Abstract

The Rio Grande rift in central New Mexico provides an excellent location to study the interaction between high-angle and low-angle (15°–35°) normal faults during crustal extension. Here we evaluate the relative importance of low-angle normal faults (LANFs) in the Albuquerque basin of central New Mexico with goals of testing two conflicting models of rift geometry and producing evolutionary models for the northern and southern parts of the basin. Using physiographic relationships, field observations, structural data analysis, and thermal history modeling, we document two brittle LANF systems on salients in adjacent opposite-polarity half-grabens. These fault systems were both active ca. 20–10 Ma and are locations of maximum fault slip as indicated by thickness of sedimentary fill in adjacent sub-basins and highest elevation rift flanks. Average fault dip increases basinward, and outbound faults were abandoned while intrabasinal faults cut Quaternary units, supporting an evolutionary model where master normal faults initiated at a higher dip, were shallowed by isostatic footwall uplift in regions of highest slip, and became inactive while younger normal faults emerged basinward. These geometrical and kinematic observations are predicted by the rolling-hinge model for the formation of LANFs. This mechanism has been widely applied to core complexes in highly extended terranes (e.g., Basin and Range), regions of orogenic collapse, and mid-ocean ridges, and it is shown here to also be applicable to narrow continental rifts of modest (∼35%) extension. Similarities to core complexes include a physiographic expression of domal uplifts, evolution of a master detachment horizon that initiated as a breakaway, and isostatically rotated low-angle normal faults. Although the degree of extension was too low to juxtapose ductile footwall rocks against brittle hanging-wall rocks, if extension had progressed in the Albuquerque basin, eventually a mature metamorphic core complex would have formed, similar to those preserved in the adjacent Basin and Range Province. The Rio Grande rift, therefore, provides a snapshot of the embryonic stages of core complex formation, bridging the gap between mature core complexes and incipient extensional environments.

INTRODUCTION

The Rio Grande rift, extending more than 1000 km from Colorado to Mexico, is one of the world’s premier and best studied continental rift systems. Although research on understanding the geology of the rift has been ongoing for decades, many questions still remain regarding the timing and style of its structural development. Continued interest in the rift is driven not only by the need to understand continental extensional processes but also because of a need to understand and characterize potential seismic hazards (e.g., Wong et al., 2004), the valuable water resources of the basin-fill aquifers in Colorado and New Mexico (e.g., Bartolino and Cole, 2002; Plummer et al., 2004; Johnson et al., 2013), and potential oil and gas production (Black, 2013). These issues are especially important in the Albuquerque metropolitan area, the fastest growing region in New Mexico. To address important earthquake, oil and gas, and water-related issues in New Mexico, it is necessary to have a firm understanding of young deformation (e.g., Berglund et al., 2012; Ricketts et al., 2014a) as well as the longer term evolutionary history of rift faults and basin geometry. Continuing efforts to understand these issues related to the Rio Grande rift have led to the development of two Geological Society of America (GSA) Special Papers: GSA Special Paper 291, edited by Keller and Cather (1994), and GSA Special Paper 494, edited by Hudson and Grauch (2013).

One aspect of rift formation that is relevant to these issues is the relative importance of high-angle versus low-angle normal faults (LANFs) in accommodating crustal extension and how fault geometries may vary as extension progresses. Low-angle normal faults (commonly referred to as “detachment faults”) in the Rio Grande rift have been recognized for decades (e.g., Black, 1964), where they typically exist as isolated faults or fault fragments within a predominantly high-angle normal fault environment (Fig. 1) (Baldridge et al., 1984). However, because of the relative paucity of LANFs in the rift compared to their high-angle counterparts, recent models of basin formation in the central Rio Grande rift tend to emphasize high-angle normal faults to accommodate crustal extension (e.g., Connell, 2008; Grauch and Connell, 2013), and models including LANFs are generally lacking. This is in contrast to LANFs in the adjacent Basin and Range Province, where large-magnitude Eocene–late Miocene crustal extension in western North America formed a continuous belt of detachment faults and metamorphic core complexes that extends from British Columbia, Canada, to Sonora, Mexico (Fig. 1) (e.g., Coney, 1980; Axen et al., 1993; Dickinson, 2002, 2009). The timing and style of extension to form metamorphic core complexes in the Basin and Range has been extensively studied, and these structures are typically thought to have formed via a rolling-hinge mechanism, where isostatic rebound of footwall rocks causes initially high-angle normal faults to rotate to shallower angles (e.g., Spencer, 1984; Buck, 1988; Hamilton, 1988; Wernicke and Axen, 1988; Axen and Bartley, 1997).

Within this general setting, this paper examines and evaluates the importance of several exposed LANFs in the central Rio Grande rift to test the hypothesis that they formed through a similar rolling-hinge mechanism as LANFs that are part of metamorphic core complexes in the Basin and Range Province. This study aims to help answer the question of the overall importance of exposed LANFs in the modest-extension (∼17%–35%; Russell and Snelson, 1994) Albuquerque basin of the Rio Grande rift, central New Mexico. We apply a combination of physiographic observations, fault orientations, kinematic studies, and low-temperature thermal history modeling to understand the evolution of some long-recognized (Black, 1964), but poorly understood (e.g., Baldridge et al., 1984), LANF segments. We examine whether these LANFs initiated and moved at low or high angle and what can be learned in this moderate extension, narrow, continental rift system to inform models for rift basin segmentation, rift geometry, and early stages of core complex development in extensional domains. On a scale of the Rio Grande rift, we explore the importance of fault rotation to form LANFs, as opposed to models involving only high-angle faults (e.g., Grauch and Connell, 2013), to explain rift-flank geometries and strain compatibility between upper-crustal and deeper-crustal extension. Within this context, and with the ultimate goal of producing a robust fault evolutionary model for the Albuquerque basin, below we first briefly discuss extensional fault geometries and then summarize observations and constraints on fault geometry and fault age in the Albuquerque basin, central New Mexico.

