Accommodation of Plate Motion in an Incipient Strike‐Slip System: The Central Walker Lane

Geodesy shows that ∼7 mm/yr of dextral shear is accumulating across the Central Walker Lane in the absence of through‐going strike‐slip faults. To better understand how this shear is accommodated, we describe and quantify the patterns and slip rates of active faults extending between the Lake Tahoe and Walker Lake basins. Lidar data and geomorphic mapping show linear fault traces and stepping fault geometries consistent with the accommodation of dextral oblique‐slip motion along the Wassuk and Smith Valley faults, whereas the Mason and Antelope valley faults are primarily dip‐slip. Vertical slip rates based on cosmogenic ages in Antelope, Smith, and Mason valleys are 0.5+0.5/−0.3, 0.5+0.7/−0.4, and 0.04+0.05/−0.03 mm/yr, respectively. A strike‐slip fault in the Pine Grove Hills has a dextral slip rate of 0.3–0.8 mm/yr. The remaining unaccounted‐for shear is expected to be accommodated by off‐fault deformation, including block rotations, broad co‐seismic warping, and complex rupture patterns. Together these faults form a left‐stepping en echelon series of dextral, oblique, and normal faults that extend from south of Walker Lake to north of Lake Tahoe, similar to patterns observed in the initial stages of dextral shear laboratory models. GPS profiles spanning the entire Walker Lane are compared to these new and previously published slip rates, and show that while kinematic efficiency is ~60% and ~90% in the Northern and Southern Walker Lane, respectively, it is only ~25% in the Central Walker Lane.

To achieve these goals, we use high-resolution topographic data and imagery to search for evidence of previously unreported strike-slip faulting and to map the tectonic geomorphology of a part of the Walker Lake basin, the Pine Grove Hills, and Antelope, Mason, and Smith valleys. We present observations of northwest oriented dextral motion along both previously mapped and newly mapped sections of faults throughout the study region. We calculate rates of slip for each of the major range bounding faults in these basins using cosmogenic nuclide boulder and sediment samples from faulted alluvial fan surfaces. We then present GPS profiles of transects across the entire Walker Lane. These observations form the basis of discussions focused on resolving geologic and geodetic discrepancies and the structural evolution of shear zones.

Regional Context
The Walker Lane Belt (Stewart, 1988) is a ∼500-km-long by 100-km-wide northwest trending zone of discontinuous active faults and disrupted topography that sits between the Sierra Nevada to the west and the Basin and Range to the east (Figure 1). Roughly 20% of the ∼50 mm/yr of dextral shear between the Pacific and North American plates (Bennett et al., 2003;Dixon et al., 2000;Thatcher et al., 1999;Unruh et al., 2003) is accommodated by the normal and strike-slip faults that compose the Walker Lane. The Walker Lane is well defined geodetically by a zone of ∼9 mm/yr of northwest directed dextral shear at the latitude of Owens Valley, decreasing to ∼7 mm/yr at the latitude of Reno, NV (e.g., Bormann et al., 2016;Hammond et al., 2011;Wesnousky et al., 2012). Cumulative dextral shear also decreases from ∼110 km across the southern part (e.g., Guest et al., 2007) to 20-30 km across the northern part (Faulds et al., 2005). These decreases and the broad, discontinuous nature of faulting in the Walker Lane have been cited as evidence of the youthful expression of an incipient transform boundary (e.g., Faulds & Henry, 2008;Faulds et al., 2005).
The Central Walker Lane is here further subdivided into four domains based on orientations and kinematics of faults ( Figure 1). The northerly Carson domain (e.g., P. H. Cashman & Fontaine, 2000;Li et al., 2017;Wesnousky, 2005) is composed of the northeast-striking, sinistral Olinghouse fault (Briggs & Wesnousky, 2005), and the Carson and Wabuska lineaments (Li et al., 2017). This domain has accommodated >50° of clockwise vertical axis rotation since the Miocene (Carlson, 2017;P. H. Cashman & Fontaine, 2000). The eastern domain, modified here after the Walker Lake block of Stewart (1988), includes the northwest-striking dextral strike-slip Benton Springs, Indian Head, Gumdrop, and Petrified Springs faults (Angster et al., 2019;Wesnousky, 2005). Together, these closely spaced faults accommodate ∼4 mm/yr of dextral shear (Angster et al., 2019). The southerly Mina Deflection (e.g., Wesnousky, 2005) is composed of east-west-striking sinistral faults and forms a large right step between the dextral faults of the Southern Walker Lane and the eastern domain of the Central Walker Lane.
The western domain of the Central Walker Lane encompasses the basins and their north-striking range-bounding faults that are the focus of this paper, including the Lake Tahoe (Kent et al., 2005;Maloney et al., 2013;Pierce et al., 2017) and Walker Lake Dong et al., 2014;Kroeger, 2015) basins, andCarson (dePolo &Sawyer, 2005;Ramelli et al., 1999), Smith (Wesnousky & Caffee, 2011), Mason, and Antelope (Sarmiento et al., 2011) valleys ( Figure 2). Dong et al. (2014) report 1 mm/ yr of dextral slip near the Wassuk rangefront, but there are no other reported rates of dextral oblique-or strike-slip faults in the western or Carson domains to accommodate the ∼4 mm/yr of dextral shear measured west of the Wassuk Range. GPS transects show that the observed shear is evenly spaced across the Central Walker Lane, and is not isolated to the faults of the eastern part (e.g., Bormann et al., 2016). PIERCE ET AL.

10.1029/2019TC005612
To explain the discrepancy between geologic and geodetic measurements, Wesnousky et al. (2012) propose that in the absence of dextral faults, range scale clockwise block rotations and asymmetric, northward-opening basins accommodate the observed geodetic shear across the western domain of the Central Walker Lane. However, GPS driven block models (e.g., Bormann et al., 2016) show that dextral shear in the western domain cannot be accommodated through normal faulting and block rotations alone, suggesting that some component of dextral oblique slip is needed to accommodate shear in this region. Thus, unrecognized oblique-slip faulting in the western domain may play a larger role in accommodating Walker Lane shear than previously thought.

