Geologic mapping in southern Death Valley, California, demonstrates Mesozoic contractional structures overprinted by two phases of Neogene extension and contemporaneous strike-slip deformation. The Mesozoic folding is most evident in the middle unit of the Noonday Formation, and these folds are cut by a complex array of Neogene faults. The oldest identified Neogene faults primarily displace Neoproterozoic units as young as the Johnnie Formation. However, in the northernmost portion of the map area, they displace rocks as young as the Stirling Quartzite. Such faults are seen in the northern Ibex Hills and consist of currently low- to moderate-angle, E-NE–dipping normal faults, which are folded about a SW-NE–trending axis. We interpret these low-angle faults as the product of an early, NE-SW extension related to kinematically similar deformation recognized to the south of the study area. The folding of the faults postdates at least some of the extension, indicating a component of syn-extensional shortening that is probably strike-slip related. Approximately EW-striking sinistral faults are mapped in the northern Saddlepeak Hills. However, these faults are kinematically incompatible with the folding of the low-angle faults, suggesting that folding is related to the younger, NW-SE extension seen in the Death Valley region. Other faults in the map area include NW- and NE-striking, high-angle normal faults that crosscut the currently low-angle faults. Also, a major N-S–striking, oblique-slip fault bounds the eastern flank of the Ibex Hills with slickenlines showing rakes of <30°, which together with the map pattern, suggests dextral-oblique movement along the east front of the range.

The exact timing of the normal faulting in the map area is hampered by the lack of geochronology in the region. However, based on the map relationships, we find that the older extensional phase predates an angular unconformity between a volcanic and/or sedimentary succession assumed to be 12–14 Ma based on correlations to dated rocks in the Owlshead Mountains and overlying rock-avalanche deposits with associated sedimentary rocks that we correlate to deposits in the Amargosa Chaos to the north, dated at 11–10 Ma.

The mechanism behind the folding of the northern Ibex Hills, including the low-angle faults, is not entirely clear. However, transcurrent systems have been proposed to explain extension-parallel folding in many extensional terranes, and the geometry of the Ibex Hills is consistent with these models. Collectively, the field data support an old hypothesis by Troxel et al. (1992) that an early period of SW-NE extension is prominent in the southern Death Valley region. The younger NW-SE extension has been well documented just to the north in the Black Mountains, but the potential role of this earlier extension is unknown given the complexity of the younger deformation. In any case, the recognition of earlier SW-NE extension in the up-dip position of the Black Mountains detachment system indicates important questions remain on how that system should be reconstructed.

Collectively, our observations provide insight into the stratigraphy of the Ibex Pass basin and its relationship to the extensional history of the region. It also highlights the role of transcurrent deformation in an area that has transitioned from extension to transtension.

The Death Valley extensional terrane has long been recognized as a prime example for studying continental extension (e.g., Snow and Wernicke, 2000; Miller and Pavlis, 2005) due to spectacular rock outcrops, abundant markers that can be used as pre-extensional piercing points and evaluate displacements, and its relatively young age, allowing tight age constraints on deformation. Nonetheless, because any correlation, be it stratigraphic or structural, is an interpretation open to debate, there are questions that surround the nature of the geologic structures that accommodated Neogene extension in the Death Valley region. Understanding these structures has been a focus of numerous studies (e.g., Wernicke, 1981; Wernicke et al., 1988; Topping, 1993; Miller and Prave, 2002; Miller and Pavlis, 2005; Pavlis et al., 2014), and there is no consensus on the kinematic history of extension across the region. The rolling hinge model is a widely cited concept applied to this region, and this kinematic model is coupled with an inference of ~140 km of NW-directed extension across the Death Valley extensional terrane based on a suite of inferred offsets of Mesozoic and older markers (e.g., Snow and Wernicke, 2000). Inherent in the rolling hinge model is the interpretation of a single, regional detachment that underlies the region. In contrast, other workers have emphasized alternative explanations for the region's geology. Specifically they’ve suggested that displacement occurring along strike-slip faulting in the region, along with multiple, distinct detachment horizons, result in the development of discrete pull-apart basins (e.g., Wright et al., 1991; Serpa and Pavlis, 1996; Fridrich and Thompson, 2011). These models typically suggest significantly less extension than the rolling-hinge family of models, in most cases using many of the same pre-extensional markers but different interpretations of those markers (Topping, 1993; Serpa and Pavlis, 1996; Miller and Pavlis, 2005; Fridrich and Thompson, 2011; Renik and Christie-Blick, 2013). The refinement of these models for the areas within the Basin and Range Province, and the amounts of extension in them, are important for our broader understanding of continental extension and transtension. For example, in the case of Death Valley, the rolling-hinge model has been used to describe large-scale deformation of the Basin and Range region (e.g., Wernicke et al., 1988). Therefore, additional study is warranted to test alternative reconstructions.

The Ibex Hills are located in the southern part of the Death Valley extensional terrane and occupy a structural position that varies radically among different reconstructions. Two structural features in particular are especially important in this context: the Black Mountains detachment and the strike-slip Sheephead fault (Fig. 1). Holm and Wernicke (1990), Holm et al. (1992), Topping (1993), and Fridrich and Thompson (2011) interpreted low-angle normal fault systems in the Ibex Hills (Fig. 2) as the up-dip, easternmost, portion of the Black Mountains detachment system, exposing the basement gneiss in the footwall of the detachment. This interpretation has been used to support the detachment fault model with an implication for significant denudation and subsequent exposure of the footwall in the western Black Mountains (Fig. 2) (Wernicke, 1981; Snow and Wernicke, 2000). Alternatively, Troxel et al. (1992) recognized dikes and fault geometries immediately to the south of the Ibex Hills, in the Saddle Peak Hills (Fig. 1), indicative of NE-SW extension that predated NW-SE extension. This NE-SW extension is widely interpreted to be coeval with slip along the Kingston Range detachment immediately to the southeast (Davis et al., 1993), including the Miocene extensional basin system, the China Ranch basin (Scott et al., 1988). The Sheephead fault is adjacent to the Ibex Hills, just to the north, and has been cited as playing a role as a displacement transfer structure in the extension (e.g., Wright et al., 1991; Serpa and Pavlis, 1996). Nonetheless, the sense of movement on the Sheephead fault has been debated with some authors supporting sinistral slip (e.g., Holm and Wernicke, 1990; Serpa and Pavlis, 1996), while the most recent work has interpreted the fault as an overall dextral slip system (Renik, 2010), consistent with the tectonic model of Wright et al. (1991).

Given the Ibex Hills proximity to the Sheephead fault, along with its interpreted relationship to the Black Mountains detachment, an understanding of the geologic history of the range should provide better constraints on the tectonics of the area. For example, the role of strike-slip faulting, as compared to normal faults and their related detachment surfaces in the extensional history of the Death Valley region has been a point of debate (e.g., Miller and Pavlis, 2005). Despite the pivotal location of the Ibex Hills, most recent work (e.g., Corsetti and Kaufman, 2005; Petterson et al., 2011) has been limited to topical stratigraphic studies lacking a focus on the regional structure at the heart of reconstruction controversies. The most detailed published mapping in the Ibex Hills is that of Wright and Troxel (1968); however, this mapping involved only a small portion of the range with a focus on talc deposits in the region. Therefore, an analysis of the structures of the Ibex Hills is long overdue and was the goal of the work we present herein.

This study presents geologic mapping of the Ibex Hills, Saratoga Hills, northern Saddlepeak Hills, and the intervening topographic lowland, the Ibex Pass basin, which provides clear indications of a complex structural history (Plate 1). We begin with a description of the major structures and crosscutting relationships that, together with stratigraphic relationships in the associated basin system, establish two phases of Mesozoic contractional deformation and two phases of extension. The Neogene structures are also clearly associated with at least two phases of kinematically distinct strike-slip systems that include the Sheephead fault system (Fig. 1). We then assess the regional correlation of these structures and discuss their significance and relation to the Black Mountains and Kingston Range detachments (Figs. 1 and 2). In doing so, we add to the understanding of the tectonic history of the Death Valley region, which should guide future work and aid in assessment of extensional tectonic processes based on the region (e.g., Lutz et al., 2019, 2020).

The Death Valley region exposes rocks that range from middle Proterozoic to Quaternary in age. After the middle Proterozoic events that formed the crystalline basement for the region (Heaman and Grotzinger, 1992), the area experienced two cycles of sedimentary deposition: (1) middle Proterozoic deposition of the Crystal Spring Formation; and (2) Neoproterozoic to late Paleozoic strata that form a thick, miogeoclinal succession that is widely recognized as the passive margin assemblage deposited during the breakup of Rodinia (Stewart, 1972; Wright et al., 1974; Corsetti and Kaufman, 2003). The latter produced the bulk of the sedimentary sequence recognized in southern Death Valley; this sequence includes the Neoproterozoic Horse Thief Spring Formation through middle Paleozoic units of the Wood Canyon and Zabriski Quartzite Formations (Fig. 3).

