The southern Sierra Nevada foothills, central California (USA), expose a fossil pre–40 Ma bedrock pediment which we call the southern Sierra Nevada pediment. We document this landscape with multiple types of data, and also report new apatite 4He/3He, (U-Th)/He, and zircon (U-Th)/He data from the pediment that significantly expand the spatial extent of southern Sierra low-temperature thermochronology data westward into the foothills. Applying recently published thermal modeling software for thermochronologic data, which uses a transdimensional Bayesian Monte Carlo Markov chain statistical approach, we tightly constrain the thermal history of the southern Sierra Nevada pediment. Integrating this thermal history with numerous previously published data sets from across the southern Sierra, we present a chronology of tectonic and landscape evolution of the southern Sierra Nevada. For the first time we cover the entire width of the range, integrate the numerous published data sets into a single coherent geologic story, and link each phase of this story to a potential mechanism.
Modeling results are consistent with a three-phase cooling history for the southern Sierra Nevada pediment. Rapid exhumation ca. 95–85 Ma resulted in cooling to between 55 °C and 100 °C. Following this, slow cooling to surface conditions occurred from 85 Ma to 40 Ma at rates consistent with those estimated for the axial southern Sierra during the same time period by previous studies. Little if any additional cooling occurred post–40 Ma. We hypothesize that a thin sedimentary cover protected the 40 Ma bedrock landscape through much of the last 40 m.y., and that this cover eroded away post–10 Ma, re-exhuming the southern Sierra Nevada pediment as a fossil pre–40 Ma landscape. Each of these three phases of cooling links to a distinct tectonic or geomorphic regime, including the profound rapid exhumation of the southern Sierra Nevada–Mojave segment of the Cretaceous arc due to subduction of a large oceanic plateau, the formation of the low-relief landscape of the high-elevation areas of the southern Sierra Nevada with more limited tectonic forcing, and Eocene activity on the Western Sierra Fault System.
Basement outcrops along the boundary between the southwestern Sierra Nevada foothills and the San Joaquin Valley (central California, USA) expose a bedrock pediment landscape that we refer to as the southern Sierra Nevada pediment (SSNP). We will first introduce the long history of studies using low-temperature thermochronologic data to constrain the evolution of the southern Sierra Nevada, and then describe the SSNP by presenting field, geochemical, and mineralogical data. Next, we will use these data to constrain a chronology of landscape evolution and tectonic activity along the SSNP. Finally, we will interpret this chronology within the broader context of the southern Sierra Nevada–Great Valley system and discuss its implications for regional tectonics and landscape evolution.
The regional additions to basement thermochronologic data from the southern Sierra that we present here improve our understanding of the post-magmatic evolution of the southern Sierran arc. These data include new bulk apatite (U-Th)/He data (Ap-He), apatite 4He/3He data (Ap-4He/3He), and zircon (U-Th)/He data (Z-He), all from locations significantly farther west than those of any previously published data from this part of the mountain range (Fig. 1). This spatial expansion of basement thermochronometric data bears significantly on the debate in the literature about the geomorphic evolution of the southern Sierra Nevada (House et al., 1998, 2001; Clark et al., 2005; McPhillips and Brandon, 2012; Wakabayashi and Sawyer, 2001; Wakabayashi, 2013, 2015) and, more importantly, on the assumptions that underlie the key arguments in these studies. In the context of the large body of research regarding the topographic evolution of the southern Sierra and recent constraints on Eocene uplift (Sousa et al., 2016), we piece together a chronology of tectonic and landscape evolution of the southern Sierra Nevada. Furthermore, we present the first application of the (U-Th)/He chronometer to the TiO2 mineral anatase.
The SSNP runs ∼150 km along the western edge of the southern Sierran foothills from near 36°N at Fountain Springs, California, in the south to near 37° N at Friant, California, in the north (Fig. 1). Along the pediment, bedrock lithology consists of plutonic rocks of the composite Sierra Nevada batholith as well as pre-batholithic wall rocks. Locally the batholith consists of Early Cretaceous plutonic rocks emplaced ca. 115 ± 10 Ma (Chen and Moore, 1982; Saleeby and Sharp, 1980; Lackey et al., 2005; Clemens-Knott and Saleeby, 1999) at pressures of 3–4 kb (Ague and Brimhall, 1988; Ague, 1997; Nadin et al., 2016). Secondary to plutonic rocks are pre-batholithic wall rocks of the Kings-Kaweah ophiolite belt that runs along nearly the entire length of the SSNP; this belt consists of the Paleozoic Kings River ophiolite and Kaweah serpentinite mélange and encloses ophiolitic blocks and infolds of nonconformably overlying upper Paleozoic–lower Mesozoic slaty marine strata (Fig. 2; Saleeby and Sharp, 1980; Saleeby, 2011).
North of the SSNP, a section of Eocene and younger deposits known as the Superjacent Series nonconformably overlies Sierran basement (Bateman and Wahrhaftig, 1966). The southernmost outcrops of the Eocene Ione Formation (Fig. 1) occur at the northern end of our study area near Friant (Lindgren, 1911; Bates, 1945; Creely and Force, 2007; Palmer, 1978; Palmer and Merrill, 1982). At its type locality near Ione, California, in the northern Sierra Nevada, the Ione Formation is of middle Eocene age, based on a limited molluscan fauna as well as stratigraphic correlations to the Domengine Formation in the Coast Ranges and Great Valley subsurface and to the auriferous gravels of the northern Sierra Nevada (summarized in Creely and Force, 2007). The Ione Formation in our study area is the southernmost extent of the Ione and correlates to the non-marine facies of the Ione at its type location (Palmer and Merrill, 1982; Creely and Force, 2007). These outcrops (Fig. 3B) sit directly on the 114 Ma tonalite of Blue Canyon (Busacca, 1982; Bateman et al., 1983), which is locally deeply weathered beneath the Eocene nonconformity (Fig. 3D).