Quaternary Mapping
Descriptions of faulting characteristics along each of the faults and lineaments are derived from the analysis of large-scale (∼1:12,000) low-sun-angle black and white aerial photographs (Center for Neotectonic Studies library), Google Earth imagery, structure-from-motion models, and lidar data. Lidar data were compiled from three existing data sets acquired by the USGS (2009 Walker Lake Lidar), Desert Research Institute (2010 Walker River Lidar), and National Wildlife Service (2006 Walker River Lidar) and ∼334 km 2 of new data acquired in 2015 for this study through the National Center for Airborne Laser Mapping (2015 Walker Fault System NV Lidar available on OpenTopography.org). Metadata for the previously existing data sets are available at the US Interagency Elevation Inventory (coast.noaa.gov/inventory). For each basin, the lidar data sets are merged into a seamless data set with contour, hill-shade, and slope-shade maps derived for geomorphic analysis. A structure-from-motion elevation model of part of the Pine Grove Hills fault was generated using Agisoft Photoscan Pro with images collected using a DJI Phantom 3+ quadcopter. Images are georeferenced from sites located in a target region using a Trimble R10 dGPS unit, resulting in a ∼25 cm/ pixel gridded elevation model, that is available on OpenTopography.org. All faults mapped in this study are included in part in the Quaternary fault and fold database (USGS, 2020) and other prior regional geologic maps (e.g., Dohrenwend, 1982), but are presented here in greater detail as the mapping in this study is based on high-resolution topographic data and imagery, and not from prior geologic maps.
We present generalized surficial maps based on interpretations of lidar data and satellite imagery for the Pine Grove Hills and Mason, Smith, and Antelope valleys (e.g., Plate 1 (supporting information) and Figures 3,5,6,and 8). Maps are constructed following the relative geomorphic age methods outlined in Bull (1991) as used in other studies in this region (e.g., Bell et al., 2004;Koehler & Wesnousky, 2011;Li et al., 2017;Sarmiento et al., 2011;Wesnousky, 2005;Wesnousky & Caffee, 2011). All specific sites described in the following sections were field verified, but not every alluvial fan surface mapped in Plate 1 was visited in the field. Deposits are divided into units based on relative age and type of geomorphic landform/sediment: fluvial (Qfl), basin fill (Qbf), playa/lacustrine (Qp), aeolian (Qd), and four alluvial fan units (Qa-active, Qy-young, Qi-intermediate, and Qo-old). Land obscured by anthropogenic activity (anth) and undivided bedrock (bx) are also mapped. Fault traces in Figures 2, 3, 5, 6, 8, and 10 are divided as either fault scarps that are clearly expressed in Quaternary deposits (red lines) or as inferred faults and/or fault contacts between Quaternary deposits and bedrock (black lines). Alluvial fan units are divided by relative age primarily using height above modern stream grade, amount of dissection, height of fault scarp (if present), and textural differences in imagery and lidar data. Age estimates for alluvial fan units are either from cosmogenic ages presented in the following sections, or based on relative age relationships to the Qi alluvial fan surfaces, which are the youngest fans that are modified by shoreline processes when near the highstand of Lake Lahontan (Adams & Wesnousky, 1998;Reheis, 1999).
Old alluvial fan units (Qo) are the highest alluvial fan units and are often incised >10 m with well-rounded interfluves. The surfaces have soils with clay-rich Bt horizons and well-developed carbonate (Birkeland, 1984). Following others mapping in the region (e.g., Wesnousky, 2005), the Qo alluvial fans are considered to be mid-to late Pleistocene in age. Intermediate alluvial fans (Qi) sit at elevations below Qo deposits in respective map areas, are moderately incised with well-developed drainage networks, and are modified by shoreline deposits when near the highstand of Lake Lahontan. Thus, Qi alluvial fans are considered late Pleistocene in age. Qy alluvial fans are low lying, and have smooth surfaces in lidar, and a dark tone in imagery. Qy alluvial fans are Holocene and latest Pleistocene. Qa alluvial fans represent active washes and alluvial fans, and are differentiated from Qy alluvial fans by lighter tone on imagery and well-defined channel morphology. Where faults are present, Qy alluvial fans are sometimes differentiated from Qa alluvial fans by fault scarps that do not cut the younger Qa alluvial fans. Most mapped Quaternary alluvial fan units presented here match these descriptions; we note any site specific variations in the following sections.

Geochronology
Alluvial fan ages are based on measurements of cosmogenic 10 Be and 36 Cl concentrations of surface boulders and a single 10 Be depth profile (e.g., Frankel et al., 2007;Kirby et al., 2006Kirby et al., , 2008 Phillips et al., 1990). Over time, cosmic rays produce nuclides in minerals on Earth's surface, and by modeling the rate at which these nuclides accumulate we estimate the exposure age of surficial materials (Gosse & Phillips, 2001). Considerable variance in samples from alluvial fans results largely from inheritance, that is, exposure of materials prior to their final deposition (e.g., Frankel et al., 2007;Owen et al., 2011). We sampled at least six boulders per surface to help account for this variance.
Both 10 Be and 36 Cl concentrations were measured at the PRIME lab at Purdue. All 10 Be samples were processed in the Geochronology Laboratories at the University of Cincinnati following the methods of Kohl and Nishiizumi (1992). The 36 Cl samples were processed and analyzed at the PRIME lab. Detailed descriptions of sample preparation and analysis are in the supporting information.
Boulder sampling focused on the largest boulders (∼50-150-cm-diameter) from alluvial fan surfaces. Approximately 500 g samples were taken from the upper 2-5 cm of each of these boulders. Photographs of each of the sampled boulders are provided in the supporting information. 10 Be concentrations and laboratory data are listed in Table S1. The 10 Be boulder exposure ages were calculated using the Cosmic Ray Exposure Program (CREp) of Martin et al. (2017) and are listed in Table 1. The calculator requires input describing the geographic coordinates and elevation of the samples, local shielding of the sample, density of the sample, and estimation of the boulder erosion rates resulting from processes such as boulder grussification and spalling (Gosse & Phillips, 2001). These values are listed for each sample in Table S1. All the ages are minimum ages since we assume no erosion, however the rate of erosion must be low as there is a sample with an age >450 ka . We do not see evidence of a systematic boulder height-age relationship that could be a result of erosion (e.g., Behr et al., 2010;Blard et al., 2007;Garvin et al., 2005;B. Surpless & Kroeger, 2015). The age estimates are also dependent on the assumption of particular scaling models designed to estimate the long-term production rate of cosmogenic 10 Be. The 10 Be ages in Table 1 use a production rate of 4.05 ± 0.30 at/g SiO 2 /yr determined at Twin Lakes, which is located at a higher elevation than the fan surfaces here, but is within 100 km of all study sites (Balco et al., 2008;Borchers et al., 2016;Nishiizumi et al., 1989), the ERA40 atmosphere model of Uppala et al. (2005), the Lifton-VDM2016 geomagnetic database (N. Lifton, 2016), and the LSD scaling scheme (N. Lifton et al., 2014). Ages were also calculated using the time-dependent scaling model of Lal (1991) and Stone (2000), and these results are listed in Table S1. The 36 Cl boulder ages are calculated using the CRONUS calculator for 36 Cl, and are listed in Table 1, with laboratory details in Table S2.
To determine the age of a fan surface in Mason Valley using a cosmogenic depth profile, a ∼2-m-deep profile was excavated and the exposed sediment was sampled at 20, 45, 105, and 160 cm depths for 10 Be analysis. The resulting 10 Be concentrations as a function of depth were modeled using the Hidy et al. (2010) MATLAB code, which uses a Monte Carlo simulation to find the best fit of the data and resulting surface age. As soil descriptions are critical to correctly interpreting results of cosmogenic ages (e.g., Angster et al., 2019), the soil exposed in the profile was described following Birkeland (1984) and soil textural grain size was analyzed by A&L Great Lakes Laboratories, Inc.