During the latest Paleozoic and throughout the Mesozoic, much of western North America underwent a series of contractional deformation events. The Mesozoic events are collectively known as the Sevier orogeny (Armstrong, 1968; Fleck, 1970; DeCelles, 2004). The Sevier orogenic belt is a zone of generally thin-skinned thrusts that extend from southern Canada into eastern California (DeCelles, 2004). In the Death Valley area, the Sevier fold-thrust system is complicated and appears to change strike from northeast to northwest (Burchfiel and Davis, 1971; DeCelles, 2004). While some authors have included the northwest-striking structures as part of the Sevier orogenic belt (e.g., DeCelles, 2004), others have proposed they are, in fact, a younger set of Laramide thrusts overprinting older Sevier structures (Miller, 2003; Pavlis et al., 2014). While previous work suggests that overall contractional deformation in the Death Valley region is thought to have ended no later than 70 Ma (Fleck, 1970; Walker et al., 1995), Pavlis et al. (2014) challenged that view with a hypothesis of significant Laramide-age deformation in the region.

Following Mesozoic–Paleogene deformation, the Death Valley region was part of a Cordilleran highland analogous to the Andean Altiplano, commonly referred to as the Nevadaplano (Ernst, 2009). Erosion during this interval variably exhumed the previously deformed miogeoclinal strata, recorded now as a widespread unconformity beneath Neogene strata with rocks beneath the unconformity, ranging in age from Proterozoic basement to Cambrian (e.g., Pavlis et al., 2014).

Neogene extension associated with the Basin and Range began in the Miocene and continues today as active transtension (e.g., Burchfiel and Stewart, 1966; Norton, 2011). Despite the consensus that the region is undergoing transtension at the present time, there are still disagreements about the kinematic histories of Neogene structures that make up the Death Valley extensional terrane (e.g., Holm and Wernicke, 1990; Holm et al., 1992; Miller and Pavlis, 2005). The primary debate centers on the origin of low-angle detachment systems that are exposed along the eastern side of Death Valley in the Black Mountains (Figs. 1 and 2). In that area, the detailed mapping by Wright and Troxel (1984) formed the basis of some of the early hypotheses on the role of low-angle normal faults in extension (Wright and Troxel, 1973). Further studies (e.g., see review in Miller and Pavlis, 2005) revealed that successively deeper structural levels are exhumed from SE to NW across the Black Mountains, from high-grade metamorphic rocks in the Death Valley turtlebacks to low-temperature brittle deformation in the SE. Although this observation supports the concept of a single detachment system across the Black Mountains (e.g., Holm and Wernicke, 1990), several observations are at odds with this generalization. Serpa and Pavlis (1996) noted that structures within the Death Valley turtlebacks in the central Black Mountains showed evidence of an exhumed, deeper-level detachment that formed prior to the main Black Mountains detachment system. Moreover, Miller and Prave (2002) and Miller and Pavlis (2005) argued against the occurrence of even a single detachment, arguing instead for complex block rotations within a 3D, transtensional environment. Additional complications include significant structural relief on fault systems within the central Black Mountains, which include the antiforms of the Death Valley turtlebacks and curviplanar fault surfaces within the Amargosa chaos (Fig. 1). These curviplanar surfaces are often assumed to be primary corrugations in fault systems (Otten, 1976), but Mancktelow and Pavlis (1994) and Serpa and Pavlis (1996) presented evidence that these features are folds that developed coincident with the extension due to distributed transcurrent motions accompanying the extension.

The Kingston Range detachment system (Fig. 1) represents the most significant extension that predates northwest-southeast extension, and it is exposed most clearly along the front of the Kingston Range and farther south in the Halloran Hills (e.g., McMackin, 1992; Davis et al., 1993; Fowler and Calzia, 1999). Initial discussion of detachment faulting in the vicinity of the Kingston Range placed such structures within the context of major northwest-southeast extension in the region (Burchfiel et al., 1983). However, later work on the Kingston Range detachment showed a top-to-the-southwest sense of movement (e.g., McMackin, 1992; Davis et al., 1993). The timing of movement along the Kinston Range detachment is constrained locally by a 13.4 Ma hypabyssal sill, which is cut by the detachment, and the 12.4 Ma Kingston Peak pluton, which crosscuts the detachment (Calzia, 1990; Fowler and Calzia, 1999; Calzia and Ramo, 2000).

To the west and northwest of the Kingston Range lies a suite of Tertiary-aged basins that have been key in understanding the evolution of southern Death Valley (Fig. 1) (e.g., Wright, 1974; Scott et al., 1988; Topping, 1993; Prave and McMackin, 1999; Fridrich and Thompson, 2011). These basins include the China Ranch Basin (Wright, 1954; Scott et al., 1988) and the Dumont Hills Basin (Prave and McMackin, 1999) (Fig. 1); however, other work has envisioned these basins as part of a larger-scale feature, the Greater Amargosa-Buckwheat–Sperry Hills Basin (GABS Basin) (Holm et al., 1994). In the latter context, the GABS Basin has since undergone significant attenuation through a combination of the Black Mountain detachment and strike-slip on the Grand View fault (Topping, 1993; Holm et al., 1994). In general, these basins record the deposition of a sequence of alluvial fans and fanglomerates, coarse sandstones, and fine sandstones and mudstones (e.g., Prave and McMackin, 1999). Understanding the evolution of this basin system is important for the competing models of Death Valley extension. For example, interpretation of these Tertiary basins as half-grabens containing asymmetric basin facies is consistent with later extension along a detachment (e.g., Topping, 1993). Alternatively, other authors documented evidence of at least two, more symmetric, source directions for the fanglomerates of the basins, more consistent with a strike-slip–driven, “pull-apart” model of basin development (Prave and McMackin, 1999).

The presence of such a suite of complex features in the southern Death Valley region suggests an intricate extensional history. In line with that suggestion, the structure of the area is complicated by a number of crosscutting relationships and apparent structural overprinting. However, the body of published works concerning the southern Death Valley area is limited. For example, the Ibex Hills, a focus of this study, have been mentioned in relationship to the Black Mountain detachment (e.g., Holm and Wernicke, 1990; Snow and Wernicke, 2000); yet there is little published mapping of the area. Published mapping in the Ibex Hills is limited to localized mapping of talc deposits by L. Wright in the 1960s (Wright and Troxel, 1968) and reconnaissance mapping for the 1:250,000 Trona sheet (Jennings et al., 1962). These maps predate concepts of low-angle normal faults with most normal faults shown as thrust faults; yet the faults exhume crystalline basement in their footwalls, consistent with significant extensional exhumation. The presence of these low-angle faults has resulted in the interpretation that they represent an up-dip extension of the Black Mountain detachment (e.g., Holm and Wernicke, 1990; Snow and Wernicke, 2000), implying these faults represent the product of top-to-the-NW extension younger than ca. 11 Ma (Fig. 2A).

To the south of the Ibex Hills, Troxel et al. (1992) recognized that top-to-the-SW movement and low-angle normal fault systems dominated the structure of the Saddlepeak Hills. Recent mapping and compilations of that geology (Mahon and Link, 2013; Mahon et al., 2014) confirm that conclusion, indicating older extension related to the Kingston Range detachment continued to at least the Saddlepeak Hills (Fig. 1) and presumably into at least the basinal deposits to the south of the Ibex Hills. Here we examine the hypothesis that this deformation extended even farther north, into the Ibex Hills themselves (Fig. 2B).

Geologic Mapping and Cross-Section Construction

Field mapping was done electronically via a tablet computer running QGIS with a data structure similar to that described by Pavlis et al. (2010). Mapping was aided by georeferenced topographic maps and orthophotography data layers viewed in real time. Topographic maps were U.S. Geological Survey (USGS) digital raster graphics (DRGs) scanned from 7.5′ quadrangle maps, and orthophotography was from USGS, ArcGIS Online, Google maps, and Bing maps. Spatial positioning was aided by a Wide Area Augmentation System (WAAS)–corrected GPS that provided real-time positioning through the field application. The geologic map (Plate 1) is a compilation of mapping by Canalda (2009), Pavlis (2004, 2011, 2015, 2017), and Fleming and Pavlis (2018) with the pre-2018 data incorporated into an ArcGIS geodatabase through a combination of direct copy to the master and redigitizing linework to produce a final map. In addition, a portion of the northern Ibex Hills mapping was essentially a revamped version of the work of Wright and Troxel (1968) using modern orthophotos and digital elevation models (DEMs), along with some recent reconnaissance mapping.

The contact lines and orientation data were exported from QGIS as shapefiles and were then imported into Move 2017 ( A digital DEM was downloaded from the USGS National Map Download Client ( A 1° × 1° DEM of the area was imported into Move along with the geologic mapping data. In the process of the import, the DEM was cropped to only the study area. Within the Move program, the contacts and orientation data were then projected vertically to the DEM to produce a 2.5-dimensional surface model of the geologic features.

Prior to construction of cross sections, the stratigraphy of the study area was entered into Move. These data included the names of the units, their thicknesses, and their ages. The ability to preload stratigraphic data into Move is a useful feature which allows the user to draw a line in cross section, assign it a unit, and then construct the units above and below automatically, based on the provided stratigraphy. It is important to note, however, that the stratigraphic thicknesses of some of the mapped units, namely the Beck Springs Dolomite, the Kingston Peak Formation, and the Noonday Formation, change dramatically within the study area. Therefore, the thicknesses entered into the Move program were not always used in the projections of the cross sections. To counteract this issue, the program allows for the change of unit thicknesses “on the fly” when constructing different sections, thereby not affecting the thicknesses for other areas of the map.

