The Walker Lane is a broad shear zone that accommodates a significant portion of North American–Pacific plate relative transform motion through a complex of fault systems and block rotations. Analysis of digital elevation models, constructed from both lidar data and structure-from-motion modeling of unmanned aerial vehicle photography, in conjunction with 10Be and 36Cl cosmogenic and optically stimulated luminescence dating define new Late Pleistocene to Holocene minimum strike-slip rates for the Benton Springs (1.5 ± 0.2 mm/yr), Petrified Springs (0.7 ± 0.1 mm/yr), Gumdrop Hills (0.9 +0.3/−0.2 mm/yr), and Indian Head (0.8 ± 0.1 mm/yr) faults of the central Walker Lane (Nevada, USA). Regional mapping of the fault traces within Quaternary deposits further show that the Indian Head and southern Benton Springs faults have had multiple Holocene ruptures, with inferred coseismic displacements of ∼3 m, while absence of displaced Holocene deposits along the Agai Pah, Gumdrop Hills, northern Benton Springs, and Petrified Springs faults suggest they have not. Combining these observations and comparing them with geodetic estimates of deformation across the central Walker Lane, indicates that at least one-third of the ∼8 mm/yr geodetic deformation budget has been focused across strike-slip faults, accommodated by only two of the five faults discussed here, during the Holocene, and possibly half from all the strike-slip faults during the Late Pleistocene. These results indicate secular variations of slip distribution and irregular recurrence intervals amongst the system of strike-slip faults. This makes the geodetic assessment of fault slip rates and return times of earthquakes on closely spaced strike-slip fault systems challenging. Moreover, it highlights the importance of understanding temporal variations of slip distribution within fault systems when comparing geologic and geodetic rates. Finally, the study provides examples of the importance and value in using observations of soil development in assessing the veracity of surface exposure ages determined with terrestrial cosmogenic nuclide analysis.
Fault slip rates are fundamental to understanding earthquake recurrence of active faults, enabling probabilistic estimates of the timing and size of future earthquakes (e.g., Youngs and Coppersmith, 1985; Stein and Wysession, 2002; McCalpin, 2009; Burbank and Anderson, 2012). New approaches in geodetic modeling have allowed slip rate estimates to be defined within models to yield contemporary slip rates (e.g., Sauber et al., 1994; Meade and Hager, 2005; Loveless and Meade, 2010; Hammond et al., 2011; Bormann et al., 2016), which may not be representative of longer-term strain release behavior. Variations in slip along individual faults (e.g., Kirby et al., 2006; Dolan et al., 2016; Frankel et al., 2007a, 2007b, 2011; Lee et al., 2009), seismic clustering (Rockwell et al., 2000; Kenner and Simons, 2005; Perouse and Wernicke, 2016), and slip partitioning between faults (Bennett et al., 2004; Pérouse and Wernicke, 2016) can promote temporal and spatial variations of strain release within fault systems. Here, we compare prior geodetic slip rate estimates with newly determined geologic slip rates, using additional geologic observations bearing on the relative recency of rupture, to understand the temporal and spatial behavior of late Quaternary slip across a series of subparallel strike-slip faults within the Walker Lane, a broad intraplate dextral shear zone (Fig. 1; Thatcher et al., 1999; Unruh et al., 2003; Faulds et al., 2005).
We focus our study in the central Walker Lane, where transtensional deformation occurs across a series of normal and strike-slip faults that lie between the eastern Sierra Nevada and western Basin and Range (Fig. 1; Unruh et al., 2003; Wesnousky, 2005a, 2005b; Wesnousky et al., 2012). There are five major subparallel active northwest striking strike-slip faults that define the northeastern boundary of the central Walker Lane (Nevada, USA): the Agai Pah, Gumdrop Hills, Indian Head, Benton Springs, and Petrified Springs faults (Fig. 1). At this latitude, geodesy indicates ∼8 mm/yr of dextral shear across the ∼140 km width of the central Walker Lane (Fig. 1; Hammond and Thatcher, 2004; Bormann et al., 2016). Reported long-term geologic slip rate estimates for the strike-slip faults are absent to limited (e.g., Wesnousky, 2005a), thus it remains unclear of the long-term total deformation accommodated by the strike-slip faults and how deformation occurs on temporal and spatial scales within this system of strike-slip faults.
We define new geologic rates of slip and compare with estimates of the relative recency and age of surface rupture for the central Walker Lane strike-slip faults based on regional-scale Quaternary fault mapping, using (1) digital elevation models (DEMs) constructed from airborne lidar and structure-from-motion (SfM) modeling of unmanned aerial vehicle photography, and (2) terrestrial cosmogenic nuclide (TCN) and optically stimulated luminescence (OSL) dating of displaced alluvial deposits. Probabilistic bounds on the rates of slip are computed at targeted sites along four of the major strike-slip faults and provide the basis to compare geologic rates of deformation to those inferred geodetically in the central Walker Lane. The aim of the study is to better define the temporal and spatial pattern of slip across the strike-slip faults of the central Walker Lane and, as a result, also place a limit on the portion of slip that is being accommodated by faults to the west, where geologic evidence of long-term shear strain release remains elusive (e.g., Wesnousky et al., 2012; Bormann et al., 2016).