The southern terminus of the pediment, 150 km to the south, abuts the northernmost edge of the Kern Arch, a crescent-shaped active uplift along the boundary between the San Joaquin Valley and the southern Sierran foothills (Cecil et al., 2014; Fig. 1). Analogous to the stratigraphic relationship at the northern end of the pediment, Cenozoic strata of the Kern Arch are Eocene and younger, with the basal Walker Formation, containing a 40.1 ± 0.3 Ma tuff, deposited nonconformably on deeply weathered Sierran basement (Saleeby et al., 2016; Fig. 3A and 3C).
Along the western edge of the SSNP, soils and sediments of the eastern San Joaquin Valley shallowly cover low-relief bedrock outcrops, with soil depths on the order of meters to tens of meters (Sousa et al., 2013; Saleeby et al., 2013b; this study). This area hosts widespread agriculture, which makes detailed geological observations difficult. Nonetheless, field and remote sensing observations of bedrock tors interspersed amongst orchards, as well as shallow depths to basement in local water wells, confirm that this boundary is generally a low-relief bedrock landscape (this study). A few kilometers farther west, in the San Joaquin Valley subsurface, Late Cretaceous to Eocene sedimentary rocks overlie Sierran basement (Reid, 1988). East-west–trending channels with hundreds of meters of relief and a deeply weathered zone tens of meters thick mark this sub–Upper Cretaceous basement nonconformity in the San Joaquin Valley subsurface (Reid, 1988).
East of the SSNP, the southern Sierra rises rapidly to ∼2000 m elevation across a series of topographic steps. Sousa et al. (2016) showed that one of these steps is an eroded fault scarp of the Western Sierra Fault System (WSFS) and posited that the rest of the system was also active ca. 45–40 Ma (Fig. 1).
Southern Sierra Nevada Pediment
Prior to this study, much of the research in the southern Sierra Nevada (western) foothills focused on pre- and syn-batholithic petrology, geochemistry, and tectonics (e.g., Saleeby and Sharp, 1980; Clemens-Knott and Saleeby, 1999; Saleeby, 2011). However, some studies have considered the geomorphology of the southern Sierran foothills (Hake, 1928; Wahrhaftig, 1965; Saleeby and Foster, 2004; Pelletier, 2007; Figueroa and Knott, 2010). While Figueroa and Knott (2010) and Pelletier (2007) each focused on much larger areas than the SSNP, Saleeby and Foster (2004) did focus on this area. They interpreted this segment of the southern Sierran foothills as dominated by steep faceted topography buried by active sedimentation in the eastern San Joaquin Valley. This description does not bear out fully. Rather than steep topography being actively buried, the topographic features within this landscape commonly rise from a low-relief peneplain which is only shallowly, if at all, covered by sediments (Sousa et al., 2013; Saleeby et al., 2013; this study). Saleeby and Foster (2004) contended that the geomorphic differences between this and other segments of the Sierran foothills result from the surficial response to the actively foundering mantle lithospheric phenomenon known as the Isabella anomaly, which lies beneath this segment of the southern Sierra foothills (e.g., Zandt et al., 2004). The findings of this current study, as well as more recent modeling work on the surficial effects of mantle lithospheric dynamics beneath the SSNP, support this interpretation (Le Pourhiet el al., 2006; Saleeby et al., 2012, 2013b).
Western Sierran Slope
A fundamental topographic characteristic of the Sierra Nevada is the fact that in the north, the western slope is a continuous ramp, while the southern Sierra rapidly attains an elevation of 2000 m across a set of roughly range-parallel topographic steps. An early study by Hake (1928) described a set of these steps running from the San Joaquin to Kern Rivers as intra-batholithic west-down normal fault scarps (Fig. 1). However, workers neglected this idea in the literature after Wahrhaftig (1965) dismissed it based on flawed geologic interpretations and with very little mention of Hake’s (1928) observations. These erroneous interpretations include incorrect correlation of units across topographic steps (Huber, 1981) as well as the detail that the allegedly unfaulted unit (the ca. 10 Ma Kennedy Table Mountain flow) was too young to test for the existence of pre–10 Ma Cenozoic faults. Without addressing the descriptions presented by Hake (1928), Wahrhaftig argued that the steps of the southern Sierra are of a purely erosional origin. Jessup et al. (2011) tested this conclusion by measuring cosmogenic erosion rates on step treads versus risers and found their data to be generally inconsistent with Wahrhaftig’s interpretation. Recently, Sousa et al. (2016) used thermochronometric data from the main trunk and north fork of the Kings River to show that at least one of the steps described by Hake (1928) is an eroded fault scarp. This fault accommodated kilometer-scale west-down displacement in Eocene time and is a part of the WSFS. In this context, the swath of Sierran basement that composes the SSNP is the hanging wall of the WSFS. Figure 1 shows that with the exception of a few samples presented in Sousa et al. (2016), all of the previous Ap-He studies in the southern Sierra are entirely east of the WSFS.
Southern Sierra Region
Although many models exist for the antiquity and evolution of southern Sierra Nevada topography, little consensus has emerged in the literature regarding the timing of generation of the high elevations and large-relief canyons that compose the modern southern Sierra.
The assumption of Cenozoic rigid-block down-to-the-west tilt of the Sierra Nevada mountain range underpins the analysis presented in several previous studies. Some of these studies explicitly stated this assumption, and some used its implications to extrapolate geologic data over long distances and argue for a late Cenozoic origin of most of the present-day topography, particularly north of the Kings River canyon (Huber, 1981; Unruh, 1991; McPhillips and Brandon, 2012). Wakabayashi and Sawyer (2001) used long-distance projection of tilted volcanic units to argue for only a minor departure from rigidity due to east-down faulting near the Sierra crest. They also extended this rigid-block idea westward into the Great Valley, where sedimentation is balanced with erosion of the Sierra uplands during rigid west tilting (Wakabayashi and Sawyer, 2001).
