U–Pb zircon age of the Walloon Coal Measures in the Surat Basin, southeast Queensland: implications for paleogeography and basin subsidence

The Jurassic Walloon Coal Measures of the Surat Basin were previously estimated to be of Middle Jurassic age, ranging from Aalenian to Callovian, based on an uncalibrated eastern Australian biostratigraphic framework. New U–Pb dates of 162.55 ± 0.05 Ma and 158.86 ± 0.04 Ma obtained from zircons in ash-fall volcanic tuffs now place the Walloon Coal Measures of the Surat Basin in the Upper Jurassic Oxfordian. The new dates have several implications for the interpretation of the Jurassic strata in the Surat Basin. First-order subsidence rates of 61 m/Myr for the Walloon Coal Measures are more akin to those of foreland basins than the previously assumed intracratonic setting. The dates also imply deposition of the Walloon coals in substantially higher latitudes than previously assumed and that they accumulated as peats in mires that experienced more than three months’ continual darkness each winter. Zircon dating of tuffs and associated geochemistry should assist with the correlation of the laterally impersistent coals, fluvial sandstone and mudstone of the Walloon Coal Measures, which are currently difficult to correlate over distances of more than a few kilometres. Dating of the palynostratigraphic zones APJ4.2 to APJ5 (Aequitriradites norrisii Association Zone to Murospora florida Association Zone) will also need to be recalibrated.


INTRODUCTION
The Walloon Coal Measures (also formally known as the Walloon Subgroup) of the Surat Basin in southeast Queensland and northeast New South Wales contain the majority of Australia's coal seam gas (CSG) resources, accounting for 64% of the continent's proven and probable (2P) estimates (Hamilton et al. 2014a). Several studies have provided insight into the geological evolution and characteristics of the Surat Basin in light of recent CSG exploration including studies of the stratigraphy and sedimentology (Ryan et al. 2012;Martin et al. 2013;Hamilton et al. 2014b), coal petrology (Scott et al. 2007), palynology (McKellar 1998), and regional stress regimes (Baker & Skerman 2006). Despite these studies, there remains considerable uncertainty on the basic stratigraphic framework for the Walloon Coal Measures owing to the geological heterogeneity of the strata and the thin, discontinuous nature of the coal seams and sandstones. This is evident from several attempts to define and redefine the lithostratigraphic units of the Walloon Coal Measures (Cameron 1970;Scott et al. 2004;Scott 2008;Hamilton et al. 2014b). Consequently, there are numerous inconsistencies in the understanding of the formation and the basin in the literature (Martin et al. 2013;Hamilton et al. 2014b).
The primary method of age determination for the Walloon Coal Measures has been using biostratigraphy within an Australian framework constructed on spor-eÀpollen zones that have not been previously calibrated against the international geological time-scale by isotopic dating of associated tuffs (Price 1997;McKellar 1998;Jell 2013). The age ascribed to the Walloon Coal Measures, both in the Surat and in the neighbouring Clarence-Moreton Basin to the east, varies from Aalenian to Callovian (Gleadow 1990;Burger 1994;Jell 2013), with the consensus being late Bathonian to Callovian from palynological studies (Exon & Burger 1981;Wells & O'Brien 1994;Price 1997;McKellar 1998;Jell 2013). This is a wide range in absolute time from 174.1 Ma to 163.5 Ma (Cohen et al. 2013, updated). The use of biostratigraphy has been limited because of the wide age range of spor-eÀpollen taxa, with zonal index species being rare or even absent (Martin et al. 2013). This paper presents new age dates acquired by UÀPb chemical abrasion isotope dilution thermal ionisation mass spectrometry (CA-TIMS) of zircons obtained from tuff layers within coals of the Walloon Coal Measures. These ash fall tuffs have been chosen for dating, as, once deposited on a mire, such tuffs remain in situ through to diagenesis (Triplehorn & Bohor 1986). The dating of the numerous, widely distributed tuff beds, offers promise of a new understanding of the stratigraphy and event history of the basin.

