Segmentation and fault–monocline relationships in the Lapstone Structural Complex, Sydney Basin, New South Wales

Abstract The 95 km-long Lapstone Structural Complex is a segmented structural association consisting of north-trending faults and monoclines in the western Sydney Basin between the Cumberland Basin to the east and lower Blue Mountains to the west. It has developed in a compressional regime. At depth, the Lapstone Structural Complex is most likely a deeply penetrating, west-dipping thrust fault that is seismogenic in the brittle middle crust. This structure has propagated upwards into the overlying Sydney Basin in the top ∼3 km of the crust and formed a suite of distinct fault and fault–monocline structures, including: (a) in the north, inferred imbricate faults at depth in the lower Sydney Basin dying out upwards into a well-displayed, major east-facing monocline (central limb has an overall consistent <20°E dip); (b) in the middle part of the complex, an east-facing monocline with dips increasing from west to east formed as a fault-propagation fold; and (c) as found in the south, a single thrust fault (Bargo and Nepean faults). The complex has a probable late Cenozoic age and has played a role in landscape development as shown by topography, uplifted river gravels and knick points along lower order streams. Therefore, it has formed late in the uplift history of the southeastern Australian highlands postdating uplift in the Cretaceous and uplift associated with basaltic eruptions in the Cenozoic. KEY POINTS The Lapstone Structural Complex is an association of segmented faults and monoclines formed at the tip of a deep-seated thrust fault. The largest structure in the complex is a gently east-dipping monocline formed above an inferred imbricate fault system in the lower Sydney Basin succession. Much of the middle and southern parts of the complex are single faults and a fault-propagation monocline related to a thrust fault at depth. The Lapstone Structural Complex is of late Cenozoic age and is most likely seismogenic.


Introduction
It has been established from mapping at map and outcrop scales that faults typically have a segmented character with overlapping and/or terminating adjacent fault segments (e.g. Conneally et al., 2017;Fossen & Rotevatn, 2016;Manighetti et al., 2015;Pizzi et al., 2017), and these patterns have been duplicated by experimental studies (e.g. Mansfield & Cartwright, 2001). This is particularly evident in faults with limited net slip, as faults with greater offsets have had their segments linked enabling straighter and less segmented faults to have formed (Fossen & Rotevatn, 2016). It is clear from past descriptions of the Lapstone Structural Complex (LSC) in the Sydney Basin 55 km west of Sydney ( Figure 1) that it is a segmented 95 km-long fault complex with associated monoclines (Branagan & Pedram, 1990Bray et al., 2010;Fergusson, 2009;Fergusson et al., 2011). We describe the segmented structure of the LSC and relate different segments to the inferred deeper structure in the lower Sydney Basin succession using constraints from drilling, seismic reflection profiles, gravity modelling (Bray et al., 2010;Danis et al., 2011) and structural concepts developed to relate faults and monoclines from field studies, computer and experimental modelling (Allmendinger, 1998;Davis & Bump, 2009;Erslev, 1991;Hardy & McClay, 1999;Withjack et al., 1990). We also consider a number of other issues relevant to the LSC including speculation on the nature of the controlling structure at depth, discussion on the roles of extension/compression in genesis of the LSC, the role of downwarping and/or uplift, and the control exerted by basement structures in formation of the LSC.