On the other hand, the correlation of Ap-He ages with the location of major river canyons along two horizontal transects from the axial Sierra supports a Late Cretaceous antiquity of the large-amplitude, long-wavelength relief pattern common to these river canyons (House et al., 1998, 2001; Braun, 2002a, 2002b). Furthermore, vertical transects of Ap-He data from the southern Sierra show a consistent age-elevation slope of 40–60 m/m.y. and lack clear inflections that would record canyon-incising events. This implies that the high-elevation, low-relief interfluvial plateaus mimic the landscape that developed in the Late Cretaceous and was slowly exhumed at roughly this same rate until 40 Ma or later (Clark et al., 2005; Mahéo et al., 2009; House et al., 1997, 2001). Together these interpretations imply that low-relief highlands and high-relief canyons were both part of the Late Cretaceous landscape. In this view, much of the form of the modern Sierran landscape mimics regional morphology that was established in Late Cretaceous (e.g., House et al., 1998), at which time the Sierra Nevada mountains were the western flank of a high-elevation plateau referred to as the Nevadaplano (DeCelles, 2004).
In contrast, Stock et al. (2004) identified a pulse of late Cenozoic river incision in the southern Sierra using cosmogenic radionuclide burial dates from vertical transects of quartz-bearing sediments deposited on abandoned fluvial-cut terraces in carbonate caves. These data resolve late Pliocene to Pleistocene incision of the lowest ∼20% of total relief of several central to southern Sierra river canyons (e.g., 400 m in Kings Canyon). Stock et al. (2004) pointed out that they do not constrain the age of the upper 80% of relief (e.g., 1600 m in Kings Canyon).
Clark et al. (2005) identified two knickpoints in stream long profiles of the main trunks and tributaries of the Kings and Kern Rivers and argued that these knickpoints correspond to two pulses of incision responsible for most of the relief in these canyons. The authors asserted that these events must have post-dated the youngest Ap-He age on the Kings River (ca. 32 Ma). Pelletier (2007) used a numerical model to test different bedrock erosion models in the southern Sierra, and the results of his preferred model indicate that the southern Sierra Nevada experienced range-wide surface uplift in the latest Cretaceous and late Miocene.
McPhillips and Brandon (2012) integrated published Ap-He and apatite fission-track thermochronometry and igneous geobarometric data into a numerical landscape evolution model encompassing much of the modern Sierra. Their preferred model finds onset of range-wide uplift and incision at ca. 30–10 Ma.
Studies in the western foothills and eastern San Joaquin Valley subsurface reported direct measurements of Late Cretaceous and Paleogene paleo-relief. This includes a minimum of 500 m of paleo-relief in the Kaweah River drainage near the Sierra–Great Valley transition based upon interpretation of Ap-He data and bedrock pediment geomorphology (Saleeby et al., 2013b; Sousa et al., 2013, 2014; this study). Reid (1988) measured the same scale (500 m) of relief on the Late Cretaceous basement nonconformity in the eastern San Joaquin Valley subsurface.
The recent documentation of normal faulting along the western slope of the southern Sierra (Sousa et al., 2016) undermines the fundamental assumption of down-to-the-west rigid tilt included in many of these previous studies. This assumption is most critical in the western foothills, where data have been lacking and long-distance extrapolation using the rigid-block model has been necessary. By filling this gap in the basement thermochronometric data along the SSNP, we obviate the need for such an assumption and test the new model put forth by Sousa et al. (2016) for Eocene faulting, extension, and uplift.
In summary, the findings of previous studies clearly require a polyphase evolution of southern Sierra topography, with distinct topographic patterns linked to specific periods of tectonic activity and erosion. This includes large-relief river canyons dating back to Late Cretaceous time, and two Cenozoic phases of uplift and incision of these canyons in Eocene (ca. 45–40 Ma; Sousa et al., 2016) and Plio-Pleistocene time (Stock et al., 2004, 2005).
THE BEDROCK PEDIMENT
Description of Pediment Morphology
In contrast to other types of pediments that form due to differences in erodibility caused by lithologic or structural boundaries, bedrock pediments form within monolithologic areas (Oberlander, 1974; Twidale, 1981; Dohrenwend and Parsons, 2009). This type of morphology remains unexplained by theory, but modeling efforts to understand bedrock pediment formation agree in the requirement of an extended period (roughly 106–107 yr) of erosion and tectonic quiescence (e.g., Pelletier, 2010; Strudley et al., 2006).
The principal components of a bedrock pediment are a low-relief peneplain, hillslopes rising from the peneplain, and, most critically, the piedmont angle where slope changes rapidly from peneplain to hillslope without any structural or lithologic boundary (Oberlander, 1974; Twidale, 1981; Pelletier, 2010; Strudley et al., 2006). We define the SSNP as the bedrock landscape exposed along the transition from the San Joaquin Valley to the Sierra Nevada foothills, which consists of these three morphologic components: (1) low-relief bedrock outcrops within the peneplain; (2) bedrock hillslopes rising from the peneplain by meters to hundreds of meters; and (3) the transition between these two zones where slope rapidly changes, called the piedmont angle (Oberlander, 1974; Twidale, 1981; Pelletier, 2010, Strudley et al., 2006). We document these elements of the landscape using multiple methods. Where access was possible, we made field observations which are the basis for the mapping shown in Figure 1. We complemented field work with aerial and satellite images, hillshade models derived from a 10 m U.S. Geological Survey digital elevation model, and published geologic maps (Matthews and Burnett, 1965; Clemens-Knott, 2011; Macdonald, 1941; Saleeby and Sharp, 1980; Saleeby, 2011; Busacca, 1982; Bateman et al., 1983).