Kinematic Analysis

To analyze the fault along the eastern flank of the Ibex Hills, herein referred to as the Eastern Boundary fault, we used slickenside measurements within the well-exposed fault zone to evaluate the kinematics (Table 1). In this study, we use the modified odd-axis method of Bruhn et al. (2004), which is based on the original method of Krantz (1988, 1989). The modified odd-axis method assumes fault sets within a fault zone form a rhombohedral family of surfaces that slip in response to an incremental strain field. In this method, the fault planes, slip vectors, and the poles to the movement plane for each fault (equivalent to the B axis of a focal mechanism) are plotted on a stereonet. This method results in clusters of poles to fault planes to which a best-fit pole of the clusters can be determined (e.g., Carvell et al., 2014). When applied to fault zone analyses, the method typically yields either a simple cluster of B axes within the fault plane consistent with plane strain or a great-circle distribution of B axes that intersects the fault plane representing approximately uniaxial stress (Bruhn et al., 2004). The odd axis is the pole to the best-fit plane of the B axes (Krantz, 1988, 1989). The odd axis is either contractional (maximum strain) or extensional (minimum strain), which is then determined with slickenlines of a known slip sense. In both cases, the fault slip vector is simply 90° away from the B-axis maximum (plane-strain case) or the fault plane/B-axis intersection, measured in the fault plane (Bruhn et al., 2004). The intermediate and maximum (or minimum) extension directions are defined as the obtuse and acute bisectors of the movement plane pairs (Krantz, 1988; Carvell et al., 2014). The advantage of this approach compared to conventional kinematic analysis techniques (Marrett and Allmendinger, 1990) is that all slickenside measurements can be used in the analysis, even if only a few have known slip sense. In the case of the Eastern Boundary fault, we prefer the “odd-axis” method, given the well-exposed fault zones that contain a multitude of slip surfaces, which can then be analyzed as a whole.

In addition to the odd-axis method, we also made use of fault bedding cutoff lines to assess the geometry of the normal faults in the Ibex Hills, a method also used in observations of the Kingston Range (Fowler and Calzia, 1999). The reasoning for this method is that a NW-striking fault would have cutoff lines trending 90° from those of a NE-striking fault, assuming the beds they cut have a similar orientation. Therefore, if distinct groups of cutoff trends are observed, they could suggest multiple fault orientations that may be associated with different phases or extension directions. While the Ibex Hills experienced pre-extensional deformation, evidence that we support in later sections of this paper, we believe assessing the geometry of normal faults in the Ibex Hills is a useful tool, especially when used in conjunction with the other evidence we present.

Map Units: Stratigraphy of the Mapped Area

Figure 3 shows a generalized stratigraphic section for the area. Formal stratigraphic units for this area are too thick for routine mapping at scales larger than ~1:100,000, and thus, like previous workers (e.g., Wright and Troxel, 1984), we divided the formal units into subunits for more detailed mapping. The details of how we divided the units can be found in the Supplemental Material1, but two features are important in evaluating our geologic map:

  1. Neoproterozoic rock units below the Johnnie Formation (Fig. 3) vary dramatically in thickness, facies, or both, across the mapped area. Thickness variations are predominantly due to variable erosion across major disconformities at the base of the Kingston Peak Formation, within the Kingston Peak Formation, and the base of the Noonday Formation (Fig. 3), including syndepositional structure described in this paper. Lateral facies variations are prominent in the Kingston Peak and Noonday Formations due to syn-rift deposition (Macdonald et al., 2013).

  2. Neogene units in the Ibex Hills sub-basin have a complex stratigraphy, but the stratigraphic sequence is recognizable in only three sub-areas, with each recording a distinct section subject to interpretation (Fig. 4). Details are found in the Supplemental Material (footnote 1); but for the purposes of this paper, the key observation is that this succession is composed of two sequences separated by an angular unconformity: a lower volcanic sequence, tentatively correlated to the Wingate Wash Volcanics exposed to the west and north (Luckow et al., 2005; Canalda, 2009) and an upper sedimentary succession composed of interbedded sediments and landslide megabreccias. Based on our mapping, illustrated in Figure 4, Domain 1 is located in the northern Saddle Peak Hills–Ibex Pass area and is defined by a volcanic sequence overlain by two megabreccias, themselves separated by a layer of gravels. Domain 2, to the west, contains similar gravels but also a sequence of fine-grained clastic rocks in the lower portion of the section, along with distinct megabreccias of Noonday and Beck Spring Dolomites (Fig. 4). Domain 3, to the north of Domains 1 and 2, again contains gravels overlying megabreccias; but here the volcanic sequence, seen below the sediments of Domains 1 and 2, is not observed. Instead, the gravels of the basin lie unconformably atop the tilted bedrock of the Ibex Hills (Fig. 4).

Structure of the Ibex Hills Area

General Map Relationships

From the Saratoga Hills, northward to the central Ibex Hills (Plate 1), the rocks are structurally coherent and dominated by steeply dipping rocks composed primarily of the Proterozoic Pahrump Group. Bedding strikes are generally NNE but curve from NE to NS across the Saratoga Hills, indicating a large, steeply SE-plunging, open syncline. Additional structural complexity is indicated in the northern Saratoga Hills, where the Noonday and Johnnie Formations lie in low-angle fault contact with the Crystal Springs Formation and the Beck Springs Formation (Plate 1), a stratigraphic omission of at least 500 m of Kingston Peak Formation, indicating that these are low-angle normal faults. Significantly, these low-angle normal faults are also cut by an array of higher-angle faults that strike both northeast and northwest (Plate 1). The Tertiary sedimentary rocks deposited in the adjacent basin are presumably concurrent with the slip along these low-angle faults, but direct observation of this relationship is obscured in the northern Ibex Hills by the presence of the Eastern Boundary fault, which cuts the basin deposits up to the youngest megabreccias and is only overlain by the Quaternary units. A description of these units is provided in Plate 1.

In the northern Ibex Hills, the map pattern is significantly different and is dominated by the Noonday and Johnnie Formations, along with small exposures of younger rocks, especially in the northeast corner of the mapped area (Plate 1). Structural complexity increases dramatically to the north from relatively simple, homoclinal rocks of the Saratoga Hills and central Ibex Hills to complexly faulted rocks in the northern Ibex Hills (Plate 1). The orientations of the rocks in the north also change at a variety of scales, primarily due to multiple faults with complex crosscutting relationships. Nonetheless, bedding dips generally are shallower than those to the south (Plate 1).

Sinistral Faulting at the Southern Margin of the Ibex Pass Basin

The northern Saddlepeak Hills expose the southern margin of the Ibex Pass basin in a composite boundary that includes both extensive exposures of the basal Neogene unconformity and fault contacts. Throughout this area, the basal unconformity is extensively exposed along a gently north-dipping contact (Fig. 4; Plate 1) marked by either volcanic rocks lying directly on Proterozoic rocks or a 0–30-m-thick gravel lying between the volcanics and Proterozoic. Just to the south of the unconformity, however, is an array of ~EW-striking, subvertical faults that show subhorizontal slickenlines and have consistent sinistral offsets of contacts in the east-dipping Proterozoic rocks. Thus, we interpret this array of faults as a sinistral fault system, and the net slip across the zone is ~1.5 km based on offsets of the Johnnie-Stirling contact.

The relative age of this sinistral fault array is well established by crosscutting relationships with intrusives and the unconformity. Specifically, although some of the faults in the array are overlapped by the unconformity (Fig. 4), others cut the unconformity. Similarly, a small intrusive body cuts faults in the array (Fig. 4), while just to the west, a different intrusive rock is fully involved in the fault array (Fig. 4). Finally, at the northwest tip of the Proterozoic exposures in the northern Saddlepeak Hills (Plate 1), a low-angle normal fault places the upper Stirling Formation on middle Johnnie Formation but is cut by the sinistral fault array. Because this low-angle normal fault shows cutoff lines with a NW trend, we conclude this fault records top-to-the-SW motion similar to faults seen just to the south in the core of the Saddlepeak Hills (Mahon and Link, 2013). However, because of the mutual crosscutting relationships of the faults with the unconformity and igneous assemblage, we conclude that the sinistral fault array is also related to the SW-directed extensional event, presumably as a displacement transfer structure.

The unconformity seen at the base of the Neogene in the northern Saddle Peak Hills is notably absent to the north in the northern Ibex Hills (i.e., Domain 1 of Fig. 4). Instead, in Domain 1 (Fig. 4), the Proterozoic is directly overlain by Tertiary gravels, and the volcanic sequence seen in both the southern Ibex Hills and Saddle Peak Hills is absent. We suggest that this relationship necessitates a fault between Domains 1 and 3 (Fig. 4) structurally below the upper gravel sequence. Incongruities between Domains 2 and 3 are also likely the result of faults within the basin (Fig. 4). These incongruities include differences in the thicknesses of volcanic sequences below a conspicuous tuff layer (Fig. 4). In the southern Ibex Hills, Domain 2, the sequence below the tuff is significantly thinner than that seen in the northern Saddle Peak Hills, Domain 3 (Fig. 4). For a complete description of the Tertiary map units, see the Supplemental Material(footnote 1).