Setting
The Permian-Triassic Sydney Basin in the greater Sydney region of central coastal NSW contains several plateaus, including the Hornsby Plateau in the north, the Blue Mountains Plateau in the west and the Woronora Plateau in the southeast, and all are separated by the low-lying Cumberland Basin (Figure 1; Bembrick et al., 1973;Gale, 2021). The uplift history of these dissected plateaus is generally poorly constrained with estimates ranging from the Cretaceous and in one or more uplift episodes in the Cenozoic (Gale, 2021;Hatherly, 2020;Hatherly & Brown, 2022;McPherson et al., 2014). The Hornsby and Woronora plateaus are marked by a gradual rise in topography at their southern and northern margins along the Hornsby and South Coast warps, respectively (Bembrick et al., 1973). The eastern boundary of the Blue Mountains is marked by a group of north-trending monoclines and faults collectively referred to as the LSC (Branagan & Pedram, 1990. The variation from an east-facing monocline to highangle faults was outlined by Branagan and Pedram (1990), who also suggested that the LSC was controlled by a basement feature in the underlying Lachlan Orogen. Seismic reflection profiles presented by Bray et al. (2010) showed that the LSC played no part in Permian to Triassic deposition as had been earlier suggested (Pickett & Bishop, 1992). A Paleogene-Neogene age for formation of the complex (Fergusson et al., 2011) was considered consistent with the age of basalts in the Blue Mountains and a model of landscape development (van der Beek et al., 2001). More recently, Hatherly (2020) recognised that three phases of uplift were involved in formation of the Blue Mountains and that the final phase was related to development of the LSC (see also Hatherly & Brown, 2022). This was suggested to have occurred in the last 10-5 Ma reflecting the contemporary compressive stress field (Braun et al., 2009).
An east-facing monocline is well developed in the northern half of the LSC (Branagan & Pedram, 1990;Fergusson et al., 2011;Jones & Clark, 1991). Monoclines are common structures that have been recognised in either compressional settings, including uplifts associated with the Late Cretaceous-early Cenozoic Laramide Orogeny in the Colorado Plateau (Bump, 2003;Davis & Bump, 2009) or extensional settings associated with normal faulting including the Red Sea region and in many extensional sedimentary basins (Conneally et al., 2017;Gawthorpe et al., 1997;Withjack et al., 1990). Regardless of their stress regime, application of the trishear model explains the variation in structural styles of monoclines associated with faults at depth consistent with their development as faultpropagation folds (Allmendinger, 1998;Erslev, 1991). The LSC has been associated with a compressional regime (Fergusson et al., 2011;Hatherly, 2020) but in other papers no attribution to a particular stress regime has been noted (e.g. Gale, 2021;Pickett & Bishop, 1992). It has also been suggested that the LSC developed from downwarping of the Cumberland Basin rather than uplift of the Blue Mountains (Branagan, 2009;Brown, 2000;Jones & Clark, 1991). Downwarping of the coastal tract was considered to have been associated with rifting that formed the Tasman Sea in the Late Cretaceous and early Cenozoic (Branagan, 2009;Ollier & Pain, 1994). This suggestion contrasts with rapid denudation in the coastal tract as one interpretation of the younging of fission track ages towards the coast that was associated with uplift of the highlands in southeastern Australia at the onset of sea-floor spreading in the Tasman Sea (Gleadow et al., 2002). Brown (2000) argued for downwarping along the coastal tract around Ulladulla in the southern Sydney Basin as post Oligocene.
The Supplemental data include a summary of the stratigraphy, figures (Figures S1-S15) and tables (Tables S1 and S2).

Methods
We have undertaken detailed geological mapping of selected areas along the LSC concentrating on the northern part between Lapstone and Kurrajong Heights and along the southern part at the Bargo River southeast of Tahmoor (Figures 2 and S1) that complements our earlier work (Bray et al., 2010;Fergusson et al., 2011;Hatherly, 2020). Field data collection has involved mapping of unit boundaries using hand-held GPS instruments for determining precise locations as well as plotting traverses. Orientation data have been collected from bedding, but it is important to emphasise that the orientation of bedding is not easily determined in the Hawkesbury Sandstone, the main stratigraphic unit associated with the LSC. This is due to the abundant cross-bedding and primary dips associated with channel margins. Only the largest exposures of Hawkesbury Sandstone can be relied upon to recognise primary bedding features likely related to paleohorizontal to enable mapping variations in dip related to deformation. Not all orientations of bedding published in the past are reliable, and some anomalous steeper dips have given rise to fallacious structures (e.g. Branagan & Pedram, 1990, figure 6;Fergusson et al., 2011, figure 3).
Another source of important information is the LiDAR topographic data with spacing of 2 m or better available across all of the region from the ELVIS web site (https://elevation.fsdf.org.au/). These data allow the construction of detailed digital elevation models, topographic maps and more accurate cross-sections (Hatherly & Brown, 2022). In areas containing dominantly flat-lying beds (<5 ) of the Hawkesbury Sandstone, the layering controls benching reflected in the presence of many continuous cliffs developed along the steep slopes adjacent to entrenched watercourses and allows mapping of the overall orientation of layering. Although we have found that in some areas, the cliff-lines cut across layering, and thus mapping these features needs to be used with caution, especially in areas of known inclined layering such as along the central limb of the Lapstone Monocline. Where available, borehole data have also been used to constrain the subsurface stratigraphy and structure (Bray et al., 2010;Danis et al., 2011). Seismic reflection profiles have also been used, drawing on our earlier work (Bray et al., 2010).