GEOLOGICAL SETTING
The Surat Basin is widely regarded as an intracratonic basin with marine and non-marine strata (Green et al. 1997a;Martin et al. 2013) covering over 300 000 km 2 of southeast Queensland and northeast NSW (Goscombe & Coxhead 1995). It overlies the Permo-Triassic Bowen and Gunnedah basins and forms part of the Great Australian Superbasin (Goscombe & Coxhead 1995;Jell 2013 1995;Holcombe et al. 1997;Pinder 2004). The basin passes into the Eromanga Basin to the west across the Nebine Ridge and Cunnamulla Shelf (Goscombe & Coxhead 1995;Jell 2013).
The Surat Basin originated in the Rhaetian (latest Triassic) after a period of folding, uplift and peneplanation across eastern Australia (Jell 2013). Accommodation space in the Surat Basin is interpreted to have been created by two episodes of thermal relaxation and a period of lithospheric flexure through the Jurassic and into the Cretaceous, allowing 1800 m of strata to be deposited (Exon 1976;Goscombe & Coxhead 1995;Green et al. 1997a;Korsch & Totterdell 2009). Controls on the rate of subsidence in the Surat Basin remain enigmatic with multiple interpretations ranging from rifting, to thermal sag after the cessation of volcanic-arc activity, to viscous corner flow in the mantle wedge above a subducting plate (Korsch et al. 1989;McKellar 2004;Hoffmann et al. 2009;Korsch & Totterdell 2009). Six sedimentary cycles, consisting of a series of fining-upward cycles, have been recognised in the Surat Basin (Exon & Burger 1981). Each cycle consists of an upward succession of strata deposited by (a) braided streams, (b) meandering streams and (c) swamps, lakes, deltas and shallow seas (Exon 1976;Exon & Burger 1981;Jell 2013). These cycles were interpreted by Exon & Burger (1981) to be created by changes in base level driven by eustatic sea-level change. The Walloon Coal Measures form the upper part of the second sedimentary cycle of Exon & Burger (1981), with deposition in swamp, lacustrine and fluvial environments (Ryan et al. 2012;Martin et al. 2013;Hamilton et al. 2014b). Volcanism was pervasive during the deposition of the Walloon Coal Measures and overlying formations; whether this was intrabasinal or extrabasinal remains a topic of debate (Yago 1996a;Boult et al. 1998;Turner et al. 2009). Sedimentation continued in a similar fashion until the Early Cretaceous when eustatic sealevel fluctuations allowed for marine incursions into the basin with the deposition of marine and paralic formations (Jell 2013). Basin inversion in the middle to late Cretaceous led to the erosion of »2 km of strata resulting from rebound of the lithosphere after the cessation of subduction off the east coast of Australia (Hoffmann et al. 2009;Waschbusch et al. 2009). Subsequently the Surat Basin has been relatively quiescent, except for some minor post-Mesozoic adjustment of faults (Scott 2008).

STRATIGRAPHY
The Walloon Coal Measures, the formal lithostratigraphic name currently for the stratigraphic unit recognised by Geoscience Australia (2014) and the Geological Survey of Queensland (Jell 2013), was first referenced by Bryan (1944) with the official type section described by Cameron (1970). A basic outline of Jurassic lithostratigraphic units of the Surat Basin is shown in Figure 1. The Walloon Coal Measures are regionally extensive, thin to the west and 350À450 m thick (Goscombe & Coxhead 1995;Jell 2013). The formation forms the lower part of the coal-bearing Injune Creek Group, underlain conformably by the Hutton Sandstone or Eurombah Formation and overlain unconformably by the Springbok Sandstone, also part of the Injune Creek Group (Exon 1966  No attempt has been made to collate the diverse coalbed nomenclature associated with the Walloon Coal Measures, with coal-bed names varying between exploration companies, mines and geographic districts (Scott 2008). Current lithostratigraphic and coal-bed lithostratigraphic frameworks are discredited, as other correlation methods change the interpretation of depositional-basin architecture (Martin et al. 2013).
The Walloon Coal Measures in the Surat Basin were considered to have been Middle Jurassic and no younger than Callovian, using sporeÀpollen biostratigraphy (Price 1997;McKellar 1998;Geoscience Australia 2014), with some estimates of the formation, in the Clarence-Moreton Basin, to the east, being as old as Aalenian (Burger 1994). The previously estimated age of the underlying Hutton Sandstone ranges from Aalenian to Bathonian, while the overlying Springbok Sandstone has been considered to be Late Jurassic/Oxfordian from biostratigraphy (Price 1997;McKellar 1998;Geoscience Australia 2014). However, these ages now need to be revised, based on the radiometric dates obtained herein from the Walloon Coal Measures.