BACKGROUND ON EXTENSIONAL FAULT GEOMETRIES

The expression of extension in both oceanic and continental settings appears to be strongly dependent on the geotherm and the ability of the deep crust to flow (Block and Royden, 1990; Brun and van den Driessche, 1994; Lavier et al., 2000; Rey et al., 2009a, 2009b; Whitney et al., 2013). Heat flow dictates, to a large extent, the degree to which the brittle upper crust and lower ductile crust are mechanically coupled. In regions of cold to average continental heat flow, there is a strong mechanical coupling between the brittle upper crust and the lower ductile crust, and two-dimensional models predict that the lower ductile crust will thin homogeneously without responding to gravitational stresses (Whitney et al., 2013). As the lower crust thins, numerous normal faults develop in the brittle upper crust to produce graben and half-graben geometry, as well as domino-style block rotation (Fig. 2A) (e.g., Proffett, 1977; Chamberlin, 1983; Nur et al., 1986; Stewart, 1998; Brady et al., 2000). As extension progresses, the individual domino blocks and normal faults are systematically rotated, and faults become progressively less suited for continued slip. Eventually a second generation of faults will form and crosscut the first generation of faults. Based on mechanical considerations, the amount of permissible fault rotation in a stationary stress field is ∼20°–45° (based on rock strength and depth in the crust) before a new set of faults must form (Fig. 2A) (Nur et al., 1986).

In relatively hot crust, the brittle upper crust is mechanically decoupled from the lower ductile crust (e.g., Brun et al., 1994; Whitney et al., 2013). Under these conditions, the crust extends via a rolling-hinge mechanism, where strain becomes localized along a single, large-offset normal fault system, which often initiates at a high angle and rotates to shallower dips due to isostatic rebound (Spencer, 1984; Buck, 1988; Wernicke and Axen, 1988; Wdowinski and Axen, 1992; Axen and Bartley, 1997). New high-angle faults then emerge basinward, and if this process continues, eventually lower-crustal, ductilely deformed rocks can be exhumed to the surface in the footwall of the LANF to form a core complex (Fig. 2B). In this model, normal faults young toward the axis of the basin. The Basin and Range Province (e.g., Coney and Harms, 1984; Buck, 1991; Axen et al., 1993), the Aegean Sea (e.g., Lister et al., 1984; Keay et al., 2001), and Tibet (e.g., Kapp et al., 2008) are several geologic examples of a rolling-hinge mechanism operating in hot crust.

Low-angle normal faults are observed in sandbox experiments (e.g., McClay and Ellis, 1987; Brun et al., 1994; Mourgues and Cobbold, 2006), are predicted in numerical models (e.g., Buck, 1988; Rosenbaum et al., 2005; Tirel et al., 2008), and are a near-ubiquitous feature of extensional domains. However, LANFs in modest-extension, narrow (<100-km-wide), upper-crustal continental rifts are not as well documented. This paper takes the view that understanding faults in these settings may provide clues about the early development of LANFs and core complexes in general.

GEOLOGIC SETTING OF THE ALBUQUERQUE BASIN

The north-south–trending Rio Grande rift is a classic, well-exposed, and well-studied example of a narrow and still-active continental rift system (e.g., Baldridge et al., 1984; Chapin and Cather, 1994; Russell and Snelson, 1994; Berglund et al., 2012; Ricketts et al., 2014a). Early extension began ca. 36–37 Ma (Kelley and Chamberlin, 2012) and resulted in modest magnitude extension that likely began the process of separating the relatively undeformed Colorado Plateau to the west from the Great Plain province to the east. However, the modern physiography and N-S extent of the rift are the results of mainly Miocene extension, which was the time of accumulation of most of the E-W extensional strain (e.g., Kelley et al., 1992; Chapin and Cather, 1994). Normal fault geometries were influenced to some degree by reactivated earlier faults of Precambrian, Ancestral Rockies, and Laramide age (e.g., Karlstrom et al., 1999; Marshak et al., 2000). The rift is composed of a series of NS-trending basins that form an array of en echelon grabens and half-grabens that alternate polarity and are linked by NE-trending accommodation zones (e.g., Chapin and Cather, 1994) that serve to transfer the maximum displacement from one rift margin to the other (e.g., Russell and Snelson, 1994; Faulds and Varga, 1998; Minor et al., 2013). Within the Rio Grande rift, isolated exposures of LANFs have been known for decades (e.g., Baldridge et al., 1984), although they exist within a predominantly high-angle normal fault environment, and models for their formation are largely lacking.

The Albuquerque basin, located in central New Mexico, is one of the largest basins of the Rio Grande rift. Structurally it is bounded on all sides by normal fault systems, and growth of the basin has resulted in the development of multiple sub-basins that are filled with the synextensional sedimentary basin fill of the Santa Fe Group (Fig. 3). The northern (Albuquerque) sub-basin is an east-tilted half-graben, where Santa Fe Group sediments dip toward west-dipping normal fault systems (the high-angle Rincon fault and low-angle Knife Edge detachment fault) along the Sandia uplift (>3 km elevation; Fig. 3). The southern (Belen) sub-basin encompasses multiple half-grabens that are filled with gently dipping Santa Fe Group sediments. Steeper southwest-dipping beds of the Santa Fe Group are restricted to the southwestern part of the basin (Grauch and Connell, 2013) and dip into an east-dipping, low-angle normal fault system (Jeter detachment fault) that bounds the Ladron rift flank (2.7 km elevation; Fig. 3).

Several additional LANFs are located within the Albuquerque basin, including the Carrizo fault along the center of the western margin of the Albuquerque basin (Callender and Zilinski, 1976) and LANFs along the southeastern edge of the basin (Beck and Chapin, 1994; DeMoor et al., 2005) (Fig. 3). These LANFs, however, differ from LANFs in the Sandia and Ladron uplifts because the total offset along these structures is several hundred meters at most. Thus these faults may have formed through alternative processes that were not operative on a basin-wide scale, and they do not seem to play an important role in overall basin evolution.