Eastern Boundary Fault and Ibex Pass Basin

The Ibex Hills are bounded to the east by the steeply to moderately dipping Eastern Boundary fault (EBF) (Plate 1), which separates the Ibex Hills from the adjacent Ibex Pass basin. In the south, the fault is moderately well exposed in discontinuous exposures along the NE-trending, eastern flank of the Ibex Hills. Along this segment, the fault curves from a strike of ~015° near Ibex Spring in the south to ~030° near the Giant Mine with relatively uniform dips of 45°–60°. This geometry produces low-angle, fault-bedding intersections in the Proterozoic rocks. Thus, despite clear evidence of significant down-to-the-SE slip based on the outcrops of Cenozoic rocks in the basin, the stratigraphic shifts across the fault are modest below the Tertiary unconformity (e.g., Crystal Springs against Crystal Springs).

In the northern part of the Ibex Hills, the EBF becomes complex with contradictory crosscutting relationships and complex structure in both the hanging wall and footwall. From the Giant Mine to just south of the Eclipse Mine (Fig. 5; Plate 1), the fault continues at a strike of ~030° but clearly truncates a series of moderate- to high-angle, N-NW–striking normal faults in both the hanging wall and footwall. On the map (Plate 1), these faults are obvious in the footwall but are partially obscured by younger sediments in the hanging wall, including the uppermost landslide megabreccias (Fig. 5; Plate 1). The most significant of these hanging-wall structures (labeled X in Fig. 5) show evidence of extended slip during the deformation with consistent, east-side-down slip. That is, the unconformity beneath the Neogene gravels shows a small shift indicating east-side-down, but rocks beneath the unconformity show a much greater shift across the fault with Crystal Springs on the west and Johnnie Formation to the east (Plate 1). Three other NNW-trending faults show similar sub-unconformity shifts in this segment as well as a complex array of faults with smaller shifts. Clear relative age relationships are present along the segment of the fault between the Giant Mine and east of the Comet Mine where a fault sub-parallel to the EBF cuts the younger gravels (Tg2); yet, the fault and gravels are overlapped by the Beck Springs megabreccias (Tmb) (Supplemental Material [footnote 1]) (Fig. 5; Plate 1). That landslide megabreccia, in turn, is also cut by the EBF. Thus, the EBF in this segment is a pair of faults; but slip on the main segment of the EBF outlasted slip on the fault to the SE.

Just south of the Eclipse Mine (Plate 1), the EBF is associated with complex faulting that can be interpreted in several ways. In that area, the NE-trending EBF to the south meets a fault system striking ~340° with a moderate (30°–50°) east dip. This NNW-trending fault segment is either part of a curved EBF or represents a distinct fault system with mutual crosscutting relationships to the EBF. The curved fault hypothesis is supported by the NNW-trending fault showing similar geomorphic expression as the EBF, although there are no clear displacements of Neogene rocks along this segment to confirm this hypothesis. In contrast, the NNW-trending fault is clearly continuous with a fault that continues through the bend to the south as an apparently continuous structure, placing Stirling Quartzite to the east against Johnnie and Stirling Formations to the west. Nonetheless, structures with trends similar to the southern segment of the EBF also continue NE from the bend.

Finally, even farther north in the Eclipse Mine area, the NNW segment of the EBF intersects a second zone of NE-striking faults. Like the bend to the south, this zone displays contradictory crosscutting relationships and allows multiple interpretations. Nonetheless, traces of the NNW segments appear to be present up into the northernmost Ibex Hills and consistently place Stirling Quartzite and younger formations along the eastern flank of the range. Given this correlation, the Eastern Boundary fault appears to have an approximate heave of 4 km, based on the map distance between the Stirling Quartzite in the central Ibex Hills to that in the north (Plate 1). Complicating this correlation, however, is the loss of the high-angled EBF in the northern Ibex Hills, which suggests that the fault may become part of the currently low-angle fault system (Plate 1).

Kinematic analysis on the Eastern Boundary fault in the northern Ibex Hills and along the eastern front near the Wonder Mine (Plate 1) shows an orthorhombic fault geometry within the fault zone (i.e., two sets of conjugate fault systems), albeit with a more dominant N-NE–striking set (Fig. 6). In general, the slickenlines have rakes of 30° or less, mostly from the S-SW (Fig. 6). The B axes show a distinct cluster (Fig. 6B) plunging moderately to steeply to the north. With a master fault dipping steeply to the ESE, this B-axis cluster, together with the corresponding odd axis (Fig. 6B) and known stratigraphic shifts, indicate the odd axis is an extension direction, and the eastern boundary fault is an oblique, dextral-normal slip fault. This is consistent with the average orientation of the fault.

Collectively, these observations suggest the EBF is a composite structure with the main fault to the south representing an oblique, dextral-normal fault that interacts complexly with a NNW-striking fault system to produce two apparent bends in the fault trace of the EBF. However, based on the kinematic analysis of the southern portion, south of the Comet Mine area, a component of dextral movement clearly displaces units as young as Stirling Quartzite. This fault relationship is maintained to the north, albeit with some complex crosscutting relationships.

Younger Normal Faults

A set of nearly orthogonal high-angle normal faults crosscuts the low-angle normal faults of the northern Ibex Hills (Fig. 5). The dips of these faults range from ~50° to >80° (Fig. 7). The majority of the measured faults in the northern Ibex Hills strike to the northeast with a range of ~030° to ~070° with a second group striking predominately north-south to northwest-southeast, consistent with mapped faults (Plate 1). These N-S–striking faults generally dip at higher angles than the predominately NE-striking faults, with many near vertical (Fig. 7).

Fault-bedding cutoff angles for the high-angle normal faults in the northern Ibex Hills are generally >60° with a few as low as 30° (Fig. 8). The range of cutoff angles is presumably due to the dip of beds before faulting. Cutoff lines for these faults also consistently trend NE-SW (Fig. 8), consistent with an origin as post–11 Ma, extensional structures developed during NW-SE extension. Figure 8E shows the expected geometry of that scenario, in which east-dipping beds are cut by a NE-striking normal fault, thus producing NE-SW cutoff lines.

The influence of both the northeast- and northwest-striking normal faults in the northern Ibex Hills can be seen most clearly in the southern portion of the area (Fig. 5), where they cut into basement rock and displace the Crystal Springs Formation and the overlying Horse Thief Springs Formation. The faulting in the area results in a sequence of blocks in which the hanging walls of the faults are being dropped successively down to the northwest (Fig. 5; Plate 1).

Despite the similar orientations of the higher-angle normal faulting in the southern Ibex Hills to that in the northern Ibex Hills (Plate 1), the crosscutting relationships are somewhat different. In the southern Ibex Hills, a pair of prominent, NW-striking, high-angle faults clearly truncate the north-northeast–striking faults (Plate 1). These N-NE–striking normal faults in general have a down-to-the-NW sense of motion. However, some have a minor down-to-the-SE sense of displacement. Farther south in the Saratoga Hills, the high-angle, W-NW–striking faults appear to be truncated by NE-striking normal faults that down drop the Johnnie Formation in the hanging wall (Plate 1).

Low-Angle Normal Faults

Observations. We recognize five sets of currently low-angle normal faults in the northern Ibex Hills (Fig. 5; Plate 1). Faults A–D (Fig. 5) form a stack of subparallel faults that is best displayed in the central part of the northern Ibex Hills (Plate 1). The structurally lowest fault in this sequence (Fault A, Fig. 5) places the middle Crystal Springs Formation in the hanging wall against both crystalline basement and the lower Crystal Springs in the footwall. In map view (Plate 1), this fault, where it is exposed, is essentially bed parallel in the hanging wall but has mostly basement rock in the footwall, with only a small isolated lens of lowermost Crystal Springs Formation present. The continuation of this fault to the south is ambiguous because it is truncated by a series of moderately NW-dipping normal faults that produce multiple repetitions of the nonconformity beneath the Crystal Springs. This crosscutting relationship demonstrates that Fault A is older than these NW-dipping faults; however, map geometry is ambiguous on the kinematics of Fault A.

Structurally above Fault A is a series of subparallel, low-angle faults that carry Horse Thief Springs, the Mahogany Flats member of the Noonday Formation, and the Johnnie Formation in their hanging walls, labeled Faults B, C, and D, respectively (Fig. 5). Faults C and D show distinct fault intersections with branch lines plunging NE, whereas those of Fault B generally trend NW. This sequence of stacked faults disappears to the northwest where rocks in the hanging wall of Fault D (Johnnie Formation) overlie fault slivers of the Noonday Formation, in the Fault C hanging wall; these slivers in turn lay on crystalline basement. The structurally highest low-angle fault—Fault E (Fig. 5)—carries Stirling Quartzite and Wood Canyon Formation in its hanging wall, juxtaposed on middle Johnnie Formation in the footwall. Fault E is only exposed in the NE corner of the mapped area (Plate 1).

The NE-striking faults discussed in the previous section crosscut all structural levels of the low-angle normal faults in the mapped area (Fig. 5). However, the timing is somewhat ambiguous because they crosscut the low-angle Faults A and B. Crosscutting relationships with Faults C and above were not clearly observed (Fig. 5). In addition, the relationship between the low-angle faults and the steep frontal fault to the east, the Eastern Boundary fault (EBF), is not entirely clear (Fig. 5). In the central Ibex Hills, the EBF crosscuts Fault B (Fig. 5). However, in the northern Ibex Hills, the exact trace of the EBF is difficult to decipher, and relationships to structural higher faults are unclear.