Vertical Separation Measurements
Both normal fault slip rates and fault slip distributions require measurements of vertical separation. Fault scarp vertical separations are determined by extracting topographic profiles from lidar data. Topographic profiles are extracted approximately orthogonal to fault scarps where both hanging and footwall surfaces are similar in age and show minimal to no modification by geomorphic and anthropogenic processes unrelated to faulting (e.g., road cuts, younger alluvial fan deposition, or fluvial modification). Profiles are analyzed using a python code written for this study (see supporting information) that fits linear regressions to points selected by the user that are representative of the hanging wall, foot wall, and fault scarp surfaces (example profile in Figure 3d). To standardize measurement where hanging wall and foot wall surfaces are not parallel, the apparent vertical separation between the two surfaces is measured at the horizontal midpoint of the intersections between the regressions of the fault scarp and each of the hanging and foot walls (modified after Caskey, 1995;Rood et al., 2011).

Slip Rate Calculations
Vertical slip rates of faults are determined by dividing the probability distribution of the vertical displacement of the fault by the distribution of ages of the displaced surface using the MATLAB code of Zechar and Frankel (2009). Resulting plots are in Figure 4. In this method, the probability distribution of the age of a surface is the sum of the probability distributions of each of the samples from a single surface. The uncertainty of the vertical separation is assumed to be normally distributed, described by the mean and standard deviation of the vertical separation determined from 3 to 5 profiles extracted perpendicular to a fault scarp in a deposit of a single age. A normal distribution is used to better estimate the average fault offset from a number of measurements of fault scarps, modified unknown amounts by erosion on the hanging wall, deposition on the footwall, and the natural variability of the earthquake displacement along fault PIERCE ET AL. strike. Downdip and extensional components of slip are determined using an assumption of 55° ± 10° dipping normal fault geometries, based on gravity data, seismic reflection profiles, and the steep rangefronts throughout the study region (e.g., Cashman et al., 2009;Surpless & Kroeger, 2015). Strike-slip rates are calculated by dividing the lateral displacement by the age of the surface, and assuming a vertically dipping fault geometry. Resulting slip rates are listed in Table 2.

GPS Velocities
The Walker Lane GPS velocities presented here are from the Nevada Geodetic Laboratory MIDAS velocity solution (Blewitt et al., , 2018, accessible at http://geodesy.unr.edu/velocities/). We include velocities for all GPS stations with time series longer than 2.5 years from the semicontinuous MAGNET network operated by the Nevada Geodetic Laboratory (Blewitt et al., 2009) and neighboring continuous GPS stations. The velocities are calculated using position data collected through August 2019, presented in the NA12 reference frame , and corrected for the postseismic effects of historic ruptures in and surrounding the Walker Lane.
The MIDAS algorithm is a median trend estimator that mitigates both seasonality and step discontinuities in the times series . The resulting velocities are insensitive to the coseismic and postseismic effects of earthquakes that occurred after the midpoint of the time series , such as the July 2019 Ridgecrest, CA M w 6.4 and 7.1 sequence, but must be corrected for the postseismic effects of earthquakes that occurred prior to the middle of time series, such as the historic surface rupturing earthquakes in Central Nevada Seismic Belt and the 1993 Landers M w 7.3, and 1999 Hector Mine M w 7.0 events. We apply the viscoelastic postseismic relaxation correction from Bormann et al. (2013) that was developed using the method of Hammond et al. (2010) and the preferred western Basin and Range crustal and upper mantle viscosity model of Hammond et al. (2009). The observed and postseismic-corrected velocities are plotted GPS profiles in Figure 12 and are available in Table S4  . The postseismic-corrected velocities are shown in map view in Figure S1.

The Basins
The western domain of the Central Walker Lane is composed of the Lake Tahoe, Carson Valley, Antelope Valley, Smith Valley, Mason Valley, and Walker Lake basins. Along the western margins of each of these basins is a subparallel north-striking/east-dipping active range-bounding fault dividing a mountain range to the west from a half-graben holding the basin to the east ( Figure 2). The north-strike of these mountain ranges contrasts sharply with the northeast-strike of those of the Basin and Range to the east of this region, and the continuous high topography of the Sierra Nevada to the west. These range-bounding faults are ∼20-100 km long and together they form a rough left stepping en echelon pattern.