Depth-to-basement data from shallow water wells allows us to extend our mapping of the SSNP beyond the accessible exposures and into the subsurface west of the foothills-to-basin transition (Fig. 2). We averaged individual water well data over 1 mi2 sections and compiled the data into five cross sections covering a large portion of our study area; Figure 1 shows the locations of these cross sections, which are included in the Supplemental File1. These cross sections generally support our field based observations of the SSNP, documenting areas where the floors of small valleys along the foothills-to-basin transition are low-relief bedrock landscapes covered by only tens of meters of regolith (e.g., near the town of Orange Cove, California, and in the valley of Cottonwood Creek; cross sections B and C, respectively; see the Supplemental File [footnote 1]).
A good example of this morphology is immediately upstream of Terminus Dam along the Kaweah River. At an elevation of ∼210 m above sea level near the Horse Creek Campground (Fig. 4), the current channel of the Kaweah River opens onto a bedrock peneplain averaging 500–1000 m wide. Seasonally and in wet years this area floods, but low water levels in Lake Kaweah (e.g., during drought years) expose the low-relief bedrock peneplain. There are abundant low-relief bedrock outcrops across the pediment that show that the alluvium is shallow. Around the edge of the pediment, a transition to hillslope is exposed, where without any lithologic or structural boundary the bedrock landscape rises over 500 m to local peaks.
Mineralogical and Paleosol Occurrences on the SSNP
In the southern portion of the SSNP where bedrock lithology locally includes Kaweah serpentinite mélange of the Kings-Kaweah ophiolite belt, there are several mineralogical and paleosol occurrences that are important indicators of the pre–40 Ma landscape. Specifically, these include mineral occurrences distinctive of nickel laterites formed by chemical weathering of serpentinites (e.g., Vasconcelos and Singh, 1996; Eggleton et al., 2011). At Chrysoprase Hill, Venice Hills, and Smith Mountain (Fig. 5), nickel-rich chalcedony (the gemstone chrysoprase; Fig. 6) occurs in conjunction with deeply weathered and silicified bedrock that outcrops as an erosion-resistant ferruginous silcrete (Fig. 6). At Chrysoprase Hill and the Venice Hills there are also occurrences of hydrous Ni-Mg–rich silicates (garnierite), a nickel ore common to lateritically weathered ultramafic rocks (e.g., Thorne et al., 2012). Prior to this study, mentions of these occurrences in the literature appeared only in bulletins and reports of the mineral resources of California (e.g., Goodwin, 1958; Pemberton, 1983) and popular mention of chrysoprase as an economic gemstone mined along the SSNP for several decades in the late nineteenth and early twentieth centuries (New York Times, 1902). Together with the other data presented in this paper, these minerals and paleosols support our interpretation that the modern SSNP is a fossil pre–40 Ma landscape.
Geologic mapping of plutons and ophiolitic wall rocks exposed along the SSNP (Fig. 2; Saleeby and Sharp, 1980; Clemens-Knott and Saleeby, 1999; Saleeby, 2011, Saleeby et al., 2013a), in conjunction with our geomorphic mapping of the pediment surface, shows that the area lacks transverse faults, indicating structural continuity along this swath of Sierran basement. Accordingly, we conclude that the bedrock exposed at the sub-Eocene nonconformities near Fountain Springs and Friant represents two ends of a single strip of basement that composes the SSNP. We interpret the bedrock pediment geomorphology and the distinctive lateritically weathered paleosols and mineralogical occurrences along the SSNP to be remnant elements of a pre–40 Ma (sub–Ione and Walker) landscape.
Sampling Approach and Methods
Samples come from outcrops of Early Cretaceous plutonic rocks of the Sierra Nevada batholith along the westernmost bedrock exposures of the Sierran foothills from the towns of Fountain Springs to Friant (Fig. 1). After crushing, sieving, and standard heavy mineral separation, apatite and zircon grains from each sample were selected with a stereoscopic microscope for analysis. Grains were selected for euhedral habit and checked to exclude any grains with birefringent inclusions (examined with cross-polarized light and immersed in ethanol). The dimensions of each grain were then measured and recorded. For each sample, four to seven individual grains were first analyzed for Ap-He and Z-He age determination. Helium was measured with a Pfeiffer Prisma quadrupole mass spectrometer. After standard mineral digestions, parent concentrations were measured via isotope dilution on an Agilent 7500 inductively coupled plasma–mass spectrometer (ICP-MS) (e.g., Farley, 2002). A corrected age was calculated for each grain using the alpha-ejection correction factor (Ft) based on the measured grain dimensions (after Farley et al., 1996). Table 1 shows average bulk Ap-He and Z-He ages.
For the samples chosen for 4He/3He analysis (samples 11SS1 and 11SS6), additional apatite grains were proton irradiated to make a uniform distribution of 3He (Shuster and Farley, 2004, 2005). Individual grains were then picked following the same criteria as for bulk age determination, with particular attention paid to the absence of birefringent inclusions and complete euhedral morphology. Each individual grain was step-wise degassed using a halogen lamp as heat source (Farley et al., 1999). 4He and 3He were measured at each degassing step using a GV Instruments SFT sector field mass spectrometer at Caltech. Ap-4He/3He data are included in the Supplemental File [see Footnote 1].
(U-Th)/He and Ap-4He/3He Data
A single-grain (U-Th)/He age is generally compatible with a diversity of thermal histories. A more restricted range of thermal histories is constrained by combining multiple bulk ages from different minerals like apatite and zircon (e.g., Reiners et al., 2000), multiple grains with variations in effective U concentration (eU = U + 0.235Th) (e.g., Flowers et al., 2009), or by measuring single grain 4He rim-to-core concentration profiles via 4He/3He studies (Shuster and Farley, 2005). Variations in radiation damage result in closure temperatures that vary with eU, and different time-temperature (t-T) paths result in significantly different 4He concentration profiles based on the time-integrated balance between alpha-particle in-growth and loss by both ejection and diffusion. The Ap-4He/3He method allows us to mine the 4He rim-to-core concentration profile (Shuster and Farley, 2004, 2005), and subsequent thermal modeling allows us to constrain t-T paths.
