uppermost weathered zone in underlying cretaceous rocks, and, to a lesser extent, in regions of thinner accumulation dating to the Palaeogene-Neogene (Ransley, 2015; their Fig. 18). There were several intrusive weathering events during the Cainozoic that formed the resistant duricrust surfaces of varying composition over outcrop, as well as bleached underling zones that are modified extremely. The extremes in induration of the duricrust and in reduced permeability in the bleached zone of these weathering profiles have local modifying effects on the connectivity between the surface and the underlying aquifers.

The subsidence that occurred during the Palaeogene-Neogene in conjunction with tectonism that was relatively dramatic that was seen in the continuing uplift of the eastern Dividing Ranges are reflected in several depocentres of accumulation during the Palaeogene-Neogene. The offshore Karumba Basin and the Lake Eyre Basin are the major depocentres. The latter is divided structurally by the Birdsville track Ridge into the Callabonna and Tiari sub-basins, and within several relict drainage systems to the northeast, has related broad sublinear deposits.

Incision has formed palaeochannels closer to regions that are uplifted around the margin of the GAB. Those in the southwestern region commonly exhibit inverted reverse topography and are well documented for uranium prospectivity. In the Surat Basin to the east leads that are very deep are important for irrigation exploitation. The existence of such features elsewhere, though problematic, has yet to be established.

The Karumba Basin in the Carpentaria region has almost total coverage of the GAB sequence with the exception of around the eastern recharge zone. Aquifers in the Bulimba Formation at the base of the Karumba sequence can have direct connectivity with both of the main aquifers – the Gilbert River Formation-Eulo Queen Group equivalents and the Normanton Formation, as a result of the steeper dip of the underlying Carpentaria Basin sequence. It is expected that connectivity with the Normanton Formation exists close to the eastern coast of the Gulf of Carpentaria, while connection with the Gilbert River Formation is upland, closer to the intake beds.

Hydraulic connection is between the Wyaaba beds of the Karumba Basin and the Normanton Formation, in the southern onshore region of the Carpentaria Basin where the Bulimba Formation and equivalents are largely absent. The depocentres of the Lake Eyre Basin indicate, collectively, a downwarp zone aligned northwest to southeast, though disrupted by the Gason and Cooryanna Domes on the Birdsville track Ridge in the central to western regions of the Eromanga Basin. In the eastern part of the Poolowanna Trough adjoining the Gason Dome within the Tiari Sub-Basin as well as in the western part of the Callabonna Sub-Basin the sequence is heavily structural. The structural style is comparable to the polygonal faulting of the Rolling Down Group which underlying. The basal Eyre Formation is a sandy aquifer that apparently has connectivity with the underlying Winton Formation as aquifer on aquifer as well as with probable deeper interconnection through polygonal fault systems. Surface drainage crossing exposed regions can also recharge this formation. Highly variable salinity in the Eyre Formation is indicated by anecdotal evidence from exploitation of the aquifer for make-up water for petroleum exploration drilling.

The Coonamble Embayment over the Surat Basin proper, buried alluvial deposits of palaeovalleys, within the Cainozoic system, overlie predominantly aquitards in the GAB sequence such as Griman Creek and Wallumbilla Formations. These aquitards are variable in character; however, the deep leads can be incised into aquifer units such as the Pillaga Sandstone, over shorter tracts, closer to the margins of the GAB. Such connectivity is not well understood though it is considered to vary significantly across the region.

Concluding comments

According to Radke & Ransley, this is a brief review of the issues around a mappable and workable definition of the GAB as a geological and hydrological basin. This study also addressed through the concept of interbasinal connectivity of deeper groundwater systems that may exist in older basins, and shallow groundwater systems in Cainozoic basins.

Groundwater responds to gradients in pressure and density. There has been a growing realisation of the importance of structure, either passive, as in the case of polygonal fault systems (Ransley et al., 2012, 2015), or with major fault systems (Habermehl, 1980; Moya et al., 2014, 2016) that may provide shorter vertical pathways for the flow of fluid, though hydrostratigraphic models have in the past been the predominant paradigm for the GAB. The earlier concept of trans-basin through-flow that was in the past widely accepted, has steadily been tempered by a growing understanding of a more complex reality that needs to be fully understood in order to enable balanced and realistic risk assessment of future resource exploitation in basins that underlie the GAB.        This study provides a spatial template to help prioritise areas for research in the future, though it has not addressed the relative stagnation in deeper parts of the GAB (Radke et al., 2000), or the various structural mechanisms for upwards leakage throughout the system (Ransley & Smerden, 2012). An indication of the importance of this concept, that has recently been developed, to the management of groundwater that there has been recent federal funding for Geoscience Australia to further the research in connectivity at the base of, and within the GAB sequence.

A patchwork of older sedimentary basins, from Paleoproterozoic to Triassic age within metamorphic and crystalline basement forms the basement to the extensive GAB. Sedimentary basement approaches 50%, but is less than 10% in the Surat and western Clarence-Moreton regions and as little as 5% across the Carpentaria region. Underlying basement coverage offshore approaches 50%, but this is widely believed to have a low connectivity value.

There is variable Palaeogene-Neogene coverage across the GAB sequence. This cover is minimal in the eastern Eromanga and Surat-Clarence-Moreton regions, while the western Eromanga region to the east of the Denison-Willouran Divide is almost completely covered by the Lake Eyre Basin. Similarly, the Karumba Basin offers complete cover offshore and almost onshore, with the exception of the easternmost margin, in the Carpentaria region. Of the many small underlying basins, not much is known, not to mention larger basins that have only partially been explored, and there is much further investigation  that is  necessary; however, realistically the required gathering of data will only eventuate if resource exploration occurs across these areas, following which ensuing environmental impact assessments would be obligatory. Across the western Eromanga region there are Palaeozoic-Triassic basins that have high resource potential though they are of lower development potential at present because of their remoteness and the absence of convenient infrastructure.

Radke, B. and T. Ransley (2020). "Connectivity between Australia’s Great Artesian Basin, underlying basins, and the Cenozoic cover." Hydrogeology journal 28(1): 43-56.