Segmentation of the Lapstone Structural Complex
Analogous to the monoclines of the Colorado Plateau, it is considered that the LSC has formed by the upward propagation of a basement fault (Fergusson, 2009). The nature of the fault propagation from basement into the overlying Sydney Basin in the top 3 km of the crust has resulted in a variety of structures in the LSC reflecting segmentation with a suite of distinct fault and fault-monocline structures. A segment extends from Shaws Creek, just to the north of the Hawkesbury Lookout, northwards through Kurrajong Heights towards the Colo River (Figure 2), and is characterised by an east-facing monocline with a central limb up to 4 km wide and dipping 5-20 E (Figures 3, 4 and 5a). Additionally, the Kurrajong Fault System (Branagan & Pedram, 1990) includes high-angle reverse faults along the western margin of the northern LSC between the Colo River and Nortons Basin.
In contrast, to the south of Shaws Creek, an east-facing monocline has dips increasing from the west up to 40 to the east (Figures 6 and 7; Fergusson et al., 2011, figure 4). South of Lapstone the monocline widens and has a substantial lower dipping central planar limb similar to that north of Shaws Creek (Figure 8a). The southern boundary of this second segment coincides with the Nortons Basin diatreme ( Figure 2).
From Nortons Basin southwards, the LSC is dominated by segments with a single planar fault (e.g. the Nepean and Bargo faults), as well as sections with a simple east-facing monocline (e.g. the segment east of The Oaks) or no structures at all, as found north of Picton ( Figure 2).
From north to south, the following sections describe the segments of the LSC and their interpreted structure in the lower Sydney Basin.

Shaws Creek northwards to the Colo River
From Shaws Creek northwards, the LSC coincides with an increased elevation of the Hawkesbury Sandstone-Wianamatta Group contact, which partly reflects uplift associated with development of the Hornsby Warp and the associated Hornsby Plateau (Figure 2). The Hornsby Plateau reaches elevations up to 200-250 m and from south to north on the eastern side of the LSC rises from 20 m near the Grose River to locally over 200 m in the Wheeny Creek area ( Figure 2). Thus, some of the higher elevation in the northern part of the complex can be attributed to uplift associated with formation of the Hornsby Plateau.
At Wheeny Creek, a spectacular gorge has been cut through the LSC ( Figure S4). The eastern Lapstone Monocline is up to 3 km in width with a central planar limb dipping <5 and the upper limb truncated to the west by the steeply dipping Kurrajong Fault (see later section on Kurrajong Fault System) associated with locally westdipping beds up to 45 ( Figure S4, cross-section). A traverse across the Kurrajong and Wheeny Gap faults at Wheeny Gap has been described in detail by Clark and Rawson (2009), who attributed several tens of metres of offset along the Wheeny Gap Fault (Clark & Rawson, 2009, figure 3), but at the scale of the whole LSC, this amount of offset is relatively minor ( Figure S4).
Immediately north of Shaws Creek, the Lapstone Monocline changes trend from a general north-south direction to a northwest trend for at least 2 km towards the Grose River (Branagan & Pedram, 1990) as shown by an entrenched section of Lynchs Creek ( Figure 3). Additionally, the central planar limb of the monocline widens northwards, and topographic elevation of the upper limb increases to over 600 m at Kurrajong Heights and over 630 m around Mountain Lagoon (i.e. as pointed out by Branagan & Pedram, 1990, over three times that at Lapstone). In contrast to further south, no localised dips >30 are encountered (Figures 3-5, and S1-S4). Overall, the middle limb dips >12 E, as shown in the spectacular cliffs on the southern side of the Grose River ( Figure 5a). A small thrust fault dipping 40 E in shale of the upper Burralow Formation was found in a gully on the southern side of the Grose River (Figure 5d, e; Crook, 1957, section 17). The Lapstone Monocline north of Shaws Creek is wider and more planar in contrast to that between Shaws Creek and Lapstone.

Deep structure
The approach in structural geology has been to draw cross-sections to depth based on the principles of balanced cross-section construction (Dahlstrom, 1969). For monoclines, trishear modelling has also been utilised to portray subsurface structures (Allmendinger, 1998;Erslev, 1991) but in regions without detailed seismic reflection profiles and abundant deep bores, including the Colorado Plateau and the Blue Mountains, the deep structure is largely based on predictions from surface geology (Davis & Bump, 2009;Fergusson, 2009;Fergusson et al., 2011). The deep structure drawn in Figure 4 has been extended to depth using the kink-band method of construction whereby axial planes are drawn bisecting the upper and lower monoclinal hinges of the northern Lapstone Monocline. This allows the structures to be projected to depth but results in an unbalanced cross-section as the converging axial planes result in reducing shortening to depth. In Figure 4, we address this issue by constructing an imbricate set of thrust faults at depth that allow approximately the same amount of shortening in the near surface to be maintained to depth. Surface exposures are too shallow to expose these imbricate thrust faults, although we note that one minor thrust fault has been found in the Burralow Formation immediately below the contact with the overlying Hawkesbury Sandstone in a gully south of the Grose River (Figures 3, 5d and S2).