Analysis of the cutoff angles of the low-angle normal faults shows that for the 14 faults analyzed, half of the faults have high-angle cutoffs >45°, but 36% have cutoff angles of less than 20° (Fig. 8C). In contrast, all of the younger high-angle faults have cutoff angles greater than 30° (Fig. 8D). These observations can be interpreted a number of ways. If bedding had relatively low dips prior to extension, then approximately half of the low-angle normal faults nucleated as low-angle faults with the remainder nucleating as high-angle faults. Given the map pattern and observation of variable angular discordance beneath the Tertiary unconformity, it is more likely these variations reflect differences in the initial bedding dips, prior to the extension. Nonetheless, the orientation of the cutoff lines for the low-angle faults shows a general south trend, scattered along a low-dip great circle (Fig. 8A), which is distinct from the shallow NE-SW trends of cutoff lines for the younger high-angle faults (Fig. 8B). Most significant, however, are SE-trending cutoff lines for Faults A and B (gray dots, Fig. 8A), which are distinct from the general NE-SE trend of other cutoff lines (Figs. 8A and 8B), suggesting Faults A and B originally had NW-SE strikes. Figure 8F illustrates this concept by showing a NW-SE–striking fault cutting east-dipping beds; this fault produces SE-trending cutoff lines.

Faults A and Fault B both cut down stratigraphic section to the south (Fig. 9). Fault B cuts down section in the hanging wall from the basal Noonday unconformity in the north, through the Horsethief Spring in the south (Fig. 9), with a similar pattern in the footwall cutting downsection from uppermost Crystal Springs to near the base of the upper member in the south (Fig. 9; Plate 1). The hanging wall of Fault A is partially obstructed by the offset of Fault B, making cutoffs less clear than Fault B. Nonetheless, west of the Comet Mine (Plate 1), Fault A cuts downsection to the south from the upper Crystal Springs to the lower Crystal Springs (Fig. 5; Plate 1). The hanging-wall cutoff relationships are particularly clear at the arrow in Figure 4 where the lower-upper Crystal Springs contact is cutoff along the fault. Collectively, these observations of the orientations of bedding cutoff lines (Fig. 8A) and map-scale stratigraphic juxtapositions suggest that Faults A and B originated as NW-striking normal faults, recording an older phase of NE-SW extension. Either observation alone could be interpreted as an artifact of pre-extensional structure, but together, they are difficult to reconcile as products of NW-SE extension. For example, regional relationships (Pavlis et al., 2014) and sub-Tertiary unconformity geology (Plate 1) suggest that prior to extension, the units dipped NE, and, thus, a NE-striking normal fault, which would be expected to form from NW-SE extension, could produce the observed map-scale relationships. However, cutoff lines would trend N to NE, nearly 90° from measured cutoffs (Fig. 8).

The map-scale relationships are different for structurally higher faults. The geometry of Fault C is partially obscured by down-to-the-northwest slip along northeast-striking normal faults as well as the cutoff by Fault D. However, Fault C does appear to cut out a section of the Radcliffe Member of the Noonday Formation toward the northwest, suggesting geometry comparable to Faults A and B. Nonetheless, because of lateral stratigraphic changes of the Noonday (e.g., Corsetti and Kaufman, 2005), it is possible this juxtaposition is a combined stratigraphic and structural effect. Fault D has a more clearly defined geometry. In the east-central portions of the range, the oolite marker bed in the upper Johnnie is placed directly against the Mahogany Flats Member of the Noonday along Fault D (Fig. 9; Plate 1). Farther to the north and west, Fault D cuts downsection in the hanging wall, placing the middle member of the Johnnie Formation on Noonday (Plate 1). Faults C and D are interpreted to merge within the mapped area and thereby place the Johnnie and Noonday Formations onto basement in northernmost portions of the mapped area (Plate 1). However, Quaternary cover in the wash in the north partially obscures the contact. In addition, there are high-angle fault surfaces at the northern and northwest extent of the Noonday Formation, suggesting the possibility of a younger structure that is responsible for juxtaposing the formations against basement rock.

Although low-angle normal faults are most prominent in the northern Ibex Hills, low-angle normal faults are also present in the southernmost Ibex Hills and northern Saratoga Hills (Plate 1). The low-angle structures in this area carry the Johnnie Formation in the hanging wall and place it against the Crystal Springs Formation in the southern Ibex Hills and the Beck Springs Formation in the northern Saratoga Hills (Plate 1). Along the northern flank of the Saratoga Hills, this fault dips NW at ~45°, whereas the fault in the southern Ibex Hills dips to the SE at a similar angle, although the dip of the fault appears to steepen to the north (Plate 1).

Interpretations. These field relationships suggest strongly that Faults A and B, and probably C, formed during NE-SW extension. This is supported by the stratigraphic cutoff lines that trend south to southeast (Fig. 8), consistent with faults that cut downsection to the southwest (Fig. 9). Additionally, Fault B carries the conspicuous basal unconformity of the Noonday Formation above a nearly absent Kingston Peak Formation (Plate 1). In contrast, the structurally higher low-angle faults, as well as NE-striking, high-angle normal faults, are interpreted as products of NW-SE extension based on oppositely oriented stratigraphic cutoffs (Figs. 8 and 9). Low-angle normal faults are also present in the Saratoga Hills and have a top-to-the-NW sense of slip. Crosscutting relationships indicate that the NE-striking normal faults are younger than Faults A and B, which have a top-to-the-SW sense of slip.

Fault Kinematics

The northern Ibex Hills contain a few well-exposed fault surfaces that yield kinematic information. The fault planes were measured primarily from Faults B, C, and A (Fig. 5). The stereonet plot (Fig. 7) for measured low-angle normal faults forms a steeply dipping great circle indicating a general fault curvature about an axis plunging shallowly to the southwest (Fig. 7). This pattern is consistent with the general synformal map pattern of the northern Ibex Hills seen as both bedding dips and younger rocks surrounded by older rocks. Curiously, slickenlines for the same low-angle faults have rakes currently greater than 70° (Fig. 8) and scattered about a WNW-ESE axis (Fig. 7), which is inconsistent with the evidence that Faults A and B are top-to-the-SW normal faults. Given the fault curvature, however, it seems likely that the slickensided surfaces are recording an overprinting related to folding of the fault surface or younger top-to-the-NW faulting.

Cross-Section Observations

Cross sections (Fig. 10) support the interpretation that the structurally lower Faults A and B represent earlier top-to-the-southwest motion, whereas Fault C and the structurally higher normal faults were the result of later top-to-the-northwest movement. The projection of Fault B suggests that it crosscuts Fault A along the western flank of the northern Ibex Hills, suggesting it was a younger SW-directed extensional structure that cut Fault A (A–A′ of Fig. 10). In addition, the current fault plane orientations (Fig. 7) suggest either folding about a SW-NW axis or reflect an original corrugated fault geometry, and this curvature was incorporated into the projection of faults A–D at depth (Fig. 10). The interpretation of Fault C as a NW-directed structure was based on the field relationships and projections of the structurally lower faults in cross section (Fig. 5), which suggested the fault crosscuts the structurally lower faults to the NW. In addition, sections were refined through the process of reconstruction, and the classification of Fault C as a NW-directed structure fit well within our models.

Pre-Extensional Structures

Scattered exposures show mesoscopic folds and bedding-cleavage relationships indicative of older, Mesozoic structure, and one small exposure of a thrust system was recognized in the northern Ibex Hills (Plate 1). Nonetheless, in most of the mapped area, extensional structures are sufficiently complex that resolution of pre-extensional structure is difficult. An exception is a site in the Noonday Formation (Figs. 11 and 12) that shows a complex fold geometry. The outcrop lies within a small valley with highly folded upper Radcliff member beds bounded to the southwest and east by down-to-the-east normal faults (Fig. 5). The fault to the southwest cuts out lower Radcliff member, placing the folded upper beds against the Sentinel Peak member, which lacks the complex mesoscopic folding due to its massive bedding. To the northeast, the Mahogany Flats member of the Noonday lies in stratigraphic continuity with the Radcliff member (Fig. 5) but also lacks the conspicuous folding seen in the underlying Radcliff member.

Fleming and Pavlis (2016) described this outcrop in the context of 3D outcrop development for the site, but its significance to the regional structure was only briefly considered. The outcrop exposes a system of non-cylindrical folds with mesoscopic folds showing curved hinges and changes in fold axis orientation within ~1 m. The folds range from open to sub-isoclinal with tighter folds typically within the interbedded shales (Fig. 11). The main-phase folds are associated with a pressure-solution cleavage that is locally axial planar but is typically a divergent cleavage fan (Fig. 12). The folding is most conspicuous in the thinly bedded, shale-rich units, but the cleavage is present in both carbonates and slates. Small-scale, steeply dipping faults, with ~1–3 m of apparent offset, permeate the outcrop and complicate simple interpretation of bedding traces.

To better understand the geometry of these non-cylindrical folds, separate domains of the outcrop were analyzed, and their fold axes were determined from groupings of two or more planar measurements in each domain. These domains were determined based on perceived changes in fold orientation while in the field as well as analyzing the point cloud models. In addition to the field data, measurements from the point cloud were also plotted with the field data (Fig. 13).