One of the goals of this paper is to evaluate two conflicting models for rift geometry and evolution of the Albuquerque basin of central New Mexico. Russell and Snelson (1994) used seismic reflection and borehole data to develop one of the most widely cited models of the subsurface geometry of the Albuquerque basin, which suggested that the Albuquerque basin is composed of two individual half-grabens that alternate polarity. In their subsurface interpretations, surface faults are listric and sole into low-angle faults at shallow (5–10 km) crustal depths. Alternatively, Grauch and Connell (2013) utilized diverse geophysical data sets from the Albuquerque basin to construct a model of the subsurface that indicates a predominance of extension along high-angle faults and a more complex anticlinal accommodation zone between opposite polarity basin segments. The differences in interpretations are depicted in Figure 4, which shows ∼EW cross sections constructed across the northern Albuquerque basin by Russell and Snelson (1994) and Grauch and Connell (2013) that highlight their different interpretations of basin depth and fault geometry. The most prominent difference between these two interpretations is the importance of low-angle normal faults (LANFs) in the Russell-Snelson model and the lack of such features in the Grauch-Connell model. Assuming no movement in or out of the sections and rigid body rotations and translations, palinspastic reconstructions of these two cross sections along the Great Unconformity at the top of Precambrian rocks suggest that LANFs in the subsurface provide a better way to balance upper-crustal brittle extension with middle-crust ductile extension across shear zones, an observation that has been supported by numerous authors (e.g., McKenzie, 1978; Wernicke, 1985). The Russell-Snelson model is also supported by seismic reflection data, which image near-horizontal detachments at depth (Russell and Snelson, 1990, 1994). In addition, the Grauch and Connell (2013) cross section does not explicitly address strain compatibility issues but presumably would rely on pervasive, pure shear-dominated middle-crustal flow (McKenzie, 1978). Low-angle normal faults exposed at the surface in the Albuquerque basin (Fig. 3) also support the likelihood they are present at depth as well and contribute to the overall extension.

Southern Albuquerque Basin and the Ladron Uplift

The Ladron uplift is a high-elevation promontory that protrudes east along the southwestern edge of the Albuquerque basin, at the junction between a relatively narrow Rio Grande rift to the north and more distributed extension to the south that is similar in style to the Basin and Range Province. This uplift is bounded on its western edge by the high-angle, reverse-sense Ladron fault, which separates Precambrian granitic and metamorphic rocks in the footwall from Paleozoic sediments in the hanging wall. The rift-bounding structure along the eastern margin of the Ladron uplift is the Jeter detachment fault, the best known and most well studied LANF in the Albuquerque basin (e.g., Black, 1964; Lewis and Baldridge, 1994; Read et al., 2007). This structure dips ∼15°–30° east and contains Precambrian granitic and metamorphic rocks in the footwall (Fig. 5). The hanging wall adjacent to the Jeter detachment fault is an intensely brecciated zone containing slivers of Paleozoic and Mesozoic rocks, Neogene volcanic and volcaniclastic rocks, and Santa Fe Group rift fill. Santa Fe Group sediments dip ∼25°–35° west toward the fault plane (Read et al., 2007). The fault consists of a 3- to 4.5-m-thick brecciated zone that is composed of fine red clay surrounding intensely fractured and pulverized angular Precambrian clasts (Black, 1964). Fault plane measurements along the Jeter detachment fault define a shallowly dipping normal fault system, and slickenline orientations trend toward the southeast (Fig. 5). In the hanging wall of the Jeter detachment fault, the Silver Creek detachment fault dips ∼25°–35° east and is inferred to sole into the Jeter master detachment at depth. The Silver Creek detachment largely separates Paleozoic and Mesozoic rocks in the footwall from Santa Fe Group rift-fill sediments in the hanging wall (Read et al., 2007).

Northern Albuquerque Basin and the Sandia Uplift

The Sandia uplift is the highest elevation part of the eastern rift flanks and is located along the northeastern edge of the Albuquerque basin. It is composed of Precambrian granitic and metamorphic rocks capped by east-dipping Paleozoic sedimentary rocks (Fig. 3C). The rift-bounding structures on the west side of the Sandia uplift include the high-angle Rincon fault and disconnected LANFs at three separate locations (1, 2, and 3 in Fig. 3C). Location 1, at the base of the Sandia uplift in the Juan Tabo area, preserves a small klippe of ∼30° east-dipping Mesozoic rocks that are structurally underlain by Precambrian granite (Fig. 6). Although the fault is mostly concealed, magnetic surveys plus the geometry of the contact suggest this is a west-dipping, low-angle fault contact (Van Hart, 1999). We refer to this low-angle fault as the Juan Tabo detachment fault. At location 2, Read et al. (1999) mapped a low-angle fault zone that cuts Precambrian granitic rocks. They refer to this structure as the Knife Edge fault. New mapping, along with fault plane and slickenline measurements along this zone, suggest it dips ∼25°–35° west with top-to-the-west slip (Fig. 6). This fault is aligned with the ridge of the bedrock facet above called the Shield, suggesting that the “Knife Edge of the Shield” is an exhumed upward continuation of this fault zone as shown in Figure 6A. Riedel shears and slickenlines indicate normal-sense movement with top-to-the-west movement direction (Fig. 6B). Farther north, the Ranchos detachment fault (location 3 in Fig. 3C) separates 15° east- dipping Santa Fe Group sediments from Mesozoic footwall rocks along a normal fault dipping ∼30° (Fig. 6D; Kelley, 1975; May et al., 1994). Truncation of hanging wall strata and a ∼6-cm-thick gouge zone show this is not an unconformable contact. At all locations, deformation associated with these structures is entirely brittle.

Central Albuquerque Basin Accommodation Zone

Accommodation zones in the Rio Grande rift vary in width from >10 km to narrower transfer zones that are <3 km wide. They variably consist of single faults, multiple fault zones, single folds, multiple folds, or any combination of these (Faulds and Varga, 1998; Smith et al., 2001; Goteti et al., 2013; Koning et al., 2013). The observation of opposing polarities of half-grabens between the northern and southern parts of the Albuquerque basin necessitates the presence of an accommodation zone in the central Albuquerque basin, but its geometry has been controversial. Originally, this structure was hypothesized to be a narrow fault zone that was a continuation of the Tijeras fault at the southern end of the Sandia uplift (Fig. 3) (Cather, 1992; Lewis and Baldridge, 1994; May et al., 1994; Russell and Snelson, 1994). However, aeromagnetic surveys of the Albuquerque basin did not directly image such a structure (Grauch, 2001), calling into question its existence. More recently, utilizing diverse geophysical data sets, Grauch and Connell (2013) suggested that this accommodation zone instead consists of a series of NW-trending anticlinal structures and referred to it as an oblique anticlinal accommodation zone, after Faulds and Varga (1998).