Tuff sampling
Stratheden 4 (Geoscience Australia well code GA ENO 554973, Geological Survey of Queensland well code ARM   including Helby et al. (1987) and Price (1997) will need to be revised, to take into account not only the age assigned here to the Walloon Coal Measures (in the Surat Basin), but also the more constrained time interval for deposition of the unit. Either (1) the oldest section is unlikely to fall into the Aequitriradites norrisii Association Zone of APJ4.2 and Contignisporites glebulentus/cooksoniae Interval Zone of APJ4.3 that covers the Batho-nianÀCallovian transition, or (2) the first appearance datums need to be recalibrated in the Surat Basin (Price 1997;McKellar 1998) The extent of erosion before deposition of the overlying Springbok Sandstone can now be more accurately defined. The hiatus was likely much shorter than previously assumed (McKellar 1998;Hamilton et al. 2014b) and occurred in the late Oxfordian as opposed to across the   late Callovian/early Oxfordian boundary. Until a tuff horizon is dated in the Springbok Sandstone or the base of the Westbourne Formation, the duration of the hiatus will remain uncertain.
The new dates have interesting implications for paleogeography and paleoecology because, according to current plate tectonic reconstructions, they indicate that the coals of the Surat Basin accumulated as peats in paleolatitudes higher than 75 S (Klootwijk 1996;Blakey 2011;Boucot et al. 2013) compared with the 55 SÀ65 S previously assumed based on a Middle Jurassic age (Veevers et al. 1991;Balme et al. 1995). A position that was well within the polar zone (66 S and higher) would have experienced extreme seasonality, with winter darkness lasting at least three months and continuous daylight for three months during the summer. The climate, despite being in high latitudes, is interpreted as temperate (up to 10 C), humid and with high rainfall (>1000 mm/yr) based on palynofloral assemblages and global climate models, suitable for peat accumulation (Balme et al. 1995;McKellar 1998;Martin et al. 2013). It is well documented that similar climatic conditions existed in the Late Cretaceous and from the late PaleoceneÀEarly Eocene allowing for warm, mid-latitude flora to migrate poleward as climatic belts expanded globally, irrespective of limitations of existing at high latitudes (Spicer & Chapman 1990;Wing et al. 2005). The winter darkness would have restricted primary productivity to the spring to autumn months, even if conditions for coal accumulation and its subsequent preservation were otherwise suitable. Models of peat growth rates in the high latitudes range from approximately 0.01 to 0.05 mm/yr and can be exceeded by basin subsidence, causing mires to be inundated by clastics or drowned by flooding with the creation of lakes (McCabe 1991;Bohacs & Suter 1997;Loisel et al. 2012). This may explain the thin nature of the Walloon coal beds. Other high-latitude coals, such as those from the Cretaceous of Antarctica and Alaska, also have thin, discontinuous seams (Spicer & Parrish 1986;Macdonald & Francis 1992).
A first-order estimate of subsidence rates can be defined using the equations and methodologies of Allen & Allen (2013). Using the new UÀPb dates and stratal thicknesses from core, a basin subsidence rate of »61 m/Myr is calculated. This is similar, although slightly higher, to subsidence rates calculated for the late Mesozoic foreland basins of the western United States that on average, range from »51 to »35 m/Myr over the lifetime of a basin (Cross 1986). By comparison, the calculated rate of subsidence of »25 to 5 m/Myr for the cratonic Williston Basin in the northern United States (DeRito et al. 1983) is considerably lower. Despite the complex prehistory of the Surat Basin (rifting and foreland loading) influencing depostional architecture (McKellar 1998), theories of continental, intracratonic sag for the Surat Basin, including subduction-related dynamic platform tilting and intracratonic sag (Green et al. 1997b;Waschbusch et al. 2009;Jell 2013), are now questionable.

SUMMARY
UÀPb zircon dating of ca 159 Ma and 162 Ma tuffs place the Walloon Coal Measures in the Oxfordian, which is substantially younger than previous estimates based on eastern Australian biostratigraphic frameworks. These dates also place the Surat Basin at higher latitudes (>75 S) during the deposition of the Walloon Coal Measures than previously assumed. The coals accumulated as peats in mires that experienced at least three months of continual darkness per year. The new dates allow for precision in measuring basin subsidence rates, which may help determine the tectonic setting of the basin: subsidence rates were more similar to those of foreland rather than intracratonic basins. The dates will also provide the basis for a new, more precise biostratigraphic framework for the Jurassic of Australia. Lithostratigraphic correlation of strata of the Walloon Coal Measures within the Surat Basin has long been difficult owing to the heterolithic, laterally impersistent coals and sandstones. Dating of zircons (and associated geochemistry) from additional volcanic tuffs offers the possibility of more precise correlations across the Surat Basin and further afield that may allow for the development of better non-marine sequence-stratigraphic models.

SUPPLEMENTAL PAPERS
Analytical Methods Figure 1 Cathodoluminescence (CL) images of selected zircons extracted from the two tuff samples from Stratheden 4. Grains dated by CA-TIMS and spot analysed by LA-ICPMS are shown. Table 1 CA-TIMS data from GA2180600 and GA 2180601 in Stratheden 4. Table 2 UÀPb geochronologic analyses and trace element concentrations.
Supplemental data is available alongside the online version of the article, at http://dx.doi.org/10. 1080/ 08120099.2015.1106975