Sample 11SS6, near Friant
At the northern end of the study area near Friant, the southernmost outcrops of the Eocene Ione Formation overlie Sierran basement at an elevation of 165 m. At this location, the SSNP coincides with the sub-Ione nonconformity. Where this nonconformity crops out, bedrock is deeply weathered and nearly unrecognizable as a plutonic rock (Fig. 3D). Basement at this location is the 114 Ma tonalite of Blue Canyon (Busacca, 1982; Bateman et al., 1983). The mean zircon (U-Th)/He age from sample 11SS6 from this location is 97 ± 5 Ma (1 standard error [s.e.], n = 4), and the mean apatite (U-Th)/He age is 92 ± 4 Ma (1 s.e., n = 7) with eU ranging from 21 ppm to 66 ppm. It is worth noting that to our knowledge, sample 11SS6 has yielded the oldest Ap-He age from the Sierra Nevada batholith. Ap-4He/3He data from this sample are presented later in the paper together with thermal modeling results.
Sample 11SS1, near Fountain Springs
Our southernmost sample (11SS1) is from near Fountain Springs at an elevation of 290 m. Bedrock outcrops at the sample location include meter- to 10-m-scale corestones eroding out of the landscape, a distinct element of deeply weathered granitic rocks (Fig. 3C; e.g., Shaw, 1997). Sample 11SS1 is an altered felsic plutonic rock (Fig. 7). The nearest published U-Pb zircon age is from a 102 Ma quartz diorite a few kilometers to the east (Lackey et al., 2005; Saleeby and Sharp, 1980). A few kilometers south of this location along the White River, a 40.1 ± 0.3 Ma tuff sits on deeply weathered basement (Laser Ablation ICP-MS zircon U-Pb from Saleeby et al. ). The mean Z-He age from sample 11SS1 is 85 ± 5 Ma (1 s.e., n = 4), and the mean Ap-He age is 66 ± 4 Ma (1 s.e., n = 7), with eU ranging from 13 ppm to 38 ppm. Ap-4He/3He data from this sample are presented later in the paper together with thermal modeling results. The bulk anatase (U-Th)/He age from this sample is 97 ± 13 Ma.
The Horizontal Transect
We report bulk (U-Th)/He analyses from eight additional samples along the SSNP between Fountain Springs in the south and Friant in the north (Fig. 8). The samples were all taken from plutonic outcrops near the western edge of the southern Sierra foothills. Some samples are from isolated bedrock outcrops interspersed amongst shallow soils of the San Joaquin Valley, and others are slightly farther east in the Sierran foothills. Widespread agricultural land use in this area commonly masks the foothills-to-basin transition, but in several locations, we sampled from isolated bedrock outcrops scattered amongst orchards. Published U-Pb zircon ages along this transect are generally 115 ± 10 Ma and range from 102 Ma to 125 Ma (Saleeby and Sharp, 1980; Chen and Moore, 1982; Clemens-Knott and Saleeby, 1999; Lackey et al., 2005).
Average bulk Ap-He ages along this transect range from 69 Ma to 80 Ma (Fig. 8), with an overall average of 74 ± 4 Ma (1 standard deviation, s.d.). Average eU amongst these samples is unusually high at 116 ppm. A few samples contained grains with highly divergent eU concentrations (see the Supplemental File [footnote 1]). The best such example is sample 11SS9, which includes four individual ages averaging 75 ± 2 Ma (1 s.e.), with a range in eU from 42 ppm to 447 ppm (Fig. 9). According to the radiation damage accumulation and annealing model (RDAAM of Flowers et al., 2009), the large difference in eU amongst these grains means that they must have substantially different closure temperatures owing to variations in accumulated radiation damage. On its face, the agreement amongst the ages of sample 11SS9 apatite grains (all are within ∼10% of the mean; see the Supplemental File [footnote 1]) suggests that cooling through the helium partial retention zone (PRZ) occurred quickly. However, even though the range is small, 11SS9 is in fact the only sample in our suite that shows a compelling age versus eU correlation (Fig. 9). We incorporate the RDAAM model into our thermal modeling to extract detailed t-T information from the large variation in eU of individual grains from this sample. This modeling is presented in a later section of this paper. All individual grain zircon and apatite (U-Th)/He data are tabulated in the Supplemental File [footnote 1].
Mean Z-He ages from these samples fall into two distinct populations (Fig. 8). The first (n = 4) has an average age of 91 ± 8 Ma (1 s.d.) and is significantly younger than the local pluton ages (Table 1). The second group (n = 3) has an average Z-He age of ca. 123 Ma. Two of these samples come from Early Cretaceous plutons with emplacement age ca. 120 Ma, and the third is not near a published zircon U-Pb age. We assume that these Early Cretaceous helium ages were set during conductive cooling of their host plutons (ca. 120 Ma) and remained cooler than the Z-He PRZ during the later plutonism (ca. 115–105 Ma).
Modeling Approach and Setup
To extract quantitative information from the helium data, we utilize the thermochronologic modeling software QTQt to obtain t-T histories of individual samples (Gallagher, 2012). QTQt employs a trans-dimensional Bayesian Monte Carlo Markov chain (MCMC) statistical approach to find the best t-T paths for a sample by employing a large number of iterative perturbations (we use at least 106 iterations). After each perturbation, the model compares the proposed path to the initial path and chooses the better-fitting of the two according to an acceptance criterion (Gallagher, 2012). The model converges on the best-fit t-T path through this process during the “burn in” period (Gallagher, 2012). For each of our model runs the “burn in” period consists of at least 5 × 105 iterations (after Vermeesch and Tian, 2014). After the model has converged on the best-fit t-T path, we run a set of 5 × 105 post–“burn in” model iterations to document the distribution of best-fit t-T histories. The model outputs show the results of this post–“burn in” period. QTQt can simultaneously apply this iterative process to find a most likely t-T path with multiple different data inputs (Ap-He, Z-He, Ap-4He/3He). For each sample we input all of the available helium data into QTQt. For a detailed review of QTQt and its relation to other thermal modeling software packages, see Vermeesch and Tian (2014).