METHODS OF STUDY

The goals of this study are to construct evolutionary models of the northern and southern parts of the Albuquerque basin that explain the observed high- and low-angle normal fault networks. Here we synthesize relationships between basin geometry and fault networks, as well as timing of faulting for producing evolutionary (kinematic) models. We constructed longitudinal topographic profiles from 10 m digital elevation models to observe variation in rift-flank elevation from north to south on each rift flank (lines A–A′ and B–B′ in Fig. 3). These were compared to longitudinal profiles that were created along the base of the Santa Fe Group rift fill near the eastern and western edges of the rift which are based on structure contour maps by Grauch and Connell (2013).

Fault geometries and ages were also investigated along transverse cross-section lines C–C′ and D–D′ through the Sandia and Ladron uplifts (Fig. 3). These cross sections are parallel to the extension directions as inferred from slickenlines on fault planes. Fault dips were compiled, and geologic cross sections were constructed from published geologic maps in the northern (Connell, 2008) and southern (Machette, 1978; DeMoor et al., 2005; Connell and McCraw, 2007; Read et al., 2007) parts of the Albuquerque basin, as well as new mapping of faults.

We also report new data on timing constraints on fault activity in the Sandia and Ladron uplifts inferred from apatite fission-track (AFT) and apatite (U-Th)/He (AHe) cooling ages. We build on existing data that includes AFT ages for the Ladron uplift (Kelley et al., 1992) and AFT and AHe data for the Sandia uplift (House et al., 2003) with the ultimate goal of producing continuous thermal history models of same samples, or nearby samples, using both AFT and AHe constraints. The power of using both AFT and AHe on the same sample is that the temperature sensitivities of AFT (120–60 °C) overlap with those of AHe (90–30 °C), providing the ability to jointly invert AFT and AHe data sets to constrain the sample’s continuous cooling path through the temperature range ∼120 to ∼30 °C (e.g., Donelick et al., 2005; Ketcham, 2005; Flowers et al., 2009).

To expand upon the existing thermochronologic data set, three new samples were collected from the Ladron uplift for AFT and AHe. Apatite grains were picked from sample 88LAD-6 in the footwall of the Jeter detachment for AHe analysis. Two additional samples were collected from the hanging wall of the Jeter detachment (Fig. 3) (footwall of the Silver Creek fault). Sample 91LAD-43 provided an AFT age, while JR10-6 yielded sufficient apatite grains for AHe analysis (Tables 1–3). For AHe analysis, individual apatite crystals were selected using a standard binocular microscope. Suitable grains are sufficiently large, have euhedral crystal shape, and have no visible inclusions. Individual grain lengths were measured in order to apply an age correction (Ft) (Farley et al., 1996). Apatite (U-Th)/He analysis was performed at the Arizona Radiogenic Helium Dating Laboratory at the University of Arizona. Methods for AFT analysis are described in Kelley et al. (1992).

In total, two existing AFT and AHe samples from the Sandia uplift (House et al., 2003), one existing AFT sample from the Ladron uplift (Kelley et al., 1992), and three newly-acquired samples from the Ladron uplift were targeted to produce four new joint AFT and AHe thermal history models from four fault blocks with the purpose of identifying times of rapid exhumation associated with slip along rift-bounding normal faults. These structural blocks for which thermal history models were produced include the footwall of the Jeter fault and the fault sliver bounded by the Jeter and Silver Creek detachments from the Ladron uplift and both the footwall and hanging wall of the Knife Edge fault in the Sandia uplift (Fig. 3). Thermal history models were produced utilizing HeFTy version 1.7.5 (Ketcham, 2005), which incorporates a radiation damage accumulation and annealing model (Flowers et al., 2009). In each simulation, 10,000 random paths were generated, with resulting envelopes that encompass all “good” fit paths (goodness of fit >0.5) and all “acceptable” fit paths (goodness of fit >0.05).

Intrabasinal faults in the northern and southern parts of the Albuquerque basin typically offset the thin Quaternary alluvium that covers much of the Santa Fe Group basin fill. Thus, estimating the long-term periods of faulting along these structures is more difficult than estimating the timing of Quaternary faulting. For this study we use these available timing constraints and report the times of most recent fault activity from the U.S. Geological Survey and New Mexico Bureau of Geology and Mineral Resources (2006).

MORPHOMETRIC ANALYSIS OF THE ALBUQUERQUE BASIN

Basin Geometry and Rift-Flank Topography

Facing rift flanks on opposite sides of the Albuquerque basin (Fig. 7A) show roughly a mirror image of the other, with the Sandia (3.2 km) and Ladron (2.7 km) uplifts forming the highest rift flanks in the northern and southern Albuquerque basin, respectively. Geophysical data indicate these uplifts are adjacent to the deepest depocenters located in the northeastern and southwestern portions of the basin. The northern Albuquerque sub-basin depocenter reaches minimum elevations of <–3 km, while the southern Belen sub-basin depocenter extends to <–2 km elevation (Fig. 3) (Grauch and Connell, 2013). Using the elevation difference between rift flanks and base of the Santa Fe Group as a first-order proxy for total throw along rift-bounding normal faults, we identify a close spatial association between regions of maximum fault throw, highest rift flanks, and localities of LANFs, as it is in these two locations that LANFs are best developed (Fig. 7A).

Fault Geometry

In both the northern and southern parts of the Albuquerque basin adjacent to the Ladron and Sandia uplifts, average fault dip increases from the rift flank toward the axis of the basin (Figs. 7B and 8). In the northern Albuquerque basin, low-angle normal faults (faults at locations 1, 2, and 3 in Fig. 3) are restricted to the Sandia Mountain front. Based on total offset and relation to the high-angle faults nearby, the Ranchos and Juan Tabo detachment faults are nearly coplanar and may have once been part of the same fault system but now exist as isolated remnants that were cut by the high-angle Rincon fault (Fig. 3C).