THE CENTRAL WALKER LANE
Geodesy shows the contemporary strain field of the Walker Lane to be generally right-lateral shear dominated, with the principle strain axis oriented to the northwest, while a secondary component of north-south extension is evident in the central and southern portions of the Walker Lane (e.g., Oldow et al., 2001; Kreemer et al., 2009; Bormann et al., 2016). Active transtensional deformation occurs in the central Walker Lane across a system of active northwest-trending right-lateral strike-slip faults (e.g., Stewart, 1988; Wesnousky, 2005a), north-south–oriented normal faults (Unruh et al., 2003; Wesnousky et al., 2012), and a conjugate set of northeast striking left-lateral faults (Rogers, 1975; Wesnousky, 2005a; Li et al., 2017) (Fig. 1). The pattern and structural style of faulting indicate long-term partitioning of the strain field, with east-west extension-dominated deformation on the normal faults to the west, and dextral shear deformation focused in the eastern portion of the central Walker Lane, across the strike-slip faults of this study (Fig. 1). Yet, neotectonic (Unruh et al., 2003; Wesnousky et al., 2012) and geodetic (Kreemer et al., 2009; Hammond et al., 2011; Bormann et al., 2016) studies find shear deformation to be more evenly distributed across the central Walker Lane, and have proposed several mechanisms of right-lateral shear strain release in the absence of strike-slip faulting, including: (1) block rotations of the normal fault–bounded blocks (Cashman and Fontaine, 2000; Wesnousky et al., 2012; Carlson et al., 2013); (2) oblique slip on the normal faults (Bormann et al., 2016); and (3) partitioned slip on strike-slip faults outboard of the range-bounding faults (Dong et al., 2014). Obtaining geologically based rates from geomorphic observations of dextral slip along the normal faults has remained challenging (e.g., Wesnousky et al., 2012; Bormann et al., 2016). We focus our attention to the east, where obtaining long-term horizontal slip rates appears to be more straightforward.
The strike-slip faults of the central Walker Lane strike subparallel to the contemporary northwest oriented shear strain field and record at least 34 km of cumulative right-lateral slip since the initiation of faulting (Ekren and Byers, 1984; Hardyman et al., 2000), which began as recently as 7 Ma to 10 Ma (Hardyman and Oldow, 1991). Prior geologic studies have provided qualitative Late Pleistocene slip rate estimates for only the Benton Springs (∼1 mm/yr) and Petrified Springs (1.1–1.65 mm/yr) faults, relying on soil chronosequences and standard field techniques (Wesnousky, 2005a). Results from the geodetic block modeling of Bormann et al. (2016) predict a cumulative contemporary slip rate of 0.91–3.15 mm/yr across the strike-slip faults of the central Walker Lane, providing a basis of comparison for longer-term rates derived in this study from geologic offsets.
Toward comparing the relative recency of late Quaternary surface rupture amongst the active strike-slip faults in the central Walker Lane, we compile prior geologic mapping of Quaternary formations and fault traces with new observations resulting from ∼1:10,000-scale mapping performed in this study onto a regional map, shown in Figure 2, to highlight cross-cutting relationships of the central Walker Lane strike-slip faults with mapped Quaternary formations. Fault traces and Quaternary formation contacts are modified from the 1:48,000-scale geologic quadrangle maps of Carlson (2014), Ekren and Byers (1985a, 1985b, 1986a, 1986b), Greene et al. (1991), and Hardyman (1980), and Quaternary unit maps of Wesnousky (2005a), through analysis of modern ESRI World Imagery (https://services.arcgisonline.com/ArcGIS/rest/services/World_Imagery/MapServer) and Google Earth (https://www.google.com/earth/) satellite imagery, standard field mapping techniques, and analysis of airborne lidar data sets (1 m resolution) over the Benton Springs and Petrified Springs faults. The extent of lidar data coverage is shown in Figure 2, and the data are publicly available and described at OpenTopography (https://www.opentopography.org).
Alluvial fan formations are divided into three general units (Fig. 2; Qfo—Middle to Late Pleistocene; Qfi—Late Pleistocene; Qfy—Holocene), primarily based on variations in surficial characteristics (e.g., degree of incision, breadth and shape of interfluves, and desert pavement development) following prior methods and general age delineations commonly applied to the region (Bull, 2008; Bell et al., 2004; Wesnousky, 2005a; Frankel et al., 2007a; Koehler and Wesnousky, 2011; Wesnousky and Caffee, 2011; Li et al., 2017). Descriptions of the mapped formations are provided in Supplemental Item A1.
To better illustrate Holocene activity within detailed figures, we further subdivide
d the Holocene-age alluvial fan unit s (Qfy) into unit Qfy1, representative of active alluvium, and unit Qfy2, comprising older and higher alluvial surfaces. Fault traces are mapped within Quaternary deposits where fault scarps and lateral offsets are present. We also include the mapped traces of the faults through bedrock from the geological mapping to provide context of fault strike and location (Fig. 2).
Observations: Quaternary Expression and Recency of Surface Rupture
The Quaternary mapping along each of the central Walker Lane faults show the Indian Head and southern Benton Springs faults to exhibit fault scarps in Holocene alluvial fan deposits (unit Qfy; Fig. 2), while the surficial expressions of the Agai Pah, Gumdrop Hills, and Petrified Springs faults show scarps only within Middle to Late Pleistocene alluvial fan deposits (units Qfi and Qfo). The following describes the expression of Holocene faulting along the Benton Springs and Indian Head faults.
The Holocene fault trace of the Benton Springs fault extends discontinuously for ∼40 km, from the southern end to the northern end of Soda Springs Valley (Fig. 2). The Holocene expression is most pronounced north of Dunlap Canyon (Fig. 2) where the fault diverts from the Gabbs Valley Range front and forms subtle west-facing discontinuous fault scarps within older Holocene alluvial fans in Soda Springs Valley (Fig. 3A). An incised drainage channel is right-laterally offset by ∼3 m (Fig. 3A) within the Holocene alluvial fans. Prior trenching farther to the south in Dunlap Canyon (Fig. 2) shows evidence for an 800 14C yr B.P. surface-rupturing earthquake (Wesnousky, 2005a). The observed Holocene fault trace and associated lateral offset of ∼3 m, located <15 km northwest of the trench, are likely the result of the earthquake event observed in the trench.