Geomorphic Expression of Active Faulting
The Antelope Valley rangefront is steep, rising 800 m above the valley floor, with prominent triangular facets and frequent scarps in alluvial deposits. The range-bounding fault follows an overall northwest trace for ∼23 km, and consists of <5 km long, north-striking, left-stepping fault sections ( Figure 2). Fault scarps range in height from ∼1 to 3 m up to >20 m and are present at the mouths of most drainages in Qy, Qi, and Qo deposits ( Figure 3). Nowhere along the rangefront did we find evidence of lateral displacements.  ure 3c). The faulted surface exhibits a bench of slightly lower elevation adjacent to the fault. The surface of this bench contains a thin scattered rounded cobble deposit that was likely deposited by the Walker River. Not only does the scarp itself appear to be modified by the Walker River, but also the hanging wall surface is buried by younger alluvial fan activity and fluvial deposition from the nearby Walker River, and thus this measured vertical separation is a minimum. Boulders on the Qi surface here are mostly volcanic rocks sourced from the nearby Sierra Nevada. Here, we collected six boulder samples for 36 Cl cosmogenic analysis (Table 1). Resulting ages range from ∼33 to 111 ka (mean age of 62.4 ± 23.8 ka) (Figure 3c). Dividing the average vertical separation with this age distribution results in a vertical separation rate of 0.5 +0.5 / −0.3 mm/yr ( Figure 4). Based on an assumed 55° ± 10° dip, the downdip slip rate here is 0.6 +0.6 / −0.4 mm/yr, encompassing the 0.7 mm/yr estimate that Sarmiento et al. (2011)   their trench and the interevent time. Using the same dip results in an extension rate of 0.4 +0.4 / −0.2 mm/yr (Table 2).

Fault Slip Rate
At the Artesia Fan is a series of three progressively faulted alluvial fan surfaces (Figure 5a). In Figure 5b: Qa are the active alluvial fan deposits, forming a narrow active channel cut into young Qy deposits, and on the basin floor form a small fan overtopping Qy deposits. These Qy deposits form an alluvial fan on the hanging wall of the fault as well as on the margins of the main Qa channel on the footwall, and are composed of light gray, rounded, unweathered granitic boulders in a poorly sorted matrix. Intermediate Qi deposits are only found on the footwall of the fault and are redder in color, with a smoother, less bouldery surface than Qy. The fault has an anastomosing pattern in Figure 5b that, from south to north, starts as a single strand, splits into two, then three strands, before coalescing into a single strand again. A maximum bound on the age of the Qi fan is provided by an underlying 60-80 ka tephra exposed in a road cut in the fault scarp (Stauffer, 2003;Wesnousky & Caffee, 2011). We collected 10 boulder samples for 10 Be analysis on the Qi and Qy displaced alluvial fan units and combined these with three recalculated 10 Be boulder ages at this site from Wesnousky and Caffee (2011) (Figure 5b). Six boulder samples were collected from the footwall of the intermediate unit (Qi, Figure 5b), and range in age from 18 to 66 ka (mean age of 37.0 ± 16.3 ka), and seven samples from both the hanging and footwalls of a younger aged alluvial fan unit (Qy, Figure 5b) range in age from 10 to 33 ka (mean age of 19.7 ± 8.9 ka) (Table 1). While these ages are scattered, the mean ages of the young surface are roughly half of that from the intermediate surface (Table 2). The average of four profiles measured from the lidar data ( Figure 5) show that the scarp height of the single strand that cuts the intermediate unit is 19.5 ± 4.5 m, while two profiles of each of the two scarps in the younger unit measure 3.8 ± 0.1 and 6.6 ± 0.2 m, respectively, or 10.5 ± 0.1 m combined ( Figure 5). Dividing these scarp heights by the ages of the boulder samples from each surface leads to a vertical separation rate of 0.6 +0.8 / −0.2 mm/yr for the intermediate surface and 0.5 +0.7 / −0.2 mm/yr for the young surface ( Figure 4). These two rates over different timescales generally agree, and correspond to dip parallel rates of 0.7 +1.0 / −0.5 and 0.6 +0.9 / −0.2 mm/yr for the Qi and Qy surfaces, respectively ( Table 2). All of these rates are higher than the 0.125-0.33 mm/yr rate reported by Wesnousky and Caffee (2011) for the same Qi surface, as their maximum rate is based on the assumption that the maximum measured cosmogenic age (CWL-4: ∼60 ka) is a minimum age for the faulted surface, ignoring possible inheritance. The additional ages presented here suggest that this surface is younger than this single older outlier, and thus our slip rate is greater.

Geomorphic Expression of Active Faulting
In contrast with the two previously described basins, the rangefront along the Singatse Range is more subdued, with only ∼700 m of total relief above Mason Valley. Though there are triangular facets along the rangefront, there are few places where fault scarps are preserved in Quaternary deposits ( Figure 6). The range-bounding fault extends from ∼5 km south of where the West Fork of the Walker River enters the basin northward for ∼35 km, forming a sinuous trace of both left-and right-steps between mostly north striking rangefront sections.

Fault Slip Rate
Adjacent to the Yerington Pit ( Figure 6)  which may not be sufficient to account for variations in the seismic cycle (e.g., Styron, 2019). Here, we collected a depth profile of four samples as well as four surficial boulder samples for cosmogenic 10 Be analysis to constrain the age of the faulted surface. Boulder ages sampled from the Qo surface here range from 90.5 ± 6.9 ka to 451.1 ± 37.2 ka (mean of 204.6 ± 169.9 ka), and modeling the depth profile results in an age of 173.1 +38.7 / −32.5 ka (Figure 7). The upper two samples in the depth profile do not fit the modeled curve, which could be a result of bioturbation. Were the uppermost sample discarded, the modeled age of the surface would be older. The well-developed soil, with a 10-cm thick Bt horizon, (Figure 7b), suggests a mid to late Pleistocene age, consistent with the modeled age of the depth profile (e.g., Angster et al., 2019;Birkeland, 1984). Dividing the 173.1 +38.7 / −32.5 ka age of the depth profile by the 6.5 ± 3.6 m average displacement of the surface results in a vertical slip rate of <0.05 mm/yr. This rate correlates with extension and downdip slip rates of <0.04 and < 0.06 mm/yr, respectively. As this site is situated on the eastern side of a 2-km-wide left stepover (Figure 6a), some slip may be missing from this rate. However, we do not see evidence of youthful faulting along the western fault strand.