The Knife Edge detachment fault is most likely a separate structure from the other two because it separates Precambrian from Precambrian rocks without significant offset of the Great Unconformity. This is supported by indistinguishable AFT cooling ages on either side of the fault (House et al., 2003). Nevertheless, the orientations of the maximum and minimum strain axes calculated for the Knife Edge fault are very similar to those calculated for the Ranchos fault to the north (Fig. 6C). All other faults along the cross-section line dip >50° (Fig. 7B). In addition, much of the extension is accommodated through slip along synthetic faults, although small-offset antithetic faults are present toward the axis of the basin (Figs. 3 and 8A).

In the southern parts of the basin, faults have somewhat different geometries. Low-angle normal fault segments are restricted to the Ladron uplift, and there is a progression from low to high angle toward the center of the basin. For example, the Jeter detachment fault dips ∼15°–30°. In the hanging wall of the Jeter detachment, the Silver Creek detachment fault has multiple strands, with recorded fault plane dip measurements of 24° and 48° in the direction toward the axis of the basin (Figs. 7B and 8B; Read et al., 2007). All other intrabasinal faults in this region dip >50°, and extension is primarily accomplished through slip along faults that are synthetic with the Jeter detachment (Fig. 8B).

Thermochronometric Constraints on Timing of Fault Slip

Constraints on the timing of main fault activity associated with the Sandia and Ladron uplifts is important for producing evolutionary models for these fault systems within the northern and southern parts of the Albuquerque basin. Existing AFT ages from both locations show a roughly linear relationship with elevation (Fig. 9A) (Kelley et al., 1992; House et al., 2003). In the Ladron uplift, AFT data and track length measurements from the footwall of the Jeter detachment suggest rapid exhumation from ca. 20 to 10 Ma. Long track lengths with unimodal histograms suggest that these rocks cooled rapidly at rates of ∼12–20 °C/Ma during this time (Kelley et al., 1992). Similarly, AFT and AHe data in the footwall and hanging wall of the Knife Edge detachment in the Sandia uplift suggest rapid cooling to near-surface temperatures between ca. 22 and 14 Ma (House et al., 2003). Track length and age data indicate that this structural block cooled at rates of ∼7–12 °C/Ma during the early Miocene (Kelley et al., 1992).

In the Ladron uplift, three new samples were collected for AFT and AHe analysis (Tables 1–3). In the footwall of the Jeter detachment, sample 88LAD-6, with an existing AFT age of 14.1 ± 2 Ma (Kelley et al., 1992), yielded a single-grain AHe date of 12.4 ± 0.3 Ma (Table 3). Two samples were collected from the fault sliver bounded by the Jeter detachment and the Silver Creek fault. Sample 91LAD-43 yielded an AFT age of 59.9 ± 12.4 Ma (Table 2), and sample JR10-6 yielded an average AHe age of 9.6 ± 0.8 Ma, an average of four single-grain analyses (Table 3).

Joint AFT and AHe thermal history models for the footwall of the Jeter detachment fault suggest that it was rapidly exhumed to near-surface temperatures during a major pulse ∼20–10 Ma (Fig. 9B). In contrast, the fault sliver that is bounded by the Jeter and Silver Creek detachments (hanging wall of Jeter fault) records two separate pulses of exhumation. Initial ca. 80–60 Ma exhumation of this fault block was followed by relative stability for ∼50 m.y. A second period of uplift ca. 10–5 Ma brought this block to near-surface temperatures (Fig. 9B).

The Ladron fault along the western flank of the Ladron uplift shows reverse-sense movement (Fig. 8B), and although Lewis and Baldridge (1994) suggest that this structure could have formed during extension of the Rio Grande rift, a long belt of contractional faults along the western length of the Albuquerque basin (e.g., Callender and Zilinski, 1976; Lewis and Baldridge, 1994) suggests to us that the Ladron fault was most likely developed during the Laramide orogeny. If so, then the ca. 80–60 Ma period of cooling preserved in the hanging wall of the Jeter fault may have been accomplished through compressional deformation along the Ladron fault. The ca. 20–10 Ma period of cooling preserved in the footwall of the Jeter detachment most likely represents erosional exhumation during a time of rapid slip along the Jeter fault. Finally, the most recent ca. 10–5 Ma cooling in the hanging wall of the Jeter fault occurred after main slip along the Jeter fault and is interpreted to record normal-sense slip along the Silver Creek detachment to further exhume this fault sliver to the surface.

In the northern Albuquerque basin, thermal history models of the footwall and hanging wall of the Knife Edge detachment fault both record similar cooling histories, where both fault blocks were exhumed to the near surface ca. 20–12 Ma (Fig. 9). This suggests that movement along the Knife Edge detachment is relatively minor and that unroofing may have taken place on the larger-displacement Juan Tabo–Ranchos detachment system during this time. The Juan Tabo fault, which juxtaposes Triassic and Jurassic rocks against Precambrian granite, has accumulated more than 3 km of stratigraphic displacement (Van Hart, 1999). Low-temperature thermochronometers used in this study do not detect any Laramide-age cooling in the Sandia uplift, but previous work suggests that the Sandia uplift was affected by Laramide compression (Karlstrom et al., 1999), such that Precambrian granitic rocks were most likely not exhumed to shallow enough levels in the crust to have cooled through ∼110 °C in the Laramide.

Spatial Progression of Fault Slip

Figure 10 shows a plot of horizontal distance along cross-section lines C–C′ and D–D′ versus fault age. Although adequately constraining the times when faults became active and inactive is challenging, here we use available constraints on faults in the Albuquerque Basin from thermal history models of different fault blocks as well as latest movements reported for Quaternary faults in the Albuquerque basin. In both the northern and southern Albuquerque basin, intrabasinal faults typically record Quaternary displacements (U.S. Geological Survey and New Mexico Bureau of Geology and Mineral Resources, 2006), including the Rincon fault, with a probable latest rupture event younger than ca. 5 ka (Connell, 1995). Combining available age constraints on different faults, the early-formed faults seem restricted to the Sandia and Ladron uplifts. These faults were subsequently abandoned, and intrabasinal faults typically cut Quaternary units (Fig. 10). This observation is supported by the presence of long, narrow structural benches preserved along the flanks of the basin forming an extensional imbricate zone (Chapin and Cather, 1994). An important difference, however, is that in the Sandia region young faults tend to cut earlier-formed LANFs, whereas in the southern part of the basin there is a more systematic progression of younger synthetic faults forming basinward through time.