Faulting of Holocene deposits along the trace of the Benton Springs fault is limited to the southern segment of the fault located within Soda Springs Valley, and not recognized immediately northward where the fault cuts through the bedrock of the Gabbs Valley Range (Fig. 2). Farther north, expression of the same fault is buried by unit Qfy alluvial fans and limited to Middle to Late Pleistocene alluvial fans (units Qfi and Qfo; Fig. 2). Analysis of the northernmost lidar data set along the eastern flank of the Terrill Mountains (Fig. 2) shows the fault trace to be concealed by post-Lahontan highstand (15.5 ka; Adams and Wesnousky, 1999) alluvial fan deposits (Fig. 2). These observations indicate that the northern portion of the Benton Springs fault, beyond where the fault cuts through the Gabbs Valley Range (Fig. 2), has probably not ruptured during the Holocene, and the most recent rupture(s) was limited to the southern segment within Soda Springs Valley.
Holocene rupture on the Indian Head fault is displayed at the southern end of the Indian Head Peak ridge within Soda Springs Valley (Fig. 3B) as <0.5-m-high scarps manifested within unit Qfy2 alluvial fans, which have surfaces that are darker in color than, and sit slightly above, the active alluvium (Fig. 3B). An 8 m right-lateral offset of an alluvial fan-terrace riser is recorded by these deposits (Fig. 3B). The Holocene trace of the Indian Head fault is also observed north of the Gillis Range within Win Wan Valley (Fig. 2), showing similar characteristics of short discontinuous scarps within older Holocene alluvial fan surfaces, and after a short distance, dies out within the Holocene alluvial fan deposits in Win Wan Valley (Fig. 2).
South of U.S. Highway 95 and along strike of the Indian Head fault, an active fault trace trends along the western flank of the Black Dyke Mountain (Fig. 2). The fault trace also forms subtle (<0.5 m high) west-facing fault scarps within older Holocene alluvial fans. The expression of this southern fault strand is similar to that of the strand along the Indian Head Peak ridge. No fault scarps are observed within the younger Holocene basin-fill deposits and are likely concealed by the active basin-fill deposits in Soda Springs Valley between Black Dyke Mountain and the Indian Head Peak ridge (Fig. 2). Due to the similarity of expression and linearity of the two traces, we interpret the Black Dyke Mountain fault strand to be the southern continuation of the Indian Head fault, and the Holocene trace to be concealed by the active distal fan and aeolian deposits within Soda Springs Valley. This extends the total fault length to ∼40 km (Fig. 2).
In summary, the above observations indicate that both the Indian Head and southern Benton Springs faults have ruptured at least once during the Holocene, while the absence of displaced Holocene deposits along the Agai Pah, Gumdrop Hills, northern Benton Springs, and Petrified Springs faults indicate that they have not. Furthermore, the spatial distribution of Holocene rupture within central Walker Lane strike-slip fault system is observed to have been more focused at the southern end within Soda Springs Valley (Fig. 2).
SLIP RATE SITE DESCRIPTIONS AND HORIZONTAL DISPLACEMENTS
Geological slip rates are determined at selected locations along the faults: (1) that provide a clear measurable geomorphic offset; (2) where the age of offset deposits may be assessed with quantitative Quaternary dating methods; and (3) where the offsets are considered to be the product of at least two or more earthquakes. The site locations include: site IH on the Indian Head fault, sites GDN and GDS on the Gumdrop Hills fault, site Mina1-BS on the Benton Springs fault, and site PS on the Petrified Springs fault (Fig. 2). No measurable lateral offset along the Agai Pah fault is evident from the regional mapping, thus no rate is obtained for this fault. The reported measures of horizontal offset are made from field observations aided by high-resolution DEMs constructed from either the lidar data sets or SfM modeling of drone photography, analyzed with Agisoft Photoscan software, following the approach of Angster et al. (2016). To formalize the offset measurements and evaluate the associated uncertainty, we utilize the Matlab script LaDiCaoz_v2.1 (Zielke et al., 2012; Haddon et al., 2016), which calculates an optimal offset based on cross-correlation of topographic profiles extracted from high-resolution DEMs drawn across geomorphic features on either side of the fault.
The Indian Head fault displaces a Holocene unit Qfy alluvial fan sourced from a proximal small steep canyon near the southern end of the Indian Head Peak ridge (Figs. 2, 3A, and 4A). The alluvial fan surface sits ∼1.5 m above the active drainage and is characterized by muted bar-and-swale topography constructed by angular metasedimentary gravel and scattered small angular boulders. The fault forms an ∼1-m-high southwest-facing scarp near the bedrock contact and right-laterally offsets the southern alluvial fan-terrace riser, which is present on both sides of the fault (Fig. 4B). The riser crest and base are offset 7.6 and 8.3 m, respectively, yielding a best estimate of the offset equal to 8 ± 0.4 m (Figs. 4B and 4C). The relative consistency of offset between the measured riser and base suggest that offset occurred after deposition and full abandonment of the alluvial fan surface.
Sites GDN and GDS
The Quaternary trace at the southern end of the Gumdrop Hills fault splays into two subparallel fault traces, forming a graben within the distal portion of a unit Qfi alluvial fan in Soda Springs Valley that is sourced from a large canyon to the northeast (Fig. 2 and 5A). On the northern fault strand (site GDN; Fig. 5A), the fault exhibits an ∼3-m-high southeast-facing fault scarp within the unit Qfi alluvial fan and right-laterally deflects an incised drainage channel and offsets the southeastern alluvial fan-terrace riser of the unit Qfi alluvial fan (Fig. 5B). The crest and base of the riser record 11 and 14.2 m (12.6 ± 1.6 m) of right-lateral displacement, respectively (Figs. 5B and 5D). The southern fault strand (site GDS; Fig. 5A) forms an ∼0.5 m northeast-facing scarp within the same unit Qfi alluvial fan and right-laterally displaces the northwestern alluvial fan-terrace riser base and crest 4.2 and 5.1 m (4.6 ± 0.5 m), respectively (Figs. 5C and 5E). The measured offsets accrued post-deposition of the alluvial fan, and sum to a cumulative average displacement of 17.1 ± 2.5 m across the two fault strands.