We also impose a minimal set of manually controlled thermal history constraints. Where available, we use published zircon U-Pb ages as high-temperature constraints (650 ± 100 °C; for citations, see Table 1); elsewhere we used 115 ± 10 Ma, which encompasses the observed range of U-Pb zircon ages. We input a reasonable bounding box of temperature and time for the model to explore (150 ± 135 °C, 120 Ma to present) and a rough estimate of modern mean annual surface temperature (20 ± 5 °C) as a present-day temperature constraint. For samples 11SS1 and 11SS6 we also input a low-temperature constraint corresponding to the age of the overlying rock units (40 ± 5 Ma; 20 ± 5 °C). Table 2 lists the details of the inputs for each model run.
For samples 11SS1 and 11SS6, we input the mean Z-He age as well as the Ap-4He/3He data linked to the mean Ap-He age into the QTQt model (Table 2). The thermal modeling results from sample 11SS6 show rapid cooling ca. 95–85 Ma to <55 °C followed by slow cooling to surface conditions by 40 Ma (Fig. 10). This period of slow cooling corresponds to an erosion rate of roughly 30 m/m.y. (55 °C to 20 °C from 85 to 40 Ma with an assumed geothermal gradient of 25 °C/km [Brady et al., 2006]). The model for sample 11SS1 is consistent with rapid cooling from 95 to 85 Ma to ∼100 °C and slow cooling from ∼100 °C to ∼20 °C from 85 Ma to 40 Ma, implying an erosion rate of roughly 70 m/m.y. (Fig. 10). Both of these model results are consistent with zero additional cooling after exhumation to the surface ca. 40 Ma.
We ran an individual model for each of the other eight samples along the SSNP, none of which have Ap-4He/3He data. Individual single-grain Ap-He ages and the mean Z-He age for each sample were input into these models (Table 2). Figure 11 shows a compilation of the acceptable t-T paths for each of these model runs, and the results from these individual models and their fits to Ap-He data are included in the Supplemental File [footnote 1]. The model results are consistent with the results of samples 11SS1 and 11SS6, with samples cooled rapidly from temperatures hotter than the Z-He closure temperature (∼190 °C) to between 100 °C (11SS1) and 55 °C (11SS6) from 95 to 85 Ma. Slow cooling to the surface at rates consistent with the erosion rates determined from samples 11SS1 and 11SS6 (30–70 m/m.y.) occurred from 85 Ma to 40 Ma.
ADDITIONAL DATA FROM SAMPLE 11SS1
Anatase (U-Th)/He Chronometry
One of our samples (11SS1) is an altered granitic rock hosting a mineral assemblage of quartz, plagioclase, calcite, chlorite, anatase, and brookite. Crystalline anatase (TiO2) grains were separated from this sample using the same procedures as for apatite. A stereomicroscope was used to pick individual grains roughly 100 μm wide and 200 μm long, chosen based on size, morphology, and lack of visible inclusions. Euhedral grains had a tetragonal dipyramidal morphology and an orange color. Individual grain degassing followed the same procedure as for bulk apatite analyses. Due to our inability to recover individual grains after degassing, we used separate grains for measuring U and Th content, using the same dissolution and measurement procedure as for zircon. We calculated a raw age using the average He, U, and Th concentrations determined from several aliquots. An alpha correction was then applied using Ft calculated using a surface area–to–volume ratio determined from the grains used for the analyses, a density of ∼3.9 g/cm3, and our calculated Th/U ratio (after Farley et al., 1996; Ketcham et al., 2011). The He diffusion kinetics of anatase are presently unknown, so we treat this as a minimum anatase formation age. Table 3 lists anatase U, Th, and He data.
Calcite Clumped Isotope Thermometry and Stable Isotopes
We measured the clumped isotope composition of carbonate from an altered bedrock sample (11SS1) using a well-documented general procedure for determination of the temperature dependent mass-47 anomaly (Δ47) of carbonate samples by automated digestion, online purification, and measurement by dual-inlet gas-source mass spectrometry (e.g., Eiler, 2011; Dennis et al., 2011). Two samples of whole-rock material, 63.0 and 99.7 mg, were powdered to <106 µm and reacted under vacuum in separate McCrea-style vessels with 10% phosphoric acid for 24 h at 25 °C to react all calcite in the sample (McCrea, 1950). Evolved CO2 was extracted from the vessels and separated from water by conventional cryogenic methods on a glass vacuum line. A second reaction step for 24 h at 50 °C yielded no CO2, indicating that no dolomite was present in the sample (Al-Aasm et al., 1990). Based on manometric measurements of CO2, carbonate contents of the 63.0 mg aliquot and the 99.7 mg aliquot were calculated to be 3.39 wt% and 3.35 wt%, respectively. This calculation assumes that all carbonate was stoichiometric calcite and digestion of calcite proceeded to completion. Due to the excellent agreement of the percent carbonate values, we conclude that these assumptions are correct.
In order to obtain sufficient CO2 for a single mass-spectrometric measurement, these separate gas aliquots were combined into a single break-seal. The composite sample CO2 was purified on an automated system that includes multiple cryogenic steps and a pass through a Poropak-Q 120/80 GC column in a He carrier gas to remove potential organic contaminants, and measured on a Thermo Scientific MAT 253 gas-source mass spectrometer at Caltech. The results were projected into the absolute reference frame using standard equilibrated gases measured during the same week-long analytical session (Dennis et al., 2011).