Geomorphic Expression of Active Faulting
The character of rangefront faulting south of Mason Valley abruptly changes ∼5 km south of where the west fork of the Walker River enters the southwestern margin of the basin (Figure 2). This area is here referred to as the Pine Grove Hills, and instead of forming a rangefront composed of sharp triangular facets, the rangefront is subdued and is composed of a series of uplifted old alluvial fans and pediments. This fault zone is partly mapped as a normal fault in the USGS Quaternary fault and fold database, but new observations presented below increase the detail of mapping and show that it has a significant dextral strike-slip component.
Here, Google Earth imagery and lidar data show a discontinuous series of linear, northwest-striking faults and lineaments that extend southeast for ∼22 km. Generally, these faults have a northeast down sense of motion, and locally dextrally displace fan deposits and channels.
PIERCE ET AL.

Fault Slip Rate
Figure 8b is a Quaternary map based on a hillshade of a 0.25 m/pixel resolution structure-from-motion model ( Figure 8c) where a linear fault trace dextrally displaces Qy and Qi deposits. From north to south along this section of the fault is a small shutter ridge, a ∼33 m dextrally displaced Qi terrace riser (A' in Figures 8b and 9c), a ∼21 m dextrally deflected Qy channel (B' in Figures 8b and 9b), a ∼33 m dextrally displaced beheaded channel (B'' in Figures 8b and 9c), and a ∼20 m dextrally offset terrace riser. To show these displacements, channels A and B are backslipped 21 and 33.5 m in Figures 9b and 9c, respectively. The direction of the scarp produced by the fault changes from a subdued west-facing scarp, to a clear east facing scarp, and back to a subdued west-facing scarp along strike, further demonstrating the strike-slip nature of this fault (McCalpin, 2009;Sylvester, 1988). A maximum age of the Qi terrace risers can be estimated by the correlative ∼40-60 ka Qi surfaces measured in Antelope and Smith Valleys (Figures 3-5). Dividing the measured 18-33 m displacements by this age range results in a minimum slip rate for the Pine Grove

Walker Lake Basin
The Walker Lake basin ( Figure 10) is the largest basin of those examined and is ∼100-km-long by 20-kmwide. The western margin of the basin is bound by the Wassuk range, which hosts a prominent, active, east-dipping, normal fault, forming more than 2,200 m of vertical relief. The vertical slip rate (>0.3-0.4 mm/ yr) and paleoseismic history of this fault zone are described by Bormann et al. (2012) and B. Surpless and Kroeger (2015). Dong et al. (2014) document an active strike-slip fault in the northern part of the basin that displaces Lahontan aged shoreline deposits at a dextral slip rate of ∼1 mm/yr (Figure 10b). This strike-slip fault segment is partitioned away from the rangefront normal fault (Figure 10a). Based on variations in strike, the Wassuk rangefront fault is here divided into three sections. From north to south: a linear northwest striking section extending northwest from near Schurz, a central sinuous roughly north and northeast striking section along the western shore of Walker Lake, and a southern northwest striking linear section from Hawthorne to Whiskey Flat (Figure 10a). The central section forms a right-step in the fault system. The deepest portion of the basin, Walker Lake, is located in this right-step. This geometry of the Walker Lake fault system has been previously described as a rhomboidal pull-apart in a strike-slip system (Mann, 2007), with Walker Lake itself situated in the depocenter of this pull-apart (Link et al., 1985).
Directly west of Hawthorne ( Figure 11) is a generally NW-trending, discontinuous fault section with occasional uphill-facing scarps. This fault right laterally displaces both left and right terrace risers ∼6 m at two different drainages (Figures 11b and 11c) and a series of gullies (Figures 10c and 11d). From this area south to Lucky Boy Pass (Figure 10d), the fault forms a right-stepping pattern of north and northeast trending normal fault scarps and northwest trending linear scarps.
From Lucky Boy pass to the southeast is a linear northwest striking fault section that extends for ∼10 km (Figure 10d). The northern ∼3 km of this fault section forms NNE facing fault scarps along the rangefront. Southward from here, the fault bifurcates into a rangefront fault and a linear, basin-ward strike-slip fault trace that forms subdued uphill-facing fault scarps. The basin-ward fault section features south and north facing scarps and prominent pop-up, pressure ridge-like features of uplifted bedrock, fan, and pediment surfaces (Figure 10d). The rangefront fault to the south of the basin-ward strike-slip fault gradually bends to be more northerly striking and forms a number of east-facing scarps at canyon mouth drainages. The faulting mapped in Figure 10d is an example of strain partitioning: where a basin-ward strike-slip fault is separated from a rangefront normal fault (similar to Figure 10b and Dong et al., 2014).
The ∼30-km-long series of fault scarps extending from the southern margin of Walker Lake to Whiskey Flat (Figure 10d) are generally northwest trending. Normal scarps in north-trending bends are more numerous than those observed along the northwest-trending linear sections. The linearity of the fault traces, the alternating-facing fault scarps (Figures 10c and 11), and the several beheaded channels and lateral displacements in the northwest-striking portions of the fault zone (Figure 11), and the magnitude of scarps in the north-striking right-steps are all interpreted to be a result of northwest directed dextral faulting.