Fitzgerald, Frasers and Shaws creeks and Hawkesbury Lookout
From Lapstone northwards, there is a gradual rise in the elevation of the upper hinge of the Lapstone Monocline from a height of 200-250 m at Yellow Rock and reaching over 300 m north of Hawkesbury Lookout. The structure is relatively simple and is a continuation of the Lapstone Monocline northwards (Figures 6 and S5-S12). This is well illustrated by excellent exposures of the monocline along the southern valley sides of Fitzgeralds Creek ( Figure S8a, b) and Frasers Creek. The structure is similar to that at Mitchells Pass Road and the Old Bathurst Road (Figures S11 and S12) with moderately dipping beds in the lower part of the central limb that change to a lower dip of 30 to <10 that dominates the upper part of the central limb.
In Shaws Creek, the upper hinge of the monocline is exposed and is a gentle upright anticline with a steeply dipping axial plane ( Figure S7a, b). The fold plunges gently to the south ( Figure S6). The west-dipping limb of this fold is only 2 m in width, and immediately to the west, the beds are flat-lying and form the upper flat limb of the monocline ( Figure S7b). The Hawkesbury Lookout area is notable for the development of two shale lenses in the Hawkesbury Sandstone as shown by Branagan and Pedram (1990) and in our map and cross-sections (Figures S5 and S7c). Locally moderately to steeply dipping beds occur in places along the face of the monocline (e.g. Figure S7d). Some of these measurements in Figure S5 are shown with question marks as in some steep gullies; the possibility of downslope movement and steepening of beds cannot be ruled out for the smaller outcrops. Shale interbeds occur with the sandstone on the upper flat limb in Shaws Creek ( Figure S7e) and possibly mark the contact between the Hawkesbury Sandstone and the underlying Burralow Formation.
In contrast to the map and interpretation presented by Branagan and Pedram (1990), we have recognised no faults associated with the monocline at the surface along this section of the structure ( Figures S5 and S9). Additionally, Herbert (1989)

inferred the existence of the Mount Riverview Fault between the Old Bathurst Road and Shaws
Creek from a seismic line UD1, but we have been unable to find any topographic expression or evidence in outcrop indicative of this structure. Note that the scarp attributed to this fault in Fergusson et al. (2011,

Lapstone-Leonay
In the Lapstone-Leonay area, the Hawkesbury Sandstone and overlying units are well exposed in abundant road cuttings as well as natural exposures. Here the structure of the Lapstone Monocline is well known. Detailed geological maps of the area were published by Branagan and Pedram (1990) and Fergusson et al. (2011). An updated map including mapping by Carter (2011), Hatherly (2020) and the authors is given in Figure S11. The structure is dominated by a simple east-facing monocline ( Figure S12), although the lower limb is not flat-lying but dipping 7-8 E. Detailed mapping of the distribution of the Rickabys Creek Gravel shows that this unit has been folded along the face of the monocline (Hatherly, 2020), supporting observations of Branagan and Pedram (1997), and has not formed as several uplifted alluvial-cut terraces, as has been suggested (Pickett & Bishop, 1992). The recognition of Londonderry Clay on the Old Great Western Highway Track ( Figure  S12d) also indicates that this unit has been uplifted and folded (Carter, 2011;Gale, 2021). Local fault breccia and fractures have been documented by Branagan and Pedram (1997) in the Lapstone area, but no significant fault offset has been demonstrated associated with the Lapstone

Deep structure
In contrast to the subsurface structure north of Shaws Creek, the subsurface structure of the segment between Shaws Creek and Nortons Basin is relatively straightforward. The structure of the Lapstone Monocline, with dips increasing to the east in this segment, indicates a thrust fault at depth dipping at 30-40 W (Figure 7) and with a vertical throw of 100 m increasing northwards to over 200 m. The kink-band construction given in Fergusson (2009) has portrayed a moderately west-dipping thrust at depth that has formed the monocline as a fault-propagation fold in the upper Sydney Basin succession and has been redrawn in Figure 7 using the basement depth determined by Danis et al. (2011).