The fold axes (Fig. 13A) and bedding-cleavage intersections (Fig. 13B) indicate F1 axes scattered around an ~E-W–striking, steeply dipping great circle. This scatter is consistent with the observed curved fold axes and indicates re-folding, but orientations of the second fold generation are unclear from Figure 13 alone. Cleavage orientations from both the outcrop as well as nearby exposures of the Radcliffe Member (Fig. 14) show significant scatter but are broadly consistent with a re-folding about a NE-trending axis. We suggest the scatter in Figure 13 reflects an original fanning cleavage developed along folds with an ~NS to NW trend; these folds were re-folded about NE-trending fold axes to produce both the curved fold axes and dispersion of bedding-cleavage intersections (Fig. 13).

Cross-Section Reconstruction

Saratoga Hills

To develop a reasonable model for the structural history of the Ibex Hills and the Saratoga Hills, estimates of fault displacement were reconstructed using cross sections made throughout the ranges (Fig. 10; Plate 1). Figures 1416 show the restorations, and Table 3 summarizes the total amounts of heave and stretch for sections A–A′, B–B′, D–D′, and F–F′.

Restoration of cross sections from the Saratoga Hills (sections D and F, Fig. 15) uses a simple approach of realigning contacts that were cut by the faults in question (Fig. 15). Using the move on fault module in Move, marker beds were selected on either side of the fault and restored to pre-faulting condition, assuming cross sections are parallel to the slip direction of the faults (Fig. 15).

The structural geometry in the Saratoga Hills is much simpler than that seen to the north in the Ibex Hills (Fig. 10). The most significant difference is the lack of curved, low-angle faults such as those seen in the northern Ibex Hills (Fig. 15B; Plate 1). There is, however, a low-angle fault in the Saratoga Hills that places the Kingston Peak Formation atop the Beck Springs Dolomite (Figs. 6 and 15). This low-angle fault has clearly been cut by the higher-angle normal faults and exhibits the same relationships seen to the north in the Ibex Hills but with less ambiguity on initial fault geometries (Figs. 6), simplifying the restoration process (Fig. 15B). Therefore, analyzing the restored section of the Saratoga Hills provides a template for the more complex structure of the northern Ibex (Fig. 15B).

The cumulative heave calculated from the restoration of cross section D–D′ in the Saratoga Hills is ~3 km (Fig. 15B). After reconstruction of the extension, a northwest-trending fold pair remains within the units and is interpreted to represent a pre- to syn-extensional fold in the Saratoga Hills. We interpret this fold system as a Mesozoic, pre-extensional structure because it predates both the low- and high-angle normal faults, and the pelitic layers in the Crystal Springs Formation are slates with bedding-cleavage relationships consistent with this fold geometry.

In addition to fault restoration, the effects of range tilt were also restored using the orientations of Neogene volcanic deposits, and Figure 15B shows the results of this tilt restoration. Following this rotation and after the restoration of the moderate- and high-angle normal faults, the restored heave is ~1.7 km. However, a low-angle fault surface remains with Kingston Peak Formation and younger rocks in its hanging wall (Fig. 15B; Table 3). Based on the projections of the restored section, the heave of this low-angle surface is 1.3 km.

Restoration of section F–F′ followed the same workflow as the other sections and shows a total heave of ~800 m and a throw of ~135 m (Fig. 15D; Table 3). Given that section F–F′ is oriented E-W and therefore oblique to the regional NW- and SW-directed transport directions, it is not surprising that it has the lowest amount of extension, although it does cross one significant normal fault (Fig. 15D; Table 3). The syncline of the fold pair recorded in cross section D–D′ also projects into F–F′ and constrains the orientation of the fold as trending northeast-southwest (Fig. 15).

Ibex Hills

Restoration in the northern Ibex Hills was more complex because our interpretation of two phases of extension, along different trends, indicates a three-dimensional process (Figs. 16 and 17). Figure 15 shows sequential restoration of section B–B′. The restoration first restores motion along the high-angle normal faults (Fig. 16B) revealing the earlier geometry of the low-angle normal faults. At this stage, however, range tilt has not been restored and low-angle normal faults retain their easterly dip (Fig. 16B) with apparent thrust motion but younger-on-older stratigraphic juxtaposition. Thus, the same tilt restoration as the Saratoga Hills is used. Note that the origin of this tilt is unconstrained in the mapped area but presumably represents a deeper detachment as envisioned by Serpa et al. (1988), relatively young folding across the range (e.g., Miller et al., 2007), or both. Nonetheless, the geometry of the Tertiary geology must be considered in the restoration.

Figure 16C also includes projected beds in the footwall of Fault C, although the orientations of these beds are not well constrained because they are currently covered by the Tertiary basin to the east of the Ibex Hills (Fig. 1; Plate 1). Therefore, the footwall geology assumes the simple interpretation of homoclinal bedding based on regional range tilting and not by any more complex structure. For example, none of the folding and faulting present in the Ibex Hills is included in the footwall interpretation.

The presence of faulted Johnnie Formation atop much of the northern Ibex Hills (Plate 1) presented a space problem when attempting to restore the unit along Fault D. To address this, two inferred faults were projected into the restoration at the point of Figure 16C. Given the history of late NW-SE extension in the region, this inference of NE-striking faults in the Fault D hanging wall is a reasonable way to accommodate the space problem.

Because Fault D is the youngest low-angle normal fault based on map pattern, the next step, after correction of range tilt, was to restore Fault D, with the inferred faults of its hanging wall (Fig. 16D). This results in an estimated horizontal displacement of ~1.4 km (Fig. 16D). Following this, Fault C (Fig. 16E) was restored with a heave of ~3.2 km. The total heave along Faults C and D is ~4.6 km (Fig. 16), and the total heave of the high-angle normal faults across the section was 550 m (Fig. 16B). Taken together, this restoration indicates 5.1 km of NW-SE extension for the northern Ibex Hills (Table 3).

The restoration of section A–A′ followed the same initial steps as that of B–B′ (Figs. 17A–17C). Fault E was also contained in A–A′ and was restored after restoration of the higher-angle normal faults (Fig. 17B) because it is interpreted as a top-to-the-NW structure. The hanging walls of Faults C and D were removed for the rest of the restoration because they were restored out of the line of section from the previous steps (i.e., Fig. 16). As in the A–A′ restoration, the footwalls and hanging walls of the faults were projected upwards using the known stratigraphy to constrain the reconstructions (Fig. 17C). That is, because younger faulting and erosion have excised part of this footwall, we assume a simple, continuous section was present prior to faulting—a reasonable assumption unless unrecognized thrust systems were present prior to extension. A key difference in the projected stratigraphy for A–A′, as opposed to B–B′, was the lack of the Kingston Peak and Beck Springs Formations, which are not present below the Noonday Formation in ranges north of the Ibex Hills (Wright et al., 1974). Restored movement along Faults A and B totaled ~2.4 km of heave (Fig. 17E). The effects of the high-angle fault restoration were minimal; therefore, the NE-SW extension of the range is primarily accommodated by the low-angle faults (Fig. 17) with a net horizontal displacement of ~2.4 km across the mapped area (Table 3).

The summary of total heave and throw along the faults in the northern Ibex Hills is provided in Table 3. In general, the majority of displacement is accommodated by the low-angle normal faults, in both the NW-SE and NE-SW extension directions. The total extension in the NW-SE (as shown by section B–B′) direction is more than twice that of the NE-SW directions, and ~10% of that displacement is along the younger, high-angle normal faults. In comparison, in the NE-SW direction (as shown by section A–A′), the displacement along high-angle normal faults was found to be negligible.

Overprinting Deformation in Southern Death Valley

Older Structures

The geology of the southern Death Valley region records a multi-phase deformational history that includes at least two phases of contraction and two phases of extension. The contractional history is best displayed at the focus site where we used a high-resolution 3D outcrop model to resolve fold geometry. In this area, early NW-trending (modern coordinates) folds record an intense, ductile deformational event associated with a conspicuous cleavage. Folds of this generation range from close to sub-isoclinal. These early fold systems were then overprinted by NE-trending open folds that produced curvature in the F1 fold axes as well as folding of the pressure-solution cleavage. Note that this contractional history is geometrically distinct from overprints recognized just to the east in the Resting Spring and Nopah Ranges, where Pavlis et al. (2014) documented early NE-trending structures associated with a low-grade cleavage overprinted by NW-trending fold and thrust systems. This suggests either the Ibex Hills area experienced a distinctly different kinematic history than the area to the east, or the Ibex Hills have experienced a rigid body, vertical axis rotation of ~90°—a problem that is discussed further below.

Evidence for SW-NE Extension

The earliest extensional phase of the region is recorded in the Ibex Hills, Ibex Pass basin, and northern Saddlepeak Hills. In the Ibex Hills, we suggest that NE-SW extension is recorded as top-to-the-SW movement along Faults A and B (Figs. 5 and 9; Table 2; Plate 1). Map observations support this hypothesis because the traces of both Faults A and B cut upsection to the northeast (Fig. 9; Plate 1). Cross-section construction and restoration support this conclusion with movement along Faults A and B leading to an estimated stretch of ~1.54 in the range (Table 3).