Site Mina 1-BS
Near the southern end of the Benton Springs fault, the fault crosses in front of the Dunlap Canyon outlet at Site Mina 1-BS and forms a linear west-facing scarp within a series alluvial fans that emanated from the large canyon (Figs. 2 and 6A). On the northern side of the active drainage channel, the fault forms an ∼2-m-high scarp and right-laterally displaces the southern riser of a unit Qfi alluvial fan (Fig. 6B). Reconstruction of the alluvial fan riser crest, slope, and base provides relatively consistent offset measurements of 36, 34.2, and 33 m, respectively (Fig. 6C). The three offsets average 34.5 ± 1.5 m and occurred sometime after abandonment of the alluvial fan.
Site PS is located along the northern end of the Petrified Springs fault (Fig. 2), where the fault follows a very linear trace, defined by a predominantly northeast-facing scarp where it displaces distal portions of unit Qfi and Qfo alluvial fans (Fig. 7A). A beheaded channel, incised into a unit Qfi alluvial fan, is preserved on the east side of the fault trace and lies below a large fault scarp that preserves a high unit Qfo alluvial fan terrace surface (Figs. 7A and 7B). Reconstruction of the base and crest of the southern channel margin of the beheaded channel to the northern prow end of the bedrock ridge on the south side of the active channel yields an offset measurement of 92.5 ± 0.5 m (Fig. 7C). The offset accrued after the incision of the beheaded channel into the unit Qfi alluvial fan.
Measurements of in situ concentrations of TCNs 10Be and 36Cl are used as a primary method to place numerical age limits on the timing of alluvial fan formation at each slip-rate study site. TCNs accumulate at a relatively well-known rate resulting from the interaction of cosmic rays with certain minerals at Earth’s surface, which systematically decay with depth, allowing for the duration of surface exposure to be quantified (Gosse and Phillips, 2001). We use TCN depth profiles to mitigate the uncertainty of inheritance (e.g., Gosse and Phillips, 2001) and surface samples where possible. The details of sample collection, processing methods, and age modeling are provided in Supplemental Material B (footnote 1), and the associated Tables SB1 and SB2 provide a summary of sample data, measured 10Be and 36Cl concentrations from the Purdue Rare Isotope Measurement (PRIME) Laboratory (West Lafayette, Indiana, USA), and parameters used for age modeling. Results of the 10Be and 36Cl profile age modeling are provided in Tables 1 and 2, respectively. At one location, OSL analysis of a buried sand lens within an alluvial fan deposit is used to further constrain the age of a displaced alluvial fan surface. Details of the OSL sample collection and laboratory processing methods performed at the University of Cincinnati (Ohio, USA) are also provided in Supplemental Material B. In addition to the numerical dating methods, soil profile descriptions of each pit exposure are recorded following the techniques and terminology described by Birkeland (1984) and Soil Survey Division Staff (1993). Characteristics such as soil thickness, clay content, carbonate stage, and development of B-horizons have been observed to increase as a function of time (e.g., Bachman and Machette, 1977; Machette, 1985; Birkeland, 1984; Harden et al., 1991), and are used to assess relative ages of the offset alluvial surfaces and compared with the results from the TCN and OSL analyses. Soil profile characteristics of each pit are provided in Table SB3 (footnote 1).
Site IH Pit
The location of the pit excavated on the surface of the displaced unit Qfy2 alluvial fan surface at site IH along the Indian Head fault is shown in Figure 4B and sits ∼2 m above the active alluvium. The displaced surface is characterized by muted bar-and-swale topography and has poorly developed desert pavement within the interfluves. The pit was dug into an interfluve surface and exposed a poorly sorted unconsolidated massive conglomerate composed of angular pebbles and cobbles, capped by faint soil characterized by a 7-cm-thick silt-rich Av-horizon underlain by an 11-cm-thick Bw-horizon and stage I carbonate development (Machette, 1985) (Fig. 8A; Table SB3 [footnote 1]). The weak soil development suggests this surface to be no older than Holocene in age. The 10Be concentrations of six samples, collected within the pit at 30 cm intervals, are similar throughout the entire depth profile (Fig. 8A), showing no systematic decrease with depth. Modeling of the concentrations using the Hidy et al. (2010) depth profile Monte Carlo age modeler provides a surface age of 10.4 +37.1/−0.3 ka (Table 1).
There are several observations that lead us to disregarding the upper age bound of the 10Be profile age result. First, the lack of a systematic decrease of 10Be concentration with depth observed in the site IH cosmogenic profile (Fig. 8A) is generally symptomatic of a younger deposit (Owen et al., 2011), where the inherited TCN signal is likely exceeding the concentrations of 10Be produced in situ. Second, the large positive uncertainty (+37 ka) might not reflect a natural condition but rather be a result of the software modeling settings selected for the best fit, which might have forced an expected increase in concentration near the surface (<30 cm) (Fig. 8A). The increase expressed by a rollover of the depth profile near the surface is not reflected by the sample concentrations and is likely not real. Finally, the weak soil and the geomorphic expression of the alluvial fan, exhibited by bar-and-swale topography and little desert pavement development, corroborate a Holocene surface. Together, these observations indicate that the lower end of the 10Be age (10.8–10.5 ka) more likely represents the depositional age of the unit Qfy2 alluvial fan (Fig. 4B).
Site GDN and GDS Pits
Pits were dug into the dissected unit Qfi alluvial fan at slip-rate sites GDN and GDS along the Gumdrop Hills fault (Figs. 5B and 5C). The surfaces of the unit Qfi alluvial fan at both sites are correlative and display relatively smooth gently dipping surfaces that have a moderately developed desert pavement and are sparsely vegetated. Both pits exposed similar lithologies of poorly sorted angular to subangular massive pebble-cobble conglomerate (Figs. 9A and 9B). The soil in the site GDS pit is characterized by a 19-cm-thick sandy clay loam Av-horizon underlain by an ∼25-cm-thick sandy loam Bw-horizon (Fig. 8B) with stage I carbonate development (Machette, 1985) extending ∼1.5 m below the surface (Table SB3 [footnote 1]). The soil observed in the site GDN pit is found to be similar, characterized by a 10-cm-thick silty loam Av-horizon underlain by an ∼15-cm-thick sandy loam Bw-horizon (Fig. 9B), and also having stage I carbonate development extending through the entire profile (Table SB3). The thicker Bw-horizons observed in these two pits show that these soils are more developed than the soil observed in the site IH pit, and are comparable to the regional soils associated with the 15.5 ka Lahontan highstand and calibrated with geochronology data (e.g., Adams and Wesnousky, 1999; Benson, 1978; Reheis et al., 1989; McFadden et al., 1998). On this basis, we interpret this surface to be ca. 15.5 ka in age.