The Kurrajong Fault System
The Kurrajong Fault System was the name given by Branagan and Pedram (1990) to the high-angle faults developed west of the Lapstone Monocline from south of Glenbrook northwards to beyond Mountain Lagoon. These structures at the surface are mainly developed in the Hawkesbury Sandstone, although in places offsets along them are constrained by different elevations of the Hawkesbury Sandstone-Wianamatta Group contact, which all have marked topographic expression with west-facing escarpments (<100 m relief) and are hard to find in outcrop, despite the relative abundance of outcrop. The Kurrajong Fault System includes the Kurrajong, Burralow, Grose, Blue Gum, Fraser, Yellow Rock and Glenbrook faults (Figures 4, 5f, 6, 7, S9 and S10).
Features associated with these faults, including the prominent flexure associated with the Kurrajong Fault at 'Cut Rock' (Figure 2), were described by David (1902) and Branagan and Pedram (1990). The Glenbrook Fault is also associated with a prominent flexure in the Hawkesbury Sandstone with dips increasing from 2 to 3 W 150 m east of the fault to local dips adjacent to the fault of 35 W ( Figure S10). Nearly all of the 55 m offset along the Glenbrook Fault indicated by the topographic scarp is explained by this flexure, and no actual fault plane has been identified in Glenbrook Creek where continuous exposure allows identification of the western-most hinge of the flexure. In contrast, despite detailed field examination, no flexure has been found along the Yellow Rock Fault, nor is a fault plane exposed, despite abundant outcrop. Offset along the Yellow Rock Fault is indicated by the prominent topographic escarpment (60 m relief) in addition to offset along the Hawkesbury Sandstone/ Wianamatta Group contact ( Figures S9 and S10). The Grose Fault is also marked by an impressive topographic scarp (100 m relief) and has some associated irregular dips in the Grose River (Figures 3 and S2). West-dipping beds in the head of a gully on the southern side of the Grose River indicate the local presence of a west-facing flexure similar to the Kurrajong and Glenbrook faults, but this has not been found further to the southeast along Shaws Ridge where, similar to the Yellow Rock Fault, the Grose Fault is indicated only by an impressive topographic scarp (80 m relief; Figure S9).
Evidence of neotectonic activity on the Kurrajong Fault System is given by fault-angle depressions on the western Figure 7. Cross-section along northing 6272500 through Hawkesbury Lookout based on kink-band method reconstruction given in Fergusson (2009). Location shown on Figure 6. side of the system. As discussed by Rawson and Clark (2009) these depressions form swamps, the largest at Mountain Lagoon, which abuts the Kurrajong Fault (McPherson et al., 2014), and Burralow Swamp, which abuts the Burralow Fault ). Drilling in these two swamps revealed bedrock bars rising approximately 15 m at Mountain Lagoon and 5 m at Burralow Swamp, which were interpreted to be due to minor faults. Pollen analysis of sediments from drilling within the swamps suggests a late Pleistocene to Holocene age at Burralow Swamp

The Nepean Fault
The Nepean Fault extends from Nortons Basin southwards for 20 km and is well developed at Bents Basin (Branagan & Pedram, 1990;Fergusson et al., 2011) where the fault is steeply dipping and has 100 m of throw. Seismic reflection line CD85-115 (Figure 8b) shows that the fault is a west-dipping reverse fault (Bray et al., 2010, figure 4b), whereas adjacent to the fault resolution in other seismic reflection lines is too poor to indicate its dip.

Spring Creek Monocline east of The Oaks
East of The Oaks, the Nepean Fault is replaced by the Spring Creek Monocline, a relatively narrow feature <100 m in width containing a central limb dipping 12-15 E. It has a throw of 30-40 m from east to west across the structure (Fergusson et al., 2011), considerably less than further north. The full extent of this structure has not been mapped.

Lack of structure north of Picton
Southeast of The Oaks and north of Picton, the southern extension of the LSC is devoid of any structure, whereas small monoclines and faults are developed in and south of Picton (Fergusson et al., 2011). South of Picton, the scarp across the LSC has a relief of only 20-30 m, indicating much less overall throw across the structure than further north.

The Bargo Fault
The Bargo Fault is the southernmost component of the LSC and is marked by a prominent fault scarp, 7.5 km in length (Figure 9). The scarp has a maximum of 40-50 m relief in its middle section around the Bargo River (Figure 10a, b) but in its northernmost part, 2.5 km south of Picton, it has a relief of <20 m. Most of the fault is covered by soil and talus at the base of the scarp, and the fault dip has therefore been difficult to establish. The base of the Wianamatta Group serves as a useful marker with the western block uplifted and indicates a maximum of 60 m of throw across the fault (Figure 9). The only exposure we have found of the fault occurs on the northern side of the entrenched Bargo River and has a dip of 52 W (Figure 10c, d) and thus shows that the Bargo Fault is a reverse fault or a thrust fault with a moderate west dip. The fault surface is planar and marked by ironstone bands indicative of groundwater movement and iron precipitation (cf. Vernon, 2021) along the fault (Figure 10c, d). A gouge zone up to a metre or more thick is developed along the fault plane (Figure 10d) but slickenlines are not preserved owing to the intense weathering, no doubt aided by groundwater movement.
In the Bargo River gorge, the hanging wall has a zone of gently northwest-dipping beds up to 50 m west of the fault (Figures 9 and 10e), whereas in the footwall beds dip more steeply up to 20 to the east-southeast up to 150 m downstream of the fault (Figures 9 and 10f). Overall, a gentle anticline is developed in the vicinity of the fault. Outside this narrow structure west of the fault, beds are flat-lying, whereas east of the fault, the regional dip is 1 E ( Figure S13).

Deep structure
The deep structures of the segments between Nortons Basin and the southern end of the Bargo Fault are all indicative of a moderately west-dipping thrust fault at depth but with only limited throw (<70 m) and therefore much more reduced than further north.