Evidence for the timing of this extension, as well as direct connection to previously recognized NE-SW extension, is illustrated by the structure of the Ibex Pass basin and northern flank of the Saddlepeak Hills. In the Ibex Pass area, the volcanic rocks have been interpreted as part of the 14–12 Ma volcanic assemblage based on correlations to the Wingate Wash assemblage (e.g., Luckow et al., 2005; Canalda, 2009) and one unpublished K-Ar date (Calzia and Rämö, 2005). In the Ibex Pass area, these volcanics dip NE, consistent with tilting during a SW-directed extensional event, but are overlain along a nearly flat-lying angular unconformity with distinctive megabreccias that Topping (1993) correlated to rock-avalanche deposits now exposed to the north, in the Amargosa Chaos. Topping (1993) showed that these granitic megabreccias were sourced from the ca. 14 Ma Kingston Peak pluton and were deposited between ca. 10–8 Ma. Taken together, the data suggest that the majority of the volcanic deposits of the Ibex Hills area were deposited between ca. 14–12 Ma with the angular unconformity developed between 12 and 10 Ma. Thus, deformation had begun with some tilting prior to the development of the unconformity. Similarly, in the northern Saddlepeak Hills, mutual crosscutting relationships among EW-trending sinistral faults and the igneous succession as well as truncation of earlier low-angle normal faults by that fault array (Fig. 4) place the top-to-the-SW faults within the time windows of the 14–12 Ma igneous event. Finally, in parts of the Ibex Hills basin (lower insert, Fig. 4), a NW-striking, east-dipping fault is clearly cut by NE-striking faults associated with younger, NW-SE–directed extension. However, this NW-striking fault and another fault to the SE also cut the angular unconformity, indicating the NE-SW extension continued after development of the unconformity with the basal megabreccias (Table 2). This timing is consistent with previous studies that place the transition to the younger, NW-SE extension in the 12–10 Ma time interval (Fridrich and Thompson, 2011).

The presence of potential NE-SW extension in the Saratoga Hills is ambiguous, but an older, currently low-angle fault surface is exposed in the northern Saratoga Hills. That fault is cut by top-to-the-NW normal faults (Figs. 6 and 15). Thus, based on similar crosscutting relationships to the north, this low-angle fault is probably part of older, NE-SW extension. While it is possible that these low-angle faults in the Saratoga Hills are coeval with those seen in the Ibex Hills, we believe that is unlikely given the lack of exhumation in the footwall between the two ranges. That is to say, if the low-angle faults of the northern Ibex Hills extended south, we would expect the presence of deeper structural levels in the footwalls of the Saratoga Hills; however, such a presence is not observed. Instead, we infer that the southern and northern Ibex Hills represent two distinct fault systems.

Importantly, although a period of NE-SW extension has long been known to have occurred in the Saddlepeak Hills (Troxel et al., 1992), no previous mapping has documented any related structures farther north into the Ibex Hills. Recognizing these older extensional structures in the Ibex Hills is important because previous models have either ignored its geology or included the structures as part of the Black Mountain detachment fault (Fig. 2) (e.g., Holm and Wernicke, 1990; Snow and Wernicke, 2000). We note, however, that this conclusion has further implications for strike-slip offsets in the system. Studies of the Kingston Range detachment (e.g., Davis et al., 1993) have generally only projected the detachment a few km north of the Kingston Range (Fig. 1). Projecting that northern limit to the SW is consistent with observed extensional structures in the Saddlepeak Hills (Fig. 1) but not the Ibex Hills. However, our work in the Ibex Hills indicates the presence of top-to-the-SW faulting occurred ~15 km to the northwest of the Saddlepeak Hills. In order to explain this, either the northern edge of this extensional belt has been offset ~15 km, or NE-SW extension was present farther north into the Nopah and Resting Spring Ranges and later displaced to the SW (Figs. 1 and 2). This potential offset of ~15 km can at least partly be explained by our estimates of ~5 km of heave along the younger NE-striking faults (Table 3). The remainder of the offset is probably due to dextral slip on the Grandview fault (Fig. 1), consistent with Topping's (1993) hypothesis based on other evidence. Thus, future studies need to consider this more complex kinematic history.

Overprinted Normal Faults

Crosscutting Faults A and B in the northern Ibex Hills are Faults C and D, interpreted here as top-to-the-NW low-angle normal faults (Fig. 16; Table 2). These faults record ~4.6 km of heave to the NW and place the upper Noonday Formation and younger rocks against crystalline basement (Plate 1). More recent top-to-the-NW structures are also present in the Ibex Hills as well as farther south in the Saratoga Hills (Figs. 15 and 16; Plate 1). This extension is represented by moderate- to high-angle normal faults that clearly crosscut the currently low-angle Faults A–E (Table 2). The presence of these structures is important because it does indicate this region was overprinted by top-to-the-NW structures, supporting the hypothesis that this region also contains the up-dip equivalents of the Amargosa–Black Mountain fault system, or similar structures.

The most recent faulting in the map area occurred along the Eastern Boundary fault (Fig. 5; Table 2; Plate 1). The presence of the Eastern Boundary fault is interpreted here as related to the most recent phase of transtension in the Death Valley region because it crosscuts earlier normal faulting in the Ibex Hills (Fig. 5). Although the trace of the Eastern Boundary fault is obscured in the northern portion of the study area, the fault displaces the Johnnie and Stirling Formations ~4 km to the southeast relative to the footwall positions of these rocks (Plate 1). In addition, the gravels that overlie the Tertiary sections of the basin, and are interbedded with the granitic megabreccias (Fig. 4), presumably continue to the north and lap onto the flank of the Ibex Hills and contain clasts of the Noonday and Beck Springs megabreccias (Plate 1). Given that the megabreccias of the Noonday and Beck Springs Formations are presumed to be coeval with the latest phases of extension and transtension in the range, the interbedded upper gravels and megabreccias of the Tertiary Ibex Pass basin are also coeval with those events. Active extension and transtension continue to the west of the Ibex Hills, but faulting in the study area ended by the Pliocene (Table 2). Remaining activity in the area is now dominated by contraction and transpression (Menges et al., 2005).

Implications for Extensional Models of the Region

Previous workers have associated the faulting in the northern Ibex Hills with the Black Mountain detachment fault (e.g., Holm and Wernicke, 1990; Snow and Wernicke, 2000), and one of the goals of this study was to test this hypothesis. Based on our interpretations, the data are not consistent with a single detachment surface and instead indicate a more complex history with at least two extensional systems superimposed. However, one variant on the detachment model that is allowable from the data is that the Ibex Hills, and specifically Faults C and above, represent a piece of the extensional detachment system that was close to, or at, the breakaway for the Black Mountains detachment. That is, motion on the fault at the structural level of Fault C and higher is limited to ~5 km, which could be because these rocks were left behind as deformation progressed to the northwest. Moreover, the presence of the thick megabreccias derived from Kingston Range (unit Tmg) and the presence of Kingston Range granite in younger gravels (Tg2) (see Supplemental Material [footnote 1] for their full description) support some original continuity between the Amargosa Chaos and the Ibex Pass basin, prior to the potential NW-directed movement along the detachment. Ultimately, while the Ibex Hills have no doubt been transported along other regional structures (for example, the Sheephead fault), Grandview fault, some portion of the Black Mountain–Amaragosa fault system, and/or faults within the Ibex Pass area (Canalda, 2009), a single structure cannot account for the multi-phase deformation recorded in the range.

In addition to enhancing our understanding of the local Death Valley geology, refining models, such as those mentioned in the previous paragraph, also improves our broader understanding of continental extension. The Basin and Range Province is one of the most well studied areas of continental extension on Earth (Cemen et al., 2002), and Death Valley provides insight into many extensional processes (Miller and Pavlis, 2005; Hussein et al., 2007; Lima et al., 2018). To that point, recent models that seek to understand strain partitioning in extensional regions rely on a number of data sets, including piercing lines and Cenozoic basin history (Lutz et al., 2019, 2020). While our work here represents only a small piece in a larger puzzle, it nonetheless lies in a key position and will provide an updated interpretation of one of southern Death Valley's extensional basins.

Genesis of Curved Fault Surfaces in the Ibex Hills

At large scale, the northern Ibex Hills is a broad synformal structure as indicated by the map traces of units and faults (Plate 1). Stereonet analysis of low-angle fault orientations in the area supports this map-scale observation and suggests a curvature of the fault surfaces about a northeast-southwest–directed axis equivalent to the orientation of the synform axis (Fig. 7). While some of this curvature is likely due to different initial geometries of the various low-angle normal faults mapped in the area (e.g., Fault A vs. Fault D), the curved geometry is observed at all structural levels. Thus, because the faults are all Neogene extensional structures, both the synform and the fault systems are presumed to be syn-extensional products. The phenomenon of fault curvature is common in extensional terranes and can originate as primary fault curvature and/or corrugations with axes parallel to extension (e.g., Fowler and Calzia, 1999), as long-wavelength folds with axes perpendicular to extension related to isostatic rise of a detachment footwall (Wernicke and Axen, 1988), a result of folding of fault planes in transtension (e.g., Mancktelow and Pavlis, 1994; Serpa and Pavlis, 1996), or some combination of these processes.