The 10Be and 36Cl concentrations of six profile samples collected from the site GDS pit are plotted in Figures 8B and 9A, respectively. The 10Be concentrations of the site GDS pit show a systematic decrease with depth (Fig. 8B), despite the sample from 100 cm depth showing a slight variance from the predicted modeling of Hidy et al. (2010). The best-fit profile provides an age of 73.8 +39.6/–18.6 ka (Table 1). The 36Cl concentrations for the site GDS pit also display a general decrease of concentration with depth, however there is more variance of the concentrations from the predicted modeling (Fig. 9A), which yields a preferred age of 61.4 ± 1.2 ka (Table 2). The 36Cl concentrations for the site GDN pit profile show more agreement with the modeling (Fig. 9B) and provide an age of 70.2 ± 5.7 ka (Table 2), providing an age range of 60–71 ka for this unit Qfi alluvial fan surface based on TCN 36Cl. We were unable to perform 10Be analysis within the site GDN pit due to low quartz yield for some fractions of the depth profile, and 10Be results from the GDS depth profile provide an age range of 113–55 ka. Although 36Cl age calculations fall within 10Be age range, both TCN analyses are much older than the interpreted ca. 15.5 ka younger age of the soils observed in both pits. Despite the lack of quartz, we successfully obtained an OSL sample from a lens of fine- to medium-grained sand at ∼1.5 m below the surface within the site GDN pit (Fig. 9B). Details of sample collection and processing are provided in Supplemental Material B (footnote 1). The OSL sample yields an age of 18.9 ± 1.2 ka (Fig. 10).
The age results for the unit Qfi surface at sites GDN and GDS determined from the TCNs, OSL analysis, and interpreted age of the soil are compared in Figure 11. The TCN ages span 113.4 to 55.2 ka and do not overlap with the much younger age of the OSL sample (18.9 ± 1.2 ka) (Fig. 11). There are several observations that lead us to interpret that the younger OSL age of the deposit is more representative of the age of the surface than the age range indicated by the TCN analysis. First, a soil developed on a surface of 60–80 ka or slightly older as suggested by the cosmogenic analysis would be expected to display a significant Bt-horizon and a more advanced carbonate development stage (e.g., Harden et al., 1991; Machette, 1985), which are not observed. Second, the weak soil development observed within the two pits (Fig. 11) is consistent with the age result of the OSL sample, which is representative of the deposition time of the sand lens. The sand lens lies ∼1.5 m below the surface, thus it is older than the alluvial fan surface, and the OSL age provides a maximum age for the fan surface. On these bases, we interpret the OSL age (18.9 ± 1.2 ka) to represent the maximum age of the unit Qfi alluvial fan surface at sites GDN and GDS.
Site Mina1-BS Pit
The site Mina1-BS pit is located on the displaced unit Qfi alluvial fan surface along the Benton Springs fault (Fig. 6B). The unit Qfi fan surface at this site sits ∼5 m above the active channel and is mantled by a well-developed desert pavement that is sparsely vegetated. The pit exposes a poorly sorted, massive, subrounded pebble conglomerate with a moderately developed soil established within the upper ∼90 cm of the surface, characterized by an ∼20-cm-thick rubified Bt-horizon (Fig. 8C) and stage II carbonate development (Machette, 1985) (Table SB3 [footnote 1]). The elevated clay content and increased carbonate stage of this soil indicate that this surface is older than the soils observed in the site GDN and GDS pits and likely predates the Lahontan highstand (ca. 15.5 ka; e.g., Adams and Wesnousky, 1999). We interpret this surface to be >15.5 ka in age based on the soil.
The 10Be concentrations of six depth profile samples from the site Mina1-BS pit generally display a decrease in concentration with depth, with the exception of the sample at 30 cm depth (Fig. 8C). Modeling of the profile concentrations using the Hidy et al. (2010) modeler provides a most probable age of 21.7 +10.8/−7.1 ka (Table 1). This age result is in general agreement with the interpretations based on the soil development and is interpreted here to represent the depositional age of the alluvial fan.
Site PS Pit (PSP1) and Surface Boulders
The PSP1 pit was emplaced on the unit Qfi surface into which the beheaded channel is incised at the PS site along the Petrified Springs fault (Fig. 7B). The alluvial fan surface sits ∼2 m above the active alluvium and has a desert pavement scattered with varnished andesite boulders (Fig. 9C). The pit exposes a poorly sorted angular pebble to cobble conglomerate with a sandy matrix. A well-developed ∼15-cm-thick rubified argillic Bt-horizon is present within the upper 25 cm of the deposit (Fig. 9C), with stage III carbonate development (Machette, 1985) below (Table SB3 [footnote 1]). The advanced soil development and elevated carbonate stage displayed in this pit suggest that this surface is the oldest of all of the sites. Prior observations of the soil at this site led Wesnousky (2005a) to correlate this surface to a 60–90 ka alluvial fan deposit within Bettles Well Canyon (Bell, 1995), located near the southern end of the Benton Springs fault (Fig. 2).