These observations are in good agreement with previous studies in the southwest Basin and Range that also document basinward migrations in faulting through time (e.g., Axen et al., 1999; Spelz et al., 2008; Fletcher and Spelz, 2009). These studies used a geomorphic approach of intrabasinal Quaternary scarp-forming faults to help determine the relative ages of scarp-forming earthquakes. Although beyond the scope of this paper, future work on Quaternary intrabasinal faults in the Albuquerque basin would be useful to further test the hypothesis that faulting migrated basinward through time, which would support a rolling-hinge mechanism for formation of fault networks.

DISCUSSION

Kinematic Evolution of Rift-Flank Uplifts in the Albuquerque Basin

Synthesizing observations on topography, basin depth, fault geometry, and fault ages, we propose a model for LANFs in the Albuquerque basin that applies previous models developed for highly extended terranes, where isostatic uplift is a main driving force that rotates high-angle faults to shallower angles via a rolling-hinge mechanism (Spencer, 1984; Buck, 1988; Wernicke and Axen, 1988; Wdowinski and Axen, 1992; Axen and Bartley, 1997). In the Albuquerque basin, isostatic uplift of footwall blocks has been proposed for the Sandia uplift (May et al., 1994; Roy et al., 1999) and the Ladron uplift (Wernicke and Axen, 1988; Lewis and Baldridge, 1994). Our model expands on these previous studies and expands on the recent synthesis by Grauch and Connell (2013) by attempting to integrate the observed high- and low-angle faults, geophysical data, and thermochronometric data into a robust kinematic model for the central Rio Grande rift.

Southern Albuquerque Basin

Following initial uplift ca. 80–60 Ma during the Laramide orogeny, initial extension in the Albuquerque basin began during the Oligocene (Chapin and Cather, 1994). Continued, rapid exhumation captured by AFT and AHe techniques began prior to 20 Ma, with major slip concentrated along the Jeter fault. This main period of extension is supported by thermal history models (Fig. 9), as well as by thick syntectonic rift-fill sections (Fig 7A); Grauch and Connell (2013) estimate that more than half the thickness of Santa Fe Group rift fill was deposited from ca. 16 to 8 Ma. Thermal history models suggest that the hanging wall of the Jeter fault, which preserves Laramide-age exhumation, was further exhumed to the surface ca. 10–5 Ma (Fig. 9B), which we interpret to reflect a main period of slip along the Silver Creek fault.

A several-meter-thick belt of mylonitic rocks is preserved in the Precambrian granitic rocks immediately adjacent to the Jeter detachment (Read et al., 2007) suggesting possible ductile deformation associated with the Jeter fault. However, microtextural analysis of rocks collected from this zone suggests that movement along this shear zone was reverse sense (Fig. 11). The Jeter detachment, therefore, is most likely a brittle fault that offsets rocks that have undergone an older (Precambrian?) ductile strain. In addition, Lewis and Baldridge (1994) suggested that the Jeter fault had an original dip of ∼48° based on the cutoff angle between the Jeter fault and Paleozoic beds along the western edge of the Ladron uplift.

In our model, the basin-bounding Jeter fault (pre–ca. 20 Ma) initiated as a moderately high-angle (∼48°) structure that inherited an earlier fabric in a region that had experienced Laramide contraction (Figs. 12A–12C). Continued extension caused the Jeter fault to rebound isostatically due to the lateral removal of hanging-wall material. As discussed previously, the low-angle Jeter fault lies along the southern flank of the basin depocenter (Fig. 7). We interpret this to be a potential consequence of isostatic footwall uplift, as this would elevate portions of the deep basin to shallower levels in regions where extension is greatest. After ca. 10 Ma, the Jeter detachment is interpreted to have become inactive because it was unfavorably oriented for continued slip, and the new high-angle Silver Creek fault formed basinward to facilitate continued extension (Fig. 12D). The upwarping of the abandoned Jeter detachment during progressive slip exposed the Precambrian-cored Ladron uplift as a rift salient, and continued slip began to rotate strands of the Silver Creek fault to lower angles as younger intrabasinal faults emerged (Fig. 12E). The observed abandonment of faulting in the rift-flank uplift and the continuation of faulting within the axis of the basin (Fig. 10) are characteristic features of an operative rolling-hinge mechanism in the southern part of the Albuquerque basin and suggest that LANFs have been important contributors to the overall extension and are most likely preserved at depth as they are at the surface in the Ladron uplift.

Northern Albuquerque Basin

Thermal history models generated for the Sandia uplift, along with current fault geometry, differ slightly from the Ladron block to the south. In the Sandia uplift, both the hanging wall and footwall of the Knife Edge detachment fault were exhumed at about the same time, ca. 20–12 Ma (Fig. 9). This suggests no significant offset along that structure after this time (House et al., 2003) and that another major fault was responsible for the exhumation of the Sandia block toward the surface. We suggest that main exhumation was related to the Ranchos–Juan Tabo detachment system. In addition, published cross sections of the northern Albuquerque basin display steeper faults offsetting older, slightly shallower normal faults (e.g., Connell, 2008). A new aspect of our model is that the Sandia block preserves at least three now-separate LANF segments, with segments of the once-continuous LANF system preserved between younger high-angle normal faults.