Londonderry Sub-basin and the LSC
The Cenozoic succession in the northwestern Cumberland Basin includes the Rickabys Creek Gravel, the overlying Londonderry Clay and numerous younger upper Cenozoic to Quaternary alluvial units, including abundant gravel with a clast content indicating derivation from the older Rickabys Creek Gravel (Jones & Clark, 1991;Nanson et al., 2008). The deposition of the Rickabys Creek Gravel is indicative of rejuvenated uplift in the vicinity of the Great Dividing Range that has provided the input of either Lachlan Orogen detritus or reworked Lachlan Orogen detritus in Permian conglomerates of the lower Sydney Basin succession.
The terminology of the basinal feature containing these units needs to be reassessed. The term 'Penrith Basin' is not well defined in the literature with its outline mainly shown on maps and has been related to localised thickening of the Wianamatta Group (Bray et al., 2010;Jones & Clark, 1991). The so-called 'Penrith Basin' is centred on Windsor; no part of it includes the Penrith district, and it is therefore somewhat inappropriately named. The Cenozoic succession of the Windsor-Richmond region is more widespread than the extent shown for the so-called 'Penrith Basin' and is part of the larger physiographic Cumberland Basin, and the actual deposits are considered herein to form a depositional feature named the Londonderry Sub-basin.
The Londonderry Sub-basin along its southwestern extent shows evidence of minor northeastward tilting (Gale, 2021; based on data in Carter, 2011). North of Lapstone, eroded remnants of the Rickabys Creek Gravel and Londonderry Clay along the Great Western Highway Track ( Figure S11) have been uplifted along the lower gently dipping limb of the Lapstone Monocline, whereas at the northern end of the Londonderry Sub-basin, the Rickabys Creek Gravel has been uplifted along the Hornsby Warp  7 km north-northeast of Windsor at Wilberforce (Figure 1; Carter, 2011;Gale, 2021;Hall, 1926;Hatherly, 2020). Thus, deposition of the Rickabys Creek Gravel and Londonderry Clay provide constraints on development of both the LSC and Hornsby Warp. Uplift and/or rejuvenation of the Hornsby Warp has ponded the Rickabys Creek Gravel and allowed deposition of the Londonderry Clay in a lacustrine environment (Carter, 2011;Gale, 2021;Hall, 1926) No definitive ages are known for units in the Londonderry Sub-basin apart from them predating the late Quaternary gravels along the Nepean River (Nanson et al., 2008). However, uplift of the southwestern part of the Hornsby Plateau must have postdated the 45 Ma basalt capping the Maroota Sands (Graham et al., 2010), 25 km northeast of the Londonderry Sub-basin.