Geodetic Analysis and Kinematic Efficiency
To understand how shear accommodation in the Central Walker Lane differs from shear accommodation in the Northern and Southern Walker Lane, we need first need to constrain shear budgets spanning the length of the Walker Lane. Although many recent studies have presented geodetic budgets for individual sections of the Walker Lane (e.g., Bormann et al., 2016;Hammond et al., 2011;Z. M. Lifton et al., 2013;Pardo et al., 2012), these studies present GPS velocities determined using data from different time spans and processed using different algorithms, postseismic corrections, and reference frames, which makes it difficult to determine whether differences in geodetic shear budgets are the due to changes in the deformation field or differences in data processing methods. The most recent analysis of GPS velocities spanning the entire Walker Lane system from north to south is presented in Bennett et al. (2003), dating from a time when the field of GPS geodesy was in its relative infancy and GPS stations were sparse.
In Figure 12, we present a series of 5 GPS velocity profiles illustrating the increase in interseismic GPS velocity magnitude relative to the North America along 350-km-long and 50-km-wide transects spanning the Basin and Range, Walker Lane, and Sierra Nevada transition (see Figure 13a for profile locations and extent). The velocity profiles are oriented N39°E, perpendicular to the GPS observed orientation of maximum horizontal shear strain for the Walker Lane, to capture the maximum shear deformation budget. The profiles show a steeper velocity gradient across the Walker Lane than in the neighboring Sierra Nevada microplate to the west or slowly extending Basin and Range province to the east. In all profiles, the velocities increase smoothly in a single steep ramp across the Walker Lane, and the profiles do not show discrete steps in associated with individual faults. A smooth velocity field is an expected feature of a GPS velocity field measuring interseismic deformation above locked faults where the fault spacing is less than twice the locking depth. Shear deformation budgets for each profile are determined by subtracting the velocity of GPS stations just east of the Walker Lane in the Basin and Range, from the velocity of GPS station on the Sierra Nevada microplate to the west (horizontal blue dashed lines in Figure 12). The total shear budget across the Walker Lane decreases from ∼10 mm/yr in the south (Profile E) to ∼6 mm/yr in the north (Profile A).
The geodetic shear budgets determined from the velocity profiles shown in Figure 12 provide the basis to quantify the discrepancy between geologic and geodetic measurements of shear in the Walker Lane. We evaluate the discrepancy using the concept of kinematic efficiency, which, as defined by Hatem et al. (2017), is the ratio of total on-fault deformation to total shear. We sum the published geologic dextral fault slip rates with those determined in this study for each profile (Table 3) and divide this sum by the geodetic shear budget for that profile to calculate kinematic efficiency percentages shown in Figure 13a. These relations are also illustrated graphically by the red stepped cumulative slip rate functions in Figure 12. For simplicity, and due to both the uncertainty of fault geometry and limited contribution to northwest-oriented-shear, normal faults and transverse faults are omitted from this analysis. In general, the geologic rates reported in Table 3 are the highest rates resulting from the most recent publications. By examining shear in this way, we can make direct comparisons to the results of laboratory experiments.
Here we briefly summarize the observations of this kinematic efficiency analysis, from south to north (Figure 12). In Profile E, the Airport/Little Lake Valley, Panamint Valley, Death Valley, and Stateline faults together sum to ∼9.1 mm/yr, ∼90% of the 10 mm/yr of dextral shear observed by geodesy (Figure 12e). Profile D only crosses the White Mountains and Fish Lake Valley faults with a combined rate of 6.5 mm/yr, ∼80% of the 8 mm/yr observed geodetically (Figure 12d). This may be a high estimate as there are several rates for the Fish Lake Valley fault in Frankel et al. (2011) that are lower than the 6.1 mm/yr rate used here. In Profile C, the Wassuk rangefront fault, along with the Gumdrop Hills, Indian Head, Benton Springs, and Petrified Springs strike-slip faults, have a summed rate of 4.8, ∼60% of the 8 mm/yr observed geodetically (Figure 12c). In Profile B, across the western part of the Central Walker Lane, the Pine Grove Hills fault and the Wassuk rangefront faults are the only known dextral strike-slip faults, with a total rate of 1.6 mm/yr, only 25% of the 7 mm/yr observed in the geodetic profile (Figure 12b). Profile A crosses the Polaris, Warm Springs, Honey Lake, Pyramid Lake, and Mohawk Valley strike-slip faults with a combined rate of 3.8 mm/yr, ∼60% of the 6 mm/yr observed by geodesy (Figure 12a). All together, these kinematic efficiencies vary from ∼25% to 90%.

Summary of Evidence of Lateral Motion in the Central Walker Lane
Only along the Wassuk rangefront and Pine Grove Hills faults were we able to find dextrally displaced piercing points. In Smith Valley, the rangefront geometry of linear, optimally oriented northwest striking fault sections separated by right-steps with increased scarp heights is similar to that of the Wassuk rangefront fault, and both fit a model of fault slip across a right-step with a significant oblique dextral component (e.g., Figure 5d). The Smith Valley fault trace is also much more linear than the other normal faults in this study, suggestive of lateral motion. While we were unable to locate any offset piercing points along this fault, any offset markers may have been buried by alluvial fan activity in the 3,530 years following the most recent earthquake (Wesnousky & Caffee, 2011).
These faults provide different examples of how dextral shear is accommodated in this region. In both Smith Valley and the Walker Lake Basin, shear is accommodated in part by large right-steps along oblique slipping rangefronts. The Pine Grove Hills fault is a primarily dextral strike-slip fault with little relative vertical motion, as demonstrated by the opposite facing scarp directions, subdued topography, and comparably large dextral offsets (Figure 8b). The oblique slipping Wassuk rangefront displays evidence of both strain partitioning and bi-modal fault behavior, where a single fault can produce either lateral or vertical slip in different earthquakes (e.g., Barnhart et al., 2015). In some locations along the Wassuk range fault system, strain is partitioned between the rangefront normal fault and a separate dextral fault outboard of the rangefront (e.g., Figure 10b). In other places dextral offsets are expressed on the rangefront as uphill facing scarps and deflected streams (e.g., Figures 10c and 11d). Repetition of events that produced these strike-slip features could not have simultaneously produced the steep rangefront, implying bi-modal earthquake behavior.