Primary corrugations parallel to extension have been proposed as a common manifestation of large-scale extension (e.g., Spencer, 2000; Spencer et al., 2019) and have been well documented within the Kingston Range detachment system just to the south of the study area (McMackin, 1992; Fowler and Calzia, 1999). Thus, the primary corrugation hypothesis is seemingly allowable; yet, we argue here it is highly unlikely for the Ibex Hills. The problem with this explanation is that while the curvature observed in the Ibex Hills is oriented along a northeast-southwest axis, consistent with corrugations of top-to-the-SW extension, it is difficult to envision a process that would preserve this geometry given the intensity of the extension that clearly overprints the older structures. For example, low-angle faults C–E (Fig. 5) all show cutoff relationships suggestive of an origin as top-to-the-NW normal faults; yet they share the curvature as underlying faults and bedding—an observation difficult to rationalize as a primary corrugation.

Alternatively, syn-extensional folding may have occurred via an isostatic response to northwest extension in the Ibex Hills. The suggestion of large isostatic responses in the region is a key component of the rolling-hinge model (Buck et al., 1988; Wernicke and Axen, 1988; Snow and Wernicke, 2000) since the deformation front propagates forming a long-wavelength fold in its wake. Combined with the fact that previous work has included the Ibex Hills as part of the up-dip portion of the Black Mountain detachment fault (e.g., Holm and Wernicke, 1990), the rolling-hinge model provides an attractive explanation. Nonetheless, although this process is well documented at the large scale, in regional detachments, the wavelength of the folding seen in the Ibex Hills is only 2–3 km, which is inconsistent with most models of isostatic flexure (Buck et al., 1988; Wernicke and Axen, 1988).

Given these complications, we suggest that the most likely scenario is that the curved fault surfaces were folded after their initial movement. This folding would have to have occurred during, or after, the development of the NE-striking normal fault systems but prior to the Quaternary when the range became part of a large, NW-trending anticline (Menges et al., 2005; Miller et al., 2007). The young folding about a NE-trending axis is broadly coaxial with the re-folding of older F1 fold axes within the Noonday Formation (Fig. 7), suggesting correlation; yet, the small-scale folds are very different in style than the map-scale structure. That is, the regional structure is a broad, relatively open synform with a wavelength of ~4 km, whereas the structures in the Noonday Formation are tight, mesoscopic structures producing complex interference patterns in outcrop. Thus, although they may be related, it is much more likely that they are structures of different age and are coincidentally coaxial.

The driver for the young, NE-trending folding is not immediately obvious. Given its relative age, it seemingly is related to the modern, transtensional kinematics in the Death Valley region (e.g., Topping, 1993; Serpa and Pavlis, 1996; Norton, 2011). Thus, some form of transcurrent-motion–related folding seems likely. Nearby strike-slip systems that are relevant to this hypothesis include: (1) the NW-trending dextral southern Death Valley system to the west; (2) the dextral-normal eastern boundary fault of the Ibex Hills described above; and (3) the enigmatic Sheephead dextral (or sinistral) fault to the north. The sense of offset along the Sheephead fault has been debated with some authors supporting a right-lateral fault (e.g., Renik and Christie-Blick, 2013), while others suggest a left-lateral sense of motion (e.g., Topping, 1993; Serpa and Pavlis, 1996) along the fault.

One potential model for the folding of the Ibex Hills is shown in Figure 18 and relies on the assumption of a left-lateral Sheephead fault. This interpretation of the Sheephead fault, when paired with the right-lateral EBF of the Ibex Hills, would generate compression in and around the northern Ibex Hills (Fig. 18). Similar cases of crustal shortening have been documented to the south at the intersection of the Garlock fault and the southern Death Valley fault zone (Brady, 1984; Serpa and Pavlis, 1997). This model also hinges upon timing of fault systems, with a requirement that the Sheephead fault and the EBF are contemporaneous.

Alternatively, if the Sheephead fault is a dextral fault, the model of Mancktelow and Pavlis (1994) potentially is applicable to the Ibex Hills (Fig. 1). That is, a NE-trending fold could result from wrench folding along an EW-trending dextral fault. Given the evidence for older, top-to-the-SW faulting in the Ibex Hills, this model would provide another explanation for the WNW kinematics shown in Figure 14D, in addition to the effects of regional vertical axis rotation (e.g., Serpa and Pavlis, 1996).

Clearly distinguishing dextral versus sinistral slip on the Sheephead fault is critical to evaluating this local tectonic problem. In addition, however, absolute timing may be critical. It is clear the folding is relatively young in the Ibex Hills with good evidence folding occurred after the low-angle faulting, but before the most recent, high-angle normal faulting (e.g., Fig. 8). Nonetheless, although they lie within the same relative time window, it is not clear the folding is contemporaneous with slip on the Sheephead, EBF, or both. Quaternary, syn- to post-extensional, folding in the southern Death Valley region has been recognized in the Tecopa Basin to the east (Miller et al., 2007), and strikingly similar geometries of folded low-angle normal faults have been mapped in the Amargosa Chaos area to the northwest (Castonguay, 2013). However, in both cases, the fold axes are trending to the northwest, nearly 90° to the synform of the Ibex Hills. The variance in the geometry of young fold axes in the southern Death Valley region may be explained by vertical axis rotation related to transtension, which may differ between these areas. Nonetheless, this rotation is unconstrained in many places (e.g., Serpa and Pavlis, 1996; Guest et al., 2003).

Collectively, these relationships indicate the need for further clarifications of both timing and vertical axis rotations in this region. Vertical axis rotations could be easily tested in the Tertiary volcanics of the Ibex Pass basin, and timing could be further clarified by dating those rocks. Similarly, thermochronology could help resolve timing issues by clarification of footwall exhumation, and studies in progress may provide some clarification. Nonetheless, more work is clearly needed on both of these issues before we can fully understand the significance of the fold system in the northern Ibex Hills.

The northern Ibex Hills expose a structurally complex sequence of normal faults, many of which are currently low-angle, placing units as young as the Johnnie Formation against Precambrian basement rock (Fig. 4; Plate 1). The low-angle normal faults appear to be folded about a NE-SW axis, consistent with the broader map pattern of the area. The fact that the low-angle faults are folded, but the younger normal faults are not, supports a folding event younger than the low-angle normal faulting but which preceded the younger faulting in the Ibex Hills.

The low-angle normal faults are cut by other normal faults that exhibit down-to-the-NW motion in their hanging walls (Figs. 4 and 8). In addition, there is compelling evidence for top-to-the-SW movement on the low-angle normal faults in the northern Ibex Hills. This SW displacement is also consistent with the offsets seen in the Ibex Pass area and the northern Saddlepeak Hills, where NW-striking, low-angle faults cut a ca. 12–10 Ma unconformity. The evidence for top-to-the-SW movement along faults in the Ibex Hills supports our assertion of a previously unrecorded multi-phase extensional history—that is, that before ca. 10 Ma, the deformation in the Ibex Hills and Ibex Pass area was dominated by NE-SW extension. This was then followed by the NW-SE extension recorded throughout the region (Table 2) (e.g., Holm et al., 1992). The multi-phase deformation of the Ibex Hills also highlights the transition from simple extension to transtension that occurred in the southern Death Valley region. We propose that the folding of the low-angle faults is related to this shift to transcurrent deformation, perhaps combined with regional block rotation, most likely tied to the nearby Sheephead fault (Fig. 18). Such polyphase deformation as seen in the Ibex Hills is important to consider for future study of the region since previous work has primarily emphasized the NW-SE extension (e.g., Holm et al., 1994) and failed to recognize the presence of NE-SW extensional structures west of the Saddle Peak Hills (Topping, 1993).

The complexity of the northern Ibex Hills is not seen farther south in the study area where the structure is dominated by steeply dipping normal and oblique faulting (Plate 1). Down-to-the-NW normal faults typical of the southern Death Valley region are well exposed in the boundary between the Saratoga Hills and the southern Ibex Hills (Plate 1). Younger, generally right-lateral to oblique faults strike to the NW and are interpreted to crosscut some of the normal faults in this area (Plate 1). These high-angle, oblique to strike-slip faults are most obvious in the Saratoga Hills, where they are clearly exposed and are the dominant structure, especially in the central and southern portions of the range (Plate 1).

Future work in the area of the Ibex Hills and Saratoga Hills could help to resolve some of the ambiguities in the observations presented herein. Most pressing might be to time the initiation of extension in the Ibex Hills because this will provide new constraints on the period of SW movement in the region. Also, dating the volcanics at the southern Ibex Hills, along with those of Ibex Pass, would be a big step in our understanding of the basin evolution and syn-extensional volcanism of southern Death Valley. In doing so, we may be able to shed light on the kinematics of regions that transition from continental extension to transtension. For example, if the oldest low-angle normal faults of the Ibex Hills are in fact related to the Kingston Range detachment, this would indicate that the range was displaced significantly along some combination of the Grandview and Sheephead faults. Such a conclusion would further highlight the importance of strike-slip faulting in the evolution of extensional provinces.

This work was supported by National Science Foundation grant EAR-1250388 to Pavlis and a Geological Society of America student grant to Fleming. We would like to thank Midland Valley Ltd., Agisoft, Inc., for software donations that made this work possible. In addition, Jim Rukofske was a vital part of completing the field work necessary for this study. We would also like to thank the two anonymous reviewers whose comments helped to improve this manuscript.

Present address: Occidental College, Department of Geology, 1600 Campus Road, Los Angeles, California 90041, USA
Science Editor: Shanaka de Silva
1 Supplemental Material. Description of map units of Ibex Hills and Saratoga Hills, as well as adjacent Ibex Pass basin. Please visit to access the supplemental material, and contact with any questions.
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.