The 36Cl concentrations of the six depth profile samples display a systematic decrease with depth (Fig. 9C). Modeling from CRONUScalc software (Marrero et al., 2016) depth profile appears to capture most of the data and provides a best-fit exposure age of 153.5 ± 11.2 ka (Table 2). Four andesite boulders were also sampled from the unit Qfi surface (PSB1–PSB4; Fig. 7B). Exposure age modeling using the CRONUScalc surface online calculator (Marrero et al., 2016) yields average age results of 127 ± 33, 305 ± 76, 119 ± 24.5, and 410 ± 150 ka for boulders PSB1, PSB2, PSB3, and PSB4, respectively (Table 3).
Comparison of the age results from the 36Cl depth profile and surface boulders at site PS is shown in Figure 12. The range of age results spans 94 to 560 ka and shows two age populations: one between 94 and 165 ka, and the other from 229 to 560 ka (Fig. 12). The younger cluster (group 1 in Fig. 12), composed of two boulders and the depth profile, is more consistent with the presence of the Bt-horizon and stage III carbonate observed in the soil, whereas
a much greater carbonate development (e.g., stage IV) would be expected for a soil approaching the age range of group 2 (Fig. 12) or surpassing 300 ka (e.g., Machette, 1985; Harden et al., 1991). On this basis, the older sample cluster, which is composed solely of surface boulder ages, is considered to contain a significant inherited signal, and the younger population more likely to represent the depositional age of the unit Qfi alluvial fan surface. We interpret the average age of the younger cluster (130.8 ± 36.9 ka; Fig. 12) as the most representative of the age of abandonment of the alluvial fan, and this age is consistent with the observed degree of soil development.
Fault slip rates are calculated by dividing the measured offsets by the estimated age of the offset surface at each site. The rate estimates calculated in this manner are formalized using the probabilistic approach of Zechar and Frankel (2009), where a probability distribution function (pdf) of the slip rate is solved from integration of the age and displacement pdfs. The slip-rate results, along with values of offset and age that are at the 95% confidence level and used for the calculations, are listed in Table 4, and plots of the pdf of the slip rate for each site are provided in Figure 13. The rates determined at each site are considered to be minimum rates, due the assumption that the measured displacements accrued sometime after the alluvial fan surfaces were abandoned. The slip rates estimated here are based on measurements collected at single sites along each the four faults, and therefore there is some additional uncertainty in the estimation of rates for the entire fault, related to the site location. For example, the slip rate sites on the Gumdrop Hills fault (sites GDN and GDS) and Benton Springs fault (site Mina1-BS) are located at the southern ends of each of the faults (Fig. 2), and it is possible that the measured offsets may underrepresent the total amount of slip on each fault (e.g., Bürgmann et al., 1994; Jonsson et al., 2002), thus the slip rate estimates from this study may underestimate the actual rate for the Gumdrop Hills and Benton Springs faults.
Indian Head Fault
The slip rate on the Indian Head fault site is determined by dividing the measured 8 ± 0.8 m offset fan terrace riser at site IH (Figs. 4B and 4C) by the lower end of the age result (10.3 ± 0.5 ka) provided by the 10Be profile of the site IH pit (Fig. 8A), yielding a minimum slip rate of 0.8 ± 0.1 mm/yr (Fig. 13). This is the shortest-term geologic rate, averaged over the Holocene, of the central Walker Lane faults.
Gumdrop Hills Fault
The slip rate for the Gumdrop Hills fault is determined by dividing the cumulative right-lateral displacement of 17.1 ± 4.1 m measured between the two offset alluvial fan terrace risers at sites GDN and GDS (Figs. 5B and 5C) by the 18.9 ± 2.4 ka age result provided by the OSL sample collected within the site GDN pit (Fig. 10). The OSL age places a maximum age bound to the displaced surface, yielding a minimum Late Pleistocene slip rate of 0.9 +0.3/−0.2 mm/yr for the Gumdrop Hills fault (Fig. 13).
Benton Springs Fault
The Benton Springs fault slip rate is derived by dividing the 34.5 ± 2.8 m of displacement measured within the unit Qfi alluvial fan terrace riser (Fig. 6B) by the 10Be depth profile age result (21.8 +4.1/−0.6 ka) obtained in the site Mina1-BS pit (Fig. 8C), yielding a minimum Late Pleistocene slip rate of 1.5 ± 0.2 mm/yr (Fig. 13). This is the highest rate of all of the central Walker Lane faults and supports prior estimates of >1 mm/yr based on soil chronosequences (Wesnousky, 2005a).
Petrified Springs Fault
The slip rate determined at site PS on the Petrified Springs fault is calculated by combining the 92.5 ± 0.5 m of right-lateral offset on the beheaded channel (Fig. 7C) with the younger age group range of 124 ± 6 ka, established by the 36Cl TCN analysis (Fig. 12). This yields a minimum Late Pleistocene slip rate of 0.7 +0.3/−0.2 mm/yr (Fig. 13). This rate is lower than prior geologic estimates (1.1–1.7 mm/yr; Wesnousky, 2005a), which were based on similar geomorphic relationships at this same site (Fig. 7C). The rate discrepancy is largely attributed to differences in the age assigned to the unit Qfi alluvial fan surface into which the beheaded channel is incised, placing a maximum age on the measured offset (Fig. 7C). The Wesnousky (2005a) study relied on the soil characteristics within the site PS pit and regional age correlations, whereas this study places an absolute age based on TCN measurements on the unit Qfi alluvial fan, offering a more direct measure of the age of the unit Qfi alluvial fan at the Petrified Springs site (Fig. 7C). On this basis, we infer the rate from this study (0.7 +0.3/−0.2 mm/yr) to best define the Late Pleistocene slip rate for the Petrified Springs fault.
The new measures of offset and the application of numerical dating techniques to the respective offset surfaces has yielded both new and revised Holocene to Late Pleistocene minimum slip rates for the Petrified Springs (0.7 +0.3/−0.2 mm/yr), Benton Springs (1.5 ± 0.2 mm/yr), Indian Head (0.8 ± 0.1 mm/yr), and Gumdrop Hills (0.9 +0.3/−0.2 mm/yr) faults. The rates, measures of offset, and characteristics of the fault traces provide a basis to discuss the characteristics of earthquake recurrence, the spatial and temporal history of late Quaternary displacement on the fault system, and compare the amount of observed geodetic slip with what is now geologically accounted for by the central Walker Lane strike-slip faults. As well, they provide a basis to illustrate the importance of complementing TCN and OSL measurements of fan age with observations of soil development.