Comparison of Geologic Rates to Block Models
The vertical separation rates and resulting extension and dip-parallel slip rates for the Walker Lake, Pine Grove Hills, and Mason, Smith, and Antelope valley faults are listed in Table 2 along with the block modeled rates for the same faults from Bormann et al. (2016). The extension rates for all faults except Mason Valley are within the margins of error for both the geologic and geodetic methods. In Mason Valley, the geologic rate of <0.05 mm/yr is an order of magnitude less than the geodetic estimate of ∼0.5 mm/yr. As the Singatse Range is much more subdued in topographic expression than other ranges in this region, and the rangefront generally lacks the uplifted, faulted fan surfaces common to the other faults in the region with slip rates of ∼0.5 mm/yr, it seems that the much lower geologic rate is more accurate. This discrepancy could be a result of the block model regularization parameters that favor smooth and simple slip rate solutions (Bormann et al., 2016) not allowing sufficient variability to match the geologic partitioning of slip across the faults in the region.
The geologic dextral rate of ∼1 mm/yr (Dong et al., 2014) for the Wassuk range is within the error of the 0.6 ± 0.4 mm/yr block modeled rate. The block model dextral rate of the Pine Grove Hills fault is 0.6 ± 0.5 mm/ yr, encompassing the ∼0.3-0.8 mm/yr estimate here. The geodetic right lateral strike-slip rate of 1.1 ± 0.4 mm/yr in Antelope Valley is among the fastest rates in the region, yet the fault lacks any observable geomorphic expression of dextral slip. The small faults sections on the east side of Antelope Valley with dextral displacements do not follow any major zone or structure as might be expected with a rate of this magnitude. In Smith Valley, the GPS block model predicts 0.5 ± 0.4 mm/yr of dextral strike-slip (Bormann et al., 2016). As the geologic rate was measured on the edge of a large right-bend that is nearly perpendicular to the overall trend of the rangefront fault ( Figure 5), part of the ∼0.4 mm/yr extension rate may be a result of a lateral component of motion along the fault, as predicted by the block model.
The sum of the block modeled dextral slip rates (Bormann et al., 2016) for the faults in Antelope, Smith, and Mason valleys, and the Wassuk rangefront, is 2.6 ± 1.6 mm/yr, spread across the faults in each of these basins (for the purposes of this transect Mason Valley and Pine Grove Hills are considered one fault). Geologically, we measure a total of ∼1.5 mm/yr of dextral slip between the Wassuk rangefront and our estimate for the Pine Grove Hills, which could be as much as ∼2.0 mm/yr if the extension measured in Smith Valley is attributed to dextral shear along the rangefront. In sum, there still remains some discrepancy between the predicted dextral rates from the block models and the geologic rates, however this discrepancy is less than previously thought (e.g., Wesnousky et al., 2012).   Table 3). There is good agreement between geodetic and geologic observations in the Southern Walker Lane, but less so in the Central and Northern Walker Lane. (b) Model of structural evolution of the Sierra Nevada Basin and Range transition zone and the Central Walker Lane. Initially the edge of the Sierra Nevada batholith resisted deformation and dextral shear followed a right step at the Mina Deflection. Continued deformation rotated and strengthened the faults of the Mina Deflection.
In the future, a through-going strike-slip fault may form as the Mina Deflection is no longer able to accommodate the right-step.

How Shear is Accommodated Without Major Strike-Slip Faults
Laboratory shear models of fault evolution and many field studies show that a large portion (often >20%) of total deformation is accommodated as off-fault deformation (e.g., Hatem et al., 2017;Herbert et al., 2014;Oskin et al., 2012;Personius et al., 2017). However, off-fault deformation is usually considered on a singular fault basis, not at a fault system scale (e.g., R. D. Gold et al., 2015). By considering the kinematic efficiency in profiles across the Walker Lane, as described above, we can estimate off-fault deformation across the entire fault system. The laboratory models of Hatem et al. (2017) show that even structurally mature through-going faults are only ∼80% efficient, with ∼20% of total shear accommodated by off-fault deformation, while immature faults, with limited total slip, can be less than 40% efficient, with >60% of the total shear accommodated off of faults. In the Walker Lane kinematic efficiencies vary from as little as ∼25% in the Central Walker Lane (profile b in Figures 12 and 13a) to ~60% in the Northern and ∼80%-90% in the Southern Walker Lane (profiles A, D, and E in Figures 12 and 13a).
The "off-fault" shear in the Central Walker Lane is likely accommodated by a combination of broad co-seismic warping, displacements that are not preserved in the paleoseismic record, and rotations of crustal blocks. Geologically accounting for near surface coseismic off-fault deformation generally requires detailed pre and postearthquake observations of a fault system, and is therefore difficult to account for in the paleoseismic record (e.g., R. D. Gold et al., 2015;Herbert et al., 2014;Milliner et al., 2015;Oskin et al., 2012;Personius et al., 2017). For example, Oskin et al. (2012) showed that during the 2010 M7.2 El Mayor-Cucapah earthquake multiple faults were linked by "enigmatic zones of distributed deformation." These zones did not rupture the surface, yet they accommodated a significant amount of strain.
Numerous small and moderate earthquakes (<M6) occur in the Walker Lane (e.g., Hatch et al., 2019;Ruhl et al., 2016), and while many of these such as the 2016 M5.   (Barnhart et al., 2019;Bell et al., 1999;Koehler et al., 2021;Rockwell, 2000;Rubin & Sieh, 1997). Subsequent studies have shown that many of the respective seismogenic faults have recurrence times on the order of ∼3-15 ka, and that, if present, scarps from prior events are heavily degraded Rockwell, 2000;Rubin & Sieh, 1997). Many of the faults that accommodated these ruptures are neither major through-going structures nor range-bounding faults Jobe et al., 2020;Koehler et al., 2021). We suggest that future M7 strike-slip events in the Central Walker Lane may be similar to these events, and might occur across a series of previously recognized and unrecognized structures.
As the lowest measured dextral slip rates in Figure 13a are ∼0.4 mm/yr, this rate may represent the threshold of preservation for multiple event offsets on strike-slip faults in the Walker Lane. Below this slip rate, geomorphic processes likely outpace the rate at which faults displace landforms, so offsets are destroyed faster than displacements can accumulate. Thus, evidence of low slip rate faults might not be preserved in the landscape.
Both GPS block models (Bormann et al., 2016) and paleomagnetic studies (Carlson, 2017;P. H. Cashman & Fontaine, 2000) agree that clockwise vertical-axis block rotations play a significant role in strain accommodation in the Carson domain and Mina Deflection. However, the role of block rotations remains unclear in the western domain of the Central Walker Lane. GPS block models predict low rate vertical-axis clockwise block rotations of 1°/Myr that act in concert with oblique faulting to accommodate observed shear strain (Bormann et al., 2016). However, limited paleomagnetic measurements in the Singatse Range, Wellington Hills, and Buckskin Range of the western domain record statistically insignificant amounts of rotation, which suggests that these blocks may be purely translating (Carlson, 2017).
Thus, while not all of the shear across the Central Walker Lane can be accounted for by summing the slip rates of known strike-and oblique-slip faults, and some "on-fault" shear may not be preserved, the sum of the lateral slip rates of the individual faults need not necessarily equal the total geodetic shear across a fault system, as distributed shear, block rotations, and complex fault relationships can accommodate a significant portion of "off-fault" shear, especially in immature systems such as the Walker Lane. This presents a problem for fault based seismic hazard models, as we do not know exactly which faults might accommodate this shear in future earthquakes. In this context, in regions with complex distributed fault patterns it may be better to use geodetic strain observations to forecast future earthquake probabilities (e.g., Lamb et al., 2018).
The Wassuk range, along with the rest of the ranges to the west, form the easternmost margin of the Sierra Nevada batholith, which is a significant structural boundary (black dotted line in Figure 13: Buer et al., 2009;Long, 2019). The westward stratigraphic tilt of the ranges in the Central Walker Lane decreases to the northwest from 60° in the Wassuk and Singatse ranges, to <20° in the Carson Range, and <5° in the Sierra Nevada (B. E. Surpless et al., 2002). Likewise, the southeasterly Singatse and Wassuk ranges have considerably higher amounts of total extension (>150%) than the ranges to the west (B. E. Surpless et al., 2002), which has been cited as evidence of the progressive northwestward encroachment of faulting into the Sierra Nevada block.
The faults in the western part of the Central Walker Lane form a left-stepping en-echelon pattern that falls roughly on trend with the Death Valley-Fish Lake Valley fault (Figures 1 and 13), the longest and highest slip rate fault in the Walker Lane (Frankel et al., 2007). Interrupting the northernmost part of the Fish Lake Valley fault is the Mina Deflection (e.g., Nagorsen-Rinke et al., 2013;Wesnousky, 2005), which accommodates a right step-over to the faults of the eastern part of the Central Walker Lane (Figures 1  and 13). Based on both paleomagnetic data (Petronis et al., 2009) and the nonoptimal, nearly east-west orientations of the sinistral faults in the Mina Deflection (Wesnousky, 2005), the crustal blocks here have accommodated clockwise vertical axis rotations of ∼20°-30°. As faults rotate, their orientations become progressively less favorable to slip, and based on Mohr Coloumb faulting theory approach a state of lockup (e.g., Hill & Thatcher, 1992;Scholz, 2019;Sibson, 1985;Thatcher & Hill, 1991). This lockup angle is ∼2ϴ, where ϴ is the angle between the maximum compressive stress, σ 1 , and the initial fault plane. Whether we assume that the faults in the Mina Deflection initially formed at ϴ = ∼30° from the roughly north-oriented σ 1 (e.g., Bellier & Zoback, 1995), based on Mohr Coulomb theory and Byerlee's law, or if they formed at ϴ = ∼45°, as Reidel shears perpendicular to the primary northwest oriented dextral faults (Wesnousky, 2005), their current 080°-085° orientation is close to the lockup angle, 2ϴ. At this stage of rotation, modeling and field studies show that a new set of faults may form (e.g., Olive & Behn, 2014;Richardson & Seedorff, 2017).
In Figure 13b, we present a model of structural evolution for the Central Walker Lane that synthesizes (1) the westward encroachment of faulting in the Central Walker Lane, (2) the rotation of the Mina Deflection, and (3) the lack of through-going strike-slip faults in the western part of the Central Walker Lane. Initially, a rigid Sierra Nevada batholith resisted deformation, and dextral faulting followed the edge of the block with a large right step at the Mina Deflection. As shear continues to today, the sinistral faults accommodating this step-over have rotated and approached the theoretical lock-up angle, while simultaneously the eastern margin of the Sierra Nevada block has been fractured by faulting. In the final pane, the faults in the Mina Deflection have further rotated, "shutting off" the right step, and shear follows a more simple, linear, through-going strike-slip fault. This model provides an example of how a major stepover in a strike-slip system can be transient in time. Here a step-over is accommodated by block rotations, but as those blocks rotate, they become less able to accommodate the step-over, and faults instead break a new path.