The composition of carbonate in sample 11SS1 is: δ13Cvpdb = –10.70‰ ± 0.01‰ (vpdb, Vienna Pee Dee belemnite), δ18Ovsmow = 14.22‰ ± 0.01‰ (using the carbonate-acid fractionation from Swart et al. ; vsmow, Vienna standard mean ocean water), and Δ47 = 0.509‰ ± 0.012‰ (all 1σ s.e.). Using our in-house high-temperature calibration, this corresponds to a crystallization temperature of 103 ± 8 °C (Bonifacie et al., 2011). We infer from the texture of the sample (Fig. 7) that the calcite, anatase, and brookite likely grew together during the same period of alteration.
The occurrence of anatase and calcite in sample 11SS1 from the southernmost exposures of the SSNP surface offers another datum for the t-T history of the surface. δ18O of carbonate is dependent on growth temperature and the δ18O of the water from which it grew. Assuming that the Δ47 value of this sample was not modified by burial heating or rock-buffered recrystallization, this calcite was in equilibrium with a fluid with a δ18Ovsmow of –0.5‰ to –2.4‰ (Kim and O’Neil, 1997). Such a composition is intermediate between low-latitude meteoric water (∼–10‰ to –5‰; Sheppard, 1986) and plutonic rocks (5‰–12‰; Taylor, 1968) and could have been produced by isotopic exchange of meteoric water with bedrock. The temperature of calcite formation and the isotopic composition of the carbonating fluid strongly suggest that the sample was subject to substantial alteration through interaction with a hot fluid of meteoric origin. This is consistent with the mineralogy and fabric of the sample, which is highly altered, hosting a mineral assemblage of quartz, plagioclase, chlorite, and calcite intergrown with anatase and brookite (Fig. 7).
We interpret this hydrothermal alteration to be related to the early rapid exhumation of the sampled area ca. 95–85 Ma. Our field observations and those of Saleeby and Sharp (1980) indicate structural and petrologic continuity between the 11SS1 sample site and the sites of the 102 Ma U-Pb zircon ages for the Fountain Springs tonalite (Lackey et al., 2005; Saleeby and Sharp, 1980). If we then bracket this igneous crystallization age with the 97 ± 13 Ma anatase (U-Th)/He age and the 85 ± 5 Ma Z-He age, we find that based on thermal modeling of the conductive cooling of the Sierra Nevada batholith (Barton and Hanson, 1989), the hosting tonalite pluton (at 3–4 kb conditions) retained enough primary heat to render the thermal conditions for anatase + calcite formation and Z-He closure during 95–85 Ma rapid exhumation. Saleeby et al. (2010) hypothesized that ca. 90 Ma major west-dipping low-angle normal faults drove rapid exhumation along the west margin of the Sierra Nevada batholith. We further posit that such an extensional regime would have fostered hydrothermal alteration of the actively exhuming basement surface as large normal faults penetrated plutons along the west margin of the batholith that were still warm from primary heat, and such faults climbed further upwards in the crust to tap meteoric water sources.
North and South Ends of the Horizontal Transect
Thermochronometric and stratigraphic data tightly constrain the thermal history of the bedrock exposed at the north end of the horizontal transect near Friant (sample location 11SS6). The tonalite of Blue Canyon was emplaced ca. 114 Ma (Chen and Moore, 1982). Thermal modeling results are consistent with rapid cooling through Z-He and Ap-He partial retention zones ca. 95–85 Ma. From 85 Ma to 40 Ma, slow cooling to the surface occurred at a rate of roughly 30 m/m.y. Around 40 Ma this bedrock was deeply weathered and at earth-surface conditions when nonconformable deposition of the Ione Formation began. Since 40 Ma, no basement exhumation has occurred at this location. The modeling results (Fig. 10) and the lack of age versus eU correlation (see the Supplemental File [footnote 1]) within Ap-He data at this location are strong evidence that the overlying Tertiary section at this location was never thick enough to disturb Ap-He ages in the underlying bedrock. Thermal modeling shown in Figure 10 indicates that samples were not heated above 40–50 °C after 40 Ma, which corresponds to a maximum possible thickness of roughly 1 km of cover.
Data from the southern end of the horizontal transect near Fountain Springs similarly constrain the thermal history of the bedrock at this location. Pluton emplacement occurred at ca. 102 Ma, followed by rapid cooling to ∼100 °C ca. 85 Ma, after which slow cooling to surface conditions occurred at a rate of roughly 70 m/m.y. until deposition of the overlying tuff at 40.1 ± 0.3 Ma (Saleeby et al., 2016).
The bulk ages from samples at the northern and southern ends of the SSNP are quite different (average Ap-He is 92 Ma at sample location 11SS6 and 66 Ma at 11SS1). Despite this difference, the QTQt thermal modeling allows us to interpret them both in the context of the same general t-T history. In conjunction with Ap-4He/3He data for each sample, the models reveal that their thermal histories are both consistent with the same three phases: rapid cooling 95–85 Ma, slow cooling 85–40 Ma, and no cooling 40 Ma to present. The thermal modeling indicates that the significant divergence in their ages is due to the different rates of slow cooling from 85 Ma to 40 Ma (∼70 m/m.y. in the south and ∼30 m/m.y. in the north), rather than a different timing of rapid cooling. These different slow erosion rates resulted in an additional ∼2 km of erosion at 11SS1 from 85 to 40 Ma compared to 11SS6. While the data and modeling results do not constrain the mechanism for this difference in slow erosion rates, it is noteworthy that these erosion rates bracket the estimates from the axial part of the southern Sierra during the same period of time, 40–60 m/m.y. (House et al., 1997, 2001; Clark et al., 2005; Mahéo et al., 2009; Sousa et al., 2016). We interpret this close agreement to indicate that the difference between these slow erosion rates is minor relative to the major phases of thermal history that are clearly present along the entire SSNP.
Thermal History of the Southern Sierra Nevada Pediment
All of the thermal modeling results are consistent with the same three-phase style of cooling history. Our primary conclusion about this history is that the SSNP was rapidly cooled to between 100 °C and 55 °C ca. 95–85 Ma, and then slowly cooled and exhumed to near the surface by 40 Ma. On average, if the entire length of the pediment were at the surface ca. 40 Ma, then the cooling rate from 85 Ma to 40 Ma would have been roughly 30–70 m/m.y. (35 to 80 °C of cooling over 45 m.y. with a geothermal gradient of 25 °C/km).