These observations lead to a kinematic model for the northern Albuquerque basin in which early-formed normal faults most likely emerged as relatively high-angle (∼50°–70°) structures (Fig. 13). Instead of being concentrated within a single narrow zone, the strain was instead partitioned among several different structures (Fig. 13C). As extension continued, faults began to rotate to shallower angles. Fault rotation in this region was likely predominantly due to isostatic rebound of the footwall block (Fig. 13D). Flexural modeling using finite-element code involving joint-inversion of topography and gravity suggests that ∼18% extension of the northern Albuquerque basin near the Sandia Mountains would result in isostatic uplift that is sufficient to produce the modern rift-flank geometry and the observed ∼15° east dip of Paleozoic sediments on the eastern side of the Sandias (Roy et al., 1999). In addition, LANF segments are found above the northern flank of the deepest depocenter in the Albuquerque basin (Fig. 7), similar to relationships in the southern parts of the basin. We again suggest this may be a potential consequence of isostatic basin uplift in a region of maximum extension, which would uplift the basin depocenter to shallower levels as it rotated early-formed normal faults to shallow levels. As the early-formed Ranchos, Knife Edge, and Juan Tabo detachment faults rotated to shallower angles, and slip along them diminished, new intrabasinal faults facilitated continued extension in the northern Albuquerque basin, including the Rincon fault, which dissected the Ranchos–Juan Tabo detachment fault system (Fig. 13E). As in the southern Albuquerque basin, the abandonment of faulting in the rift-flank uplift and the continuation of faulting within the axis of the basin suggest a rolling-hinge mechanism for the development of fault networks that accommodate extension.

The klippe of Mesozoic rocks (location 1 in Fig. 3C) has long been a conundrum in the evolutionary history of the Sandia Mountains (e.g., Read et al., 1944; Kelley and Northrup, 1975; Van Hart, 1999). The proposed evolutionary history outlined here provides a viable explanation of its development. Initially, the Mesozoic klippe formed the hanging wall of a relatively high-angle normal fault that accumulated enough displacement to place it against Precambrian granitic rocks. As slip progressed along this fault as well as others, the fault plane rotated and was eventually dissected by younger faults. The preserved structures thus require an interaction of both high- and low-angle fault segments.

Elevated Heat Flow

Our evolutionary model for the central Rio Grande rift recognizes synchronous slip on opposite mirrored active rift flanks and development of a broad accommodation zone beginning ca. 20 Ma. In addition, we perceive a link between LANF development and morphometry of both rift flanks and basin depth. This section seeks to understand the evolution of LANF systems in the southern and northern parts of the Albuquerque basin in terms of elevated heat flow that existed at the end of the Oligocene.

Widespread intermediate to silicic composition eruptions affected much of the western United States and Mexico during the Eocene–Oligocene (e.g., Elston, 1984; Lipman, 1992; Farmer et al., 2008). In Colorado and New Mexico, volcanic and magmatic activity was mainly centered in two major centers, the San Juan volcanic field in southwestern Colorado and the Mogollon-Datil volcanic field in southwestern New Mexico (Figs. 1 and 3). The Mogollon-Datil volcanic field, active from ca. 40 to 21 Ma, covers ∼40,000 km2, produced more than ten calderas in southwestern New Mexico, and is surrounded by a large apron of ignimbrites and lavas (Chapin et al., 2004a). Volcanic and magmatic rocks in the Mogollon-Datil volcanic field are thought to have been derived from lithospheric sources, although the ultimate cause of melting may have been the foundering of the underlying Farallon slab, which caused asthenospheric circulation (McMillan et al., 2000). Additional minor magmatic centers are found east of the Sandia Mountains (Fig. 3), including the Ortiz belt, active from ca. 36 to 27 Ma, which consists of at least 12 individual laccoliths (Maynard, 2005).

By ca. 21 Ma, regional volcanism and magmatism had elevated the heat flow in much of New Mexico and southern Colorado (Kelley, 2002). These magmatic centers also likely heated the crust for several million years after their emplacement. Continued Neogene heat flow is exemplified by modern-day heat-flow measurements above the Socorro magma body (Fig. 3), which are higher than the heat flow estimated from a basaltic melt of its size (Reiter et al., 2010). The calculations by Reiter et al. (2010) suggest that the measured heat flow is residual from now-solidified magmatic intrusions that were emplaced ca. 1–3 Ma. Similarly, activity in magmatic and volcanic centers in southwestern New Mexico ended ca. 21 Ma, yet residual heat flow likely elevated temperatures in the crust for several million years after volcanic and magmatic activity ceased. Based on thermal history models of the Sandia and Ladron uplifts, rapid extension along early-formed faults began prior to ca. 20 Ma in the northern and southern parts of the basin. These times of rapid extension coincide with times of elevated heat flow in much of New Mexico. In the Sandia region, House et al. (2003) use AFT data, unit thicknesses, and thermal conductivity values for different rock types to estimate geothermal gradients. Their calculations indicate that at the end of Oligocene magmatic and volcanic activity, heat flow in this region was ∼105 mW/m2, corresponding to a geothermal gradient in the upper crust of ∼38 °C/km, compared to a calculated modern-day heat flow of ∼80 mW/m2, corresponding to a modern geothermal gradient of ∼29 °C/km (House et al., 2003). Due to its close proximity to the Mogollon-Datil volcanic field, it is possible that heat flow in the southern Rio Grande rift may have even been higher than in the Sandia uplift.

In the Ladron uplift, our model suggests that rapid slip along the Jeter fault led to isostatic uplift, rotation of early-formed faults to shallower angle, and the emergence of new intrabasinal faults at a high angle through a rolling-hinge mechanism. Similar mechanisms of formation for LANFs have been widely applied to explain the formation of core complexes in regions of large magnitude (>100%) extension, such as the Basin and Range Province (e.g., Spencer, 1984; Wernicke and Axen, 1988; Axen and Bartley, 1997; Lavier et al., 1999; Fletcher and Spelz, 2009), along mid-ocean ridges (e.g., Ohara et al., 2001; Okino et al., 2004; Escartín et al., 2008; Smith et al., 2012), as well as in regions undergoing orogenic collapse such as Tibet (Chen et al., 1990; Pan and Kidd, 1992; Harrison et al., 1995; Kapp et al., 2000, 2008). As discussed previously, the development of core complexes such as these requires hot crustal conditions to expose ductilely-deformed rocks at the Earth’s surface. Deformation along the Jeter detachment fault is entirely brittle, although fault ages and geometries are consistent with early stages of a core-complex evolutionary model. The high-elevation footwall, rift promontory aspect, and domal character of a footwall core are all similar to core complex geometries seen in high-extension regions (e.g., Spencer, 1984). Thus we propose that the Ladron uplift is an example of an embryonic core complex and that a combination of insufficient extensional strain and heat flow following late Oligocene magmatic and volcanic activity inhibited the development of a mature core complex exposing ductilely-deformed rocks.