Discussion
Uplift, downwarping and the role of extension and/or compression in the origin of the LSC Fundamental to the significance of the LSC is its development under either compression or extension, or a complex history involving both. Related to the stress configuration is the development of the land surface over time. Has the present-day topography formed either by uplift of the adjacent Blue Mountains west of the structure or by downwarping of the Cumberland Basin to the east?
Downwarping associated with development of a coastal plain along the eastern Australian coastline has been associated with rifting of the highlands centred on a hypothetical Tasman Divide to form the Tasman Sea (Ollier & Pain, 1994). Thus, downwarping would have occurred in the Late Cretaceous to Paleogene during rifting and sea-floor spreading in the Tasman Sea (Branagan, 2009;Gaina et al., 1998). More recently, it has been suggested by Brown (2000) that downwarping of the coastal tract postdated Oligocene sediments and basalts in the Ulladulla region. Brown (2000, p. 248) also gave the Lapstone Monocline as another example of downwarping in the Mesozoic to Cenozoic. Thus, the lower Blue Mountains would represent a horst adjacent to the downwarped Cumberland Basin, as also suggested by Jones and Clark (1991, p. 89). Downwarping would have occurred in an extensional regime associated with the general development of the east Australian passive margin (Branagan, 2009) formed initially by extension, followed by subsequent thermal cooling and subsidence, as is typical of passive margins generally (Reston & Manatschal, 2011). For an extensional origin, the LSC would have formed in association with a steeply east-dipping normal fault producing monoclines as forced folds (Figure 11a; cf. Withjack et al., 1990).
A compressional origin would have produced a westdipping thrust fault at depth that caused the LSC to have developed in a trishear zone above the tip of the basement thrust (Figure 11b; cf. Erslev, 1991). There are several lines of evidence that indicate that the LSC formed in compression rather than extension. In the southern part of the LSC, the Bargo Fault is clearly a west-dipping reverse/thrust fault and therefore formed in compression (Figures 9 and 10). A seismic reflection profile shows that the Nepean Fault is steeply dipping to the west and a high-angle reverse fault (Figure 8b; Bray et al., 2010) and thus related to compression. Nevertheless, the steep dip of the Nepean Fault allows a more complex interpretation with formation as normal faults followed by reactivation in compression, although it is not clear why west-dipping normal faults would have formed in the early history of the LSC.
The factors we believe support development of the LSC in compression rather than extension include the relationship between the Rickabys Creek Gravel and the Lapstone Monocline not being consistent with an origin of the Cumberland Plain by downwarping and by implication extension. The Rickabys Creek Gravel covers the northwestern part of the Cumberland Plain in a particularly topographically subdued area with little relief in contrast to other parts. It is folded along the east-dipping limb of the Lapstone Monocline (Fergusson et al., 2011;Hatherly, 2020) as was recognised by numerous earlier geologists (e.g. David, 1896). On the Cumberland Plain, the Rickabys Creek Gravel overlies the Wianamatta Group, whereas along the middle and upper limbs of the Lapstone Monocline, it overlies either the Mittagong Formation or the Hawkesbury Sandstone. This relationship is explained by uplift of the western side of the LSC (Hatherly, 2020) in a compressional setting rather than by downwarping to the east.
The structure of the Lapstone Monocline between Lapstone and Shaws Creek, with the central limb increasing in dip eastwards, can be interpreted in different ways. It could be regarded as related to a steep east-dipping Figure 11. (a) Forced fold with a steeply dipping normal fault at depth overlain by a monocline (shaded area). After Withjack et al. (1990, figure 10, AAPG Bulletin, AAPG@1990, reprinted by permission of the American Association of Petroleum Geologists whose permission is required for further use). Note how the monocline increases in width up-section. (b) A monocline developed above a thrust fault at depth with development of the monocline as a fault-propagation fold during trishear. Modified after Erslev (1991, figure 4, A footwall-fixed).
normal fault (Figure 11a; forced fold of Withjack et al., 1990). Alternatively, we regard it as more likely related to formation as a fault-propagation fold above the tip of a west-dipping thrust fault in compression as envisaged for the monoclines of the Colorado Plateau and illustrated by Erslev's (1991) analysis of trishear ( Figure 11b).
All the faults in the Kurrajong Fault System have the eastern side uplifted and the western side down-dropped. If the LSC had formed by downwarping to the east, it would be anticipated that these faults should be east-side dropped, rather than their western sides. Steep east-dipping reverse faults developed west of the main structures of the LSC are more likely. We therefore regard these structures as consistent with compression rather than extension.
The earthquake record in the Sydney region is greatest in the Blue Mountains (Brown & Gibson, 2004) and consistent with a west-dipping fault (i.e. a thrust and hence compression) at depth related to the LSC (Gibson, 2009). Thus, the LSC appears to be a seismogenic fault, and this is consistent with evidence along parts of the structure for neotectonic activity at Mountain Lagoon, Burralow Creek and elsewhere (McPherson et al., 2014;. The observations of the faults in the Bargo Gorge, Glenbrook Creek, Grose River, and 'Cut Rock' at Kurrajong Heights indicate that overall, the LSC is related to a west-dipping thrust fault at depth and consistent with seismic reflection profile lines and earthquakes (see above). We also emphasise that the simplest interpretation of the uplift history of the Blue Mountains as shown by uplift of the Rickabys Creek Gravel is consistent with this interpretation. We rule out the suggestion that formation of the Cumberland Basin is related to downwarping (cf. Gale, 2021) and extension. This therefore leads us to question the role in downwarping and extension in the development of coastal landscape in southeastern Australia as suggested by Branagan (2009), Ollier andPain (1994) and Brown (2000), although the details of a reassessment of this hypothesis also need to be considered elsewhere in the context of structure and landforms (e.g. NSW South Coast).