Surface Rupture Characteristics and Implication for Holocene Earthquakes
The measured offsets and mapped traces of the central Walker Lane strike-slip faults within Holocene deposits (unit Qfy; Figs. 2 and 3) provide insight to the recent rupture characteristics. Application of empirical scaling relationships between earthquake rupture length and displacement indicates that rupture of the entire ∼40-km-long Holocene trace of the Indian Head fault would produce a coseismic slip of 1–3 m with an associated Mw ∼6.8 earthquake (Wells and Coppersmith, 1994; Wesnousky, 2008). On this basis, it would be suggested that the ∼8 m of measured right-lateral displacement at site IH (Fig. 4B) is likely the result of at least two, and possibly three, surface-rupturing events in the Holocene. Dividing the 1–3 m offset by the 0.8 ± 0.1 mm/yr slip rate estimated for the fault in this study yields an expected recurrence interval for similar-size ruptures of ∼3000–4000 yr. When taking the same approach for the ∼35-km-long Holocene trace of the southern Benton Springs fault, it would be expected for coseismic displacements to be on the order of 1–3 m, also with an associated Mw ∼6.8 earthquake. The 3 m right-lateral offset observed along the Holocene trace (Fig. 3A) is in the range of expected coseismic displacements and likely representative of the most recent surface-rupturing earthquake, occurring ca. 800 14C yr B.P. (Wesnousky, 2005a). Dividing the 3 m offset by the new slip rate determined for the Benton Springs fault (1.5 ± 0.2 mm/yr) provides a reoccurrence interval of similar-size ruptures of 1600–2300 yr.
The observations suggest that the Indian Head fault has ruptured several times during the Holocene. Because the slip rates of the other faults, excluding the Benton Springs fault, are similar to that of the Indian Head fault, it can also be expected that they have also ruptured at least once during the Holocene.
Late Quaternary Spatial and Temporal Distribution of Slip
Spatial and temporal variations of fault slip rates and return times are common within generally younger and less-developed fault systems (e.g., Cowie and Roberts, 2001), such as the Walker Lane (Wesnousky, 2005b; Faulds and Henry, 2008). In the southern Walker Lane in California, spatial and temporal variability of late Quaternary slip and the rate of slip has been reported along the Death Valley–Fish Lake Valley fault system (Frankel et al., 2007a, 2007b, 2011; Fig. 1) as well as the Owens Valley–White Mountain fault system (Kirby et al., 2006; Kirby et al., 2008; Fig. 1). Similarly, late Quaternary variations of slip rate have been observed on the Warm Springs fault (California) (Gold et al., 2013; Fig. 1) in the northern portion of the Walker Lane and attributed to transfer of slip back and forth between the Warm Springs fault and Honey Lake fault system (Fig. 1). The temporal and spatial variations in fault slip rate in each of these instances are attributed by the authors of the studies to the temporal transference of slip accommodation between proximal and subparallel faults. The slip history we observe along the central Walker Lane faults is consistent with this idea.
The Quaternary mapping show each of the central Walker Lane strike-slip faults to have been active during the Late Pleistocene, evidenced by mapped fault scarps within Late Pleistocene alluvial fans (unit Qfi) (Fig. 2). The Indian Head and southern Benton Springs faults, however, are the only two fault strands displaying unequivocal Holocene activity (Figs. 2 and 3). This would suggest that only two of the five central Walker Lane strike-slip faults have been accommodating strain during the Holocene. The absence of observed Holocene faulting along the Gumdrop Hills, Petrified Springs, northern Benton Springs, and Agai Pah fault traces may be attributed to longer recurrence intervals; however, as pointed out in the discussion section above, the faults share similar slip rates and, on that basis, it would be expected that each exhibit at least one episode of Holocene displacement. These observations infer that slip rates on the central Walker Lane strike-slip faults are susceptible to vary over the short term (<10,000 yr).
Comparing Geologic and Geodetic Slip Rates
The measured values of offset versus the age of the offset feature at each slip rate site are plotted as small stars in Figure 14. The slopes of the lines define the average fault slip rate over the time since the respective offset surfaces were abandoned and extend through the time periods of which they are observed to be active based on the observations from the Quaternary mapping. At the top of the graph, we plot the cumulative geodetic strike-slip rate estimates for the central Walker Lane strike-slip faults. Summing of the rates for the faults determined to be active in the Holocene, the Benton Springs and Indian Head faults, yields a cumulative Holocene slip rate of 2.3 ± 0.3 mm/yr (Fig. 14). This rate appears to be consistent with the geodetic estimates, falling within the range of 0.9–3.2 mm/yr (Fig. 14; Bormann et al., 2016), and further supports our interpretations of the rupture recency based on Quaternary mapping. The sum of the newly determined Late Pleistocene slip rates found for the Benton Springs, Gumdrop Hills, and Petrified Springs faults yields a cumulative rate of 3.2 ± 0.7 mm/yr (thick black line in Fig. 14). This rate appears to be at the highest end of the geodetic estimates (Fig. 14), and should be regarded as a minimum cumulative rate, as Late Pleistocene slip rates for the Indian Head and Agai Pah fault remain undetermined. Provided that these two faults also displace Late Pleistocene alluvial fans (unit Qfi; Fig. 2), it is likely the cumulative Late Pleistocene rate for all of the central Walker Lane strike-slip faults would exceed the geodetic range, allowing the interpretation that the amount of slip accommodated by the central Walker Lane strike-slip fault system may have decreased since the Late Pleistocene.