In the Sandia uplift, the style of extension is very similar to that observed in the southern Albuquerque basin, where older faults are rotated to shallow angles as extension progresses. The Knife Edge fault accumulated relatively minor amounts of slip, and unroofing of the Sandia block was most likely accomplished via slip along the Juan Tabo–Ranchos detachment fault system. In this part of the basin, it is clear that younger faults (i.e., Rincon fault) crosscut older LANFs, suggesting that perhaps fault rotation may have been due to a combination of fault block rotation coupled with isostatic uplift.

Regional Implications for Extension in the Rio Grande Rift and Basin and Range

The results of this study emphasize that an interaction between high- and low-angle normal faults resulted in the present-day geometry of the Albuquerque basin. The hypothesis that these structures formed through a rolling-hinge mechanism provides a framework for how to interpret the current geometry of the Albuquerque basin (Grauch and Connell, 2013) within the context of its structural and sedimentological evolution from the mid-Miocene to the present. We build on previous models emphasizing possible rotation of LANFs in the Ladron and Sandia uplifts (e.g., Lewis and Baldridge, 1994; May et al., 1994) and incorporate the most detailed geophysical observations of the subsurface (Grauch and Connell, 2013) to illustrate that mechanisms that operate in the highly extended Basin and Range Province also operated to a lesser degree in the Rio Grande rift. The observation that several of these LANFs in the Albuquerque basin may have formed through a rolling-hinge mechanism suggests that LANFs exposed in other parts of the rift (Fig. 1) may have also formed through a similar mechanism. For example, Blanca Peak in southern Colorado forms a salient into the San Luis basin of the Rio Grande rift that is cored by Precambrian crystalline rocks, similar to the Sandia and Ladron uplifts of the Albuquerque basin. Blanca Peak is also bounded along its western edge by a system of west-dipping LANFs (Jones, 1991; Benson, 1997), raising the possibility that fault networks in this vicinity of the rift may also represent early stages in the development of a core complex.

Fully developed metamorphic core complexes are found within a large belt of the highly extended Basin and Range Province (Fig. 1), but the recognition of a rolling-hinge mechanism operating within the Rio Grande rift at a lesser scale offers an opportunity to learn about the early stages in the development of a core complex. In the Albuquerque basin, the Sandia and Ladron uplifts represent the beginnings of a domal footwall, which, in mature core complexes, is generally thought to form through buoyancy forces due to removal of hanging-wall material (e.g., Spencer, 1984; Buck, 1988; Wernicke and Axen, 1988). Low-angle normal faults in the Albuquerque basin are also entirely brittle, in contrast to LANFs associated with mature core complexes, which can accumulate up to 50 km of slip and expose a thick (0.1–3 km) ductile shear zone (e.g., Axen, 2004). Thus, at least in the case of the Albuquerque basin, initial doming of the footwall and significant rotation of early-formed normal faults occurred during the narrow rift stage represented by the Albuquerque basin and before exposure of the brittle-ductile transition at the surface.

As seen in Figure 1, earliest extension associated with core complexes in the Basin and Range occurred in the north and swept southward through time, while a similar sweep of extension began in northern Mexico and propagated north. These trends closely followed similar sweeps in volcanism, although extension was generally postponed by several million years following volcanism (e.g., Axen et al., 1993; Humphreys, 1995, 2009; Dickinson, 2009). These space-time relationships are generally attributed to removal of the underlying Farallon plate from the base of North America in some fashion following the Laramide orogeny (e.g., Coney and Reynolds, 1977; Atwater, 1989; Chapin et al., 2004b). In contrast, while earliest extension in the Rio Grande rift began ca. 36–37 Ma (Kelley and Chamberlin, 2012), the ca. 20–10 Ma time periods of rapid cooling in the Sandia and Ladron uplifts are similar to previous studies of the rift that document synchronous extension along its length during the mid-Miocene (Kelley et al., 1992; Chapin and Cather, 1994; Landman and Flowers, 2013; Ricketts et al., 2014b). These observations on the timing of extension in the narrow Rio Grande rift differ from the sweeps of extension and volcanism in the highly extended Basin and Range Province, suggesting the possibility of slightly different driving mechanisms causing extension (e.g., Ricketts et al., 2014b). Nevertheless, despite these differences, it appears that in both domains LANFs are developed through a rolling-hinge mechanism, and if extension progressed in the Rio Grande rift, it would eventually result in the formation of a metamorphic core complex similar to those observed in the Basin and Range Province.

CONCLUSIONS

Our work sheds light on early stages in the development of LANF systems in regions of elevated heat flow and leads to the following general conclusions. (1) The Albuquerque basin evolved into a narrow, low-extension (∼17%–35%) setting characterized by an abandonment of faults within rift flanks through a rolling-hinge mechanism, 10-km-scale salients cored by basement uplifts that form the highest parts of the rift flanks, broad accommodation zones separating deep sub-basins, and, at least in this case, simultaneous slip on opposite sides of the rift. (2) We support previous work suggesting that core complexes typically form in regions of hot crust where decoupling between the upper brittle crust and lower ductile crust localizes strain. (3) In the southern Albuquerque basin, elevated heat flow and extension magnitude were insufficient to lead to a mature core complex where ductilely-deformed rocks are preserved in the footwalls of large-offset LANFs. (4) Low-angle normal faults in the northern Albuquerque basin may have formed dominantly through isostatic footwall uplift. Thus, we show that fundamental processes that are required to produce core complexes (both oceanic and continental) operated to moderate degrees within the Rio Grande rift from ca. 25 to 5 Ma. These regions, therefore, where LANFs make up a small percentage of the total exposed fault population, are important and under-utilized natural laboratories for documenting the sequential stages in development of highly extended core complexes and provide a new perspective on the importance of LANFs in narrow continental rifts.

Partial funding came from the New Mexico Geological Society, New Mexico Statemap program, the Alfred P. Sloan Foundation, and the University of New Mexico for JWR, and grants EAR-0607808, 0838575, and 1119629 for KEK. We thank Dirk Van Hart, Gary Axen, Peter Reiners, Tien Grauch, and an anonymous reviewer for fruitful discussions and feedback that improved the paper, and Adam Read, who collected samples used for microstructural analysis of the Jeter fault.