Basement control of the LSC
A previous interpretation was that the LSC is related to basement structure in the underlying Lachlan Orogen; the western margin of the Eden-Comerong 'Rift' (Branagan & Pedram, 1990). Comparison with Cooper (1992, figure 5) suggests that this is feasible, given the development of west-dipping thrusts/kink-style structures at the western margin of Eden-Comerong 'Rift' (i.e. the Budawang Syncline; Figure 12). However, structural trends in the underlying Lachlan Orogen in the southern Sydney Basin are north-northeast-trending, but further north, west of the LSC, the structural trends remain north-northeasterly, and only north of the Bathurst Granite are the structural trends more northerly (Figure 12). Thus, at least the southern part of the LSC would appear to be oblique to underlying basement structural trends ( Figure 12). However, as is apparent from Figure 12, considerable variation exists in structural trends in the Lachlan Orogen, and thus it is possible that underlying the LSC basement trends could have controlled it, although it is not possible to confidently identify just what structure and how much of it has been reactivated in forming the overlying LSC.
Deeper structure of the LSC The deeper structure of the LSC is considered to be a basement fault dipping west at 45 ; it is a brittle structure in the seismogenic part of the crust continuing to depths beyond 30 km (see Gibson, 2009, figure 4). This geometry requires that the fault is a thrust. Monoclines associated with the LSC, including the 50 km-long Lapstone Monocline, are fault-propagation folds related to trishear associated with movement on the deeper thrust fault (e.g. Erslev, 1991). The fault is depicted as a planar structure cutting through the upper to middle crust (Gibson, 2009, figure 4). In the lower Sydney Basin succession, the fault appears to continue as a planar structure at least for the southern segments of the LSC as typified by the Bargo and Nepean faults; the lack of evidence for fault-bend folding in the hanging wall implies a planar dip of the fault to depth. This is probably a result of increasing sandstone/ shale ratios that are encountered further westward in this part of the Sydney Basin (Bembrick, 1980). Thin shale horizons result in a lack of widespread flats, as may be anticipated from rheological properties of shales and widely recognised in fold-thrust belts (e.g. Butler et al., 2018).
The deeper structure of the northern segment of the LSC in the lower Sydney Basin succession is more complicated, as the problem is how to maintain the same amount of shortening to depth. The wide, planar, central monocline limb that at depth reduces in width requires that shortening be accommodated by a mechanism other than monocline formation. We propose that a series of imbricate thrusts each with relatively small offsets have maintained the shortening before linking at depth into the main westdipping thrust fault (Figure 3).

Segmentation of the LSC
A feature of many fault systems is that they consist of a segmented structure (e.g. Conneally et al., 2017;Fossen & Rotevatn, 2016;Manighetti et al., 2015;Pizzi et al., 2017), and presumably this is a reflection of the heterogeneity of rock masses and changes in the rock units encountered by faults. For the LSC, the segmented geometry is at least partly controlled by significant structures that occur at segment boundaries. For example, the Nortons Basin diatreme is at the termination of the Lapstone Monocline and the Nepean Fault (Figure 2). Similarly, the change in trend and widening of the Lapstone Monocline north of Shaws Creek ( Figure 6) coincides with the intersection of the Hornsby Warp with the LSC (Figure 1). Further south, the decreasing amount of offset along the LSC is reflected in the various fault segments around Picton, but no particular feature appears to have caused the segmentation here.

Analogues of the LSC
If the LSC were considered along with the final uplift event of the Blue Mountains (Hatherly, 2020), then a suitable analogue would be the Colorado Plateau of the southwestern United States (Fergusson, 2009). However, the Colorado Plateau is a much larger feature, over 300 km in width compared with 25 km for the Blue Mountains. The Colorado Plateau contains numerous monoclines, each associated with broad uplifts that are notable for their basement-involved style (Bump & Davis, 2003;Davis & Bump, 2009). So, the LSC and adjacent uplift are analogous to individual monoclines and uplifts in the Colorado Plateau rather than the whole plateau. The Colorado Plateau formed in a convergent continental margin setting and part of the Late Cretaceous to Paleocene Laramide Orogeny that was caused by the compressional effect of flat-slab subduction of Pacific Ocean crust under North America, with the deformation occurring 750-1500 km inboard of the active trench (English & Johnston, 2004). In contrast, the LSC has formed over 2500 km inboard of the active convergent margin in the North Island of New Zealand and the Tonga-Kermadec island arc, and is an isolated feature. Another potential analogue is the Central Bird's Head Monocline and adjacent Kemum High uplift in the Bird's Head Peninsula of northwestern New Guinea (eastern Indonesia), which also formed as a basement-cored uplift in a compressional setting (Pieters et al., 1985;Saputra, 2021).

Conclusions
The LSC is a 95 km-long association of faults and faultpropagation monoclines in the western Sydney Basin that is related to a basement thrust fault dipping to the west and underlying the Blue Mountains. The LSC has numerous segments that link at depth into the thrust fault. In the southern part of the complex, the main segments are characterised by a single west-dipping thrust (e.g. Bargo Fault) with relatively small offset (60 m) or a high-angle reverse fault (e.g. Nepean Fault) that presumably is listric and dips less steeply at depth also with minor offset (100 m). From Nortons Basin northward, the LSC is characterised by a single main monocline (the Lapstone Monocline) and is divided into two segments, a southern segment with a monocline that developed as a fault-propagation fold above the west-dipping thrust at depth, and a northern segment with a wider central limb of the monocline and shortening at depth accommodated by inferred imbricate thrust faults, each with minor offset and linking at depth into the main thrust fault. The Kurrajong Fault System represents the western extent of the fault-propagation fold. The main west-dipping controlling thrust of the LSC is seismogenic, as shown by earthquakes in the Blue Mountains in contrast to their lower abundance in the Sydney region east of the LSC.