Conclusions
We present new vertical slip rates of 0.6 +0.7 / −0.3 , 0.7 +1.0 / −0.4 , and <0.05 mm/yr for the Antelope, Smith, and Mason valley range-bounding normal faults, respectively. We estimate the dextral slip rate of the Pine Grove Hills fault to be 0.3-0.8 mm/yr, and we report new sites along the Wassuk Rangefront that accommodate dextral slip. The southern Wassuk and Smith Valley rangefronts in the Central Walker Lane accommodate previously unrecognized oblique slip, whereas in the Pine Grove Hills and northern Wassuks, discrete strikeslip faults accommodate slip. We present a series of geodetic profiles spanning the Walker Lane, showing that total shear rates decrease from 9 to 10 mm/yr at the latitude of Death Valley to ∼6 mm/yr at the latitude of Honey Lake. Our new data show that the kinematic efficiency (here defined as the ratio of summed geologic strike-slip rates to the regional geodetic shear budget) of the Central Walker Lane is only 25%, significantly lower than the ~60% and ∼90% observed in the Northern and Southern Walker Lane, respectively. The remaining unaccounted shear is presumed to be accommodated by off fault deformation, such as a combination of block rotations and faulting that are not preserved in the paleoseismic record. We present a model for the structural evolution of a crustal scale right-step in a dextral system that explains the present configuration of faulting in the Central Walker Lane. The model provides an explanation of how step overs might form in response to preexisting structural conditions but be bypassed as shear accumulates. The Central Walker Lane is an example of a region where the geologic rates of known structures underestimate the deformation rates derived from geodesy, and this might be expected in other regions characterized by broad zones of distributed deformation. 0919759 (PRIME Lab), by USGS grant G15AP00088, and by the Leverhulme Trust. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement. Thanks to two anonymous reviewers and an associate editor for very helpful reviews, which greatly improved this manuscript. We thank Paula Figueiredo and Sarah Hammer for assistance in processing and calculating TCN ages, and the National Center for Airborne Laser Mapping (NCALM) for collecting lidar data. Thanks to Steve Angster, Tabor Reedy, Chad Carlson, Noah Abramson, Annie Kell, Colin Chupik, Bill Hammond, Ken Adams, and Rachel Hatch for assisting in the field, lab, and numerous enlightening discussions. Thanks to the 2017 UNR Neotectonics class for assistance in the field. Thanks to Rich Koehler for editing an early version of this manuscript. This is Center for Neotectonics Studies contribution number 78.