This comparison between the geodetic and new geologic rates along with the observations from the Quaternary mapping are consistent with the slip calculations of Bormann et al. (2016), and indicate that the geodetically determined rate of dextral slip across the central Walker Lane strike-slip fault system has likely been accomplished by only two of the five major strike-slip faults during the Holocene. It appears that the strike-slip faults have accommodated about one-third of the ∼8 mm/yr of deformation geodetically measured across the central portion of the Walker Lane during the Holocene, and possibly one-half of the budget during the Late Pleistocene. Consequently, the remainder of the deformation budget is occurring to the southwest, where dextral shear strain is accommodated in the absence of strike-slip faults (Fig. 1). Here, paleomagnetic measurements of late Miocene volcanic rocks by Carlson et al. (2013) show the region to be dominated by vertical-axis block rotations within the Bodie Hills and Sweetwater Mountains in California that are associated with normal and left-lateral faults (Fig. 1). The calculated rates of Bormann et al. (2016) suggest relatively higher rates of block rotation and both extension and shear along the basin-bounding faults (e.g., Mono Lake fault and Silver Lake fault, California; Fig. 1), and the observations from this study would support these predictions.
In sum, it appears that dextral shear has been more or less equally distributed across the central Walker Lane throughout the late Quaternary and that the locus of slip temporally transfers within the system of strike-slip faults of this study. This pattern of equally distributed and temporally variable strain release may be common throughout the Walker Lane (e.g., Frankel et al., 2011; Gold et al., 2014). The Walker Lane has been interpreted as an incipient plate boundary (Faulds et al., 2005), and shown to be relatively structurally complex when compared to the San Andreas (Wesnousky, 2005b), the main plate boundary fault between North America and the Pacific plate (Fig. 1). It could thus be inferred that as deformation accrues in the Walker Lane, structural maturity will result in a more organized linear zone of focused dextral shear within the central Walker Lane (Wesnousky, 1988).
Utility of Soil Development in Interpreting TCN Age Results
The uncertainties associated with age results from numerical dating techniques, such as TCN dating, frequently can be derived from the assumptions needed for age calculation, such as erosion rates (Anderson et al., 1996; Gosse and Phillips, 2001; Hidy et al., 2010; Marrero et al., 2016). Soil development analysis also has uncertainty, but provides an independent measure of relative age for geomorphic surfaces, and in this study, served to be advantageous where the numerical dating techniques yielded conflicting or bimodal ages. For example, the relatively weak soil development observed in both the site GDN and GDS pits (Figs. 9A and 9B; Table SB3 [footnote 1]) along the Gumdrop Hills fault called into question the relatively consistent older TCN results from both pits. The older ages obtained through TCN analysis for the sites GDN and GDS are likely to be associated with a higher inheritance than quantified in the models, possibly the result of limitations of input assumptions needed for the model algorithms or additional surface
s processes not recognized in the scope of this study. Because the surface had no boulders or pebbles available that are suitable for cosmogenic analysis, it was not possible to cross-correlate the two TCN techniques, like the case of the PSP surface. The corroborating result of the OSL sample with the inferred soil age guided our age interpretation for the surface age at both sites of the offset alluvial fan (Fig. 11).
Similarly, the advanced stages of carbonate development and pedogenesis observed in the PSP1 pit at site PS (Fig. 9C; Table SB3 [footnote 1]) supported the relatively older 36Cl TCN age results and guided our choice amongst the two age populations (Fig. 12). Finally, the pedogenic properties of the offset alluvial fan deposits along the Indian Head and Benton Springs faults also supported the TCN age results at the site IH pit (Fig. 8A; Table SB3 [footnote 1]) and at the site Mina1-BS pit (Fig. 8C; Table SB3 [footnote 1]), respectively. The examples from this study illustrate the utility of soils as a means of comparison, and that such a comparison should be considered as a requisite internal consistency check in the evaluation of TCN surface exposure ages.
New offset measurements and the application of numerical dating techniques of displaced alluvial fan surfaces place new minimum late Quaternary dextral strike-slip rates on the Benton Springs (1.5 ± 0.2 mm/yr), Petrified Springs (0.7 +0.3⁄−0.2 mm/yr), Gumdrop Hills (0.9 +0.3⁄−0.2 mm/yr), and Indian Head (0.8 ± 0.1 mm/yr) faults of the central Walker Lane. The sum of slip rates is across this portion of the Walker Lane is generally consistent with that reported by others from geodesy. Regional Quaternary mapping of the central Walker Lane strike-slip faults, however, shows that the southern segment of the Benton Springs and Indian Head faults have likely had multi-event Holocene ruptures with estimated maximum magnitudes on the order of Mw ∼6.8 and associated coseismic horizontal displacements in the range 1–3 m, while surface rupture on the Gumdrop Hills and Petrified Springs faults have likely not occurred since the Late Pleistocene. In this regard, geology shows that the geodetically registered slip was relatively evenly divided among the four faults of the central Walker Lane during the Late Pleistocene, but has been focused on just two during the Holocene. The results thus provide an example of the temporal transfer of the locus of slip across closely spaced strike-slip fault systems, a phenomenon that cannot be recognized by geodesy alone. Finally, the observation of profiles of soil development should be considered a requisite internal consistency check in the evaluation of TCN surface exposure ages used for the assessment of fault slip rates.
This research was supported in part by National Science Foundation grants EAR-1419724 and EAR-1419789. We thank the two anonymous reviewers who provided thoughtful feedback that improved the paper. Visiting graduate student Xinnan Li from the China Earthquake Administration provided great assistance toward hand-digging many of the pits, documentation of soil characteristics, and careful sampling for cosmogenics. Many thanks to Jason Cesta who provided essential advice and assistance for the cosmogenic 36Cl sample preparation and age modeling. Useful discussions in the field and at the office with Chad Carlson, Bill Hammond, Jayne Bormann, Ian Pierce, and Tabor Reedy provided valuable insight to this study. This is University of Nevada, Reno, Center for Neotectonic Studies contribution number 77.