In our interpretation, the same batholithic swath exposed for an extended period of erosion and chemical weathering prior to 40 Ma is again exposed as the modern bedrock landscape; i.e., it is a Paleogene fossil landscape. This raises the question: How could this landscape have survived this time interval without significant erosion occurring?
The climatic conditions conducive to the chemical weathering required to form the types of nickel laterite occurrences along the SSNP are roughly >1000 mm/yr annual precipitation with cold month mean temperature ranging from 15 to 27 °C (Thorne et al., 2012). Modern conditions along our study area are not warm or wet enough to meet these criteria. However, previous workers invoked warmer and wetter global conditions in the Eocene (Pearson et al., 2007) to explain the formation of middle Eocene paleo-Oxisols within the Ione Formation in central California (Yapp, 2004) and developed on bedrock beneath the middle Eocene section in Baja California, Mexico (Abbott et al., 1976). In drier and cooler climates, these paleosols can be resistant to weathering and could potentially last for millions of years at the earth surface (e.g., Bierman and Turner, 1995). However, where plutonic bedrock crops out we expect that erosion would have been too fast to preserve the landscape for ∼107 yr. In other words, if this landscape had been exposed continuously since 40 Ma we would expect the Ap-He ages to be younger due to continued cooling.
Because of the old Ap-He ages and the presence of 500-m-scale relief within the modern landscape, we prefer a different model. Integrating the results of Saleeby and Foster (2004), Stock et al. (2004, 2005), Saleeby et al. (2012, 2013a), and Cecil et al. (2014), we hypothesize that ca. 40 Ma, Cenozoic sediments covered the SSNP and preserved the ca. 40 Ma landscape. A sedimentary thickness on the order of several hundred meters could have completely buried the modern relief without resetting Ap-He ages. This thickness is of the same order of magnitude as that of the Cenozoic section in the foothills of the northern Sierra (Bateman and Wahrhaftig, 1966).
The overlying sediments were likely removed during late Pliocene–Pleistocene erosion consistent with model predictions by Saleeby at al. (2012, 2013a) and in conjunction with the pulse of river incision documented by Stock et al. (2004, 2005). This erosion may have been due to a combination of factors including climate change related to ice age onset (e.g., Bintanja and van der Wal, 2008) and surface uplift potentially driven by dynamic processes related to the Isabella anomaly (Zandt et al., 2004). This erosion removed the overlying sediments and weathered basement which are significantly more erodible than intact basement rocks (Sklar and Dietrich, 2001). This revealed the more resistant bedrock pediment and its ferruginous silcrete carapace preserved as the fossil landscape below, re-exposing it as the modern landscape.
Potential Cause and Mechanism of Early Rapid Exhumation
From ca. 95 to 85 Ma, the southernmost Sierra Nevada–Mojave segment of the Cretaceous arc gravitationally collapsed and the batholith was rapidly exhumed to depths equivalent to ∼10 kb (e.g., Chapman et al., 2012; Saleeby, 2003; Saleeby et al., 2007). Liu et al. (2010) argued for a dynamic link between this event and the subduction of a large oceanic plateau which impinged on the Cretaceous subduction zone ca. 90 Ma. Based on the mating of basement core petrography and geochronology, and deep seismic data for the Great Valley subsurface immediately west of the SSNP, Saleeby et al. (2010) hypothesized that ca. 90 Ma major west-dipping low-angle normal faults drove rapid exhumation along the west margin of the Sierra Nevada batholith. The rapid exhumation required by our thermal modeling from 95 to 85 Ma is in general agreement with the timing hypothesized by Saleeby et al. (2010). Together these multiple lines of evidence suggest that the deep exhumation of the Cretaceous arc to the south was not spatially limited to the southernmost Sierra–Mojave region. We hypothesize that rapid exhumation of the SSNP from 95 to 85 Ma is a northern, lower-magnitude manifestation of this same event.
Regional Implications for Southern Sierra Nevada Evolution
With the exception of the area south of the Kern River (Mahéo et al., 2009), prior to this study no low-temperature thermochronometric data had been published from the southern Sierra foothills. The westernmost published data that were available for the region between 36°N and 37°N were a few Ap-He ages along the main trunk and north fork of the Kings River, ∼30 km east of our study area (Sousa et al., 2016), and a horizontal transect (referred to as T1 in House et al., 1998) between the Kaweah and San Joaquin drainages running ∼45 km east of our study area (House et al., 1998).
The southern Sierra Nevada is significantly different from the northern Sierra, both geologically and physiographically. With the exception of the Eocene rocks near Friant and Fountain Springs, the southern Sierra almost completely lacks the Paleogene deposits that are common in the northern Sierra (e.g., Busby et al., 2016). The southern Sierra also lacks the distinctive western ramp morphology that characterizes the north. Instead, the southern Sierra rapidly attains elevations of roughly 2000 m across a series of topographic steps (e.g., Hake, 1928). Despite these differences between the northern and southern Sierra and the complete lack of thermochronometric data from the western foothills, previous workers leaned on the assumption that throughout the Cenozoic the southern Sierra has behaved similarly to the northern Sierra, as a rigid west-down tilt block with a hinge line lying close to the western foothills–San Joaquin Valley boundary (e.g., Wakabayashi and Sawyer, 2001; House et al., 1998). Sousa et al. (2016) showed that the rigid-block assumption is incorrect in the vicinity of Kings Canyon, where a kilometer-scale west-down normal fault was active in Eocene time, and suggested that the WSFS was likely active in 45–40 Ma along the entire span of the southern Sierra from the San Joaquin River to the Kern River. This fault activity was part of a tectonic regime marked by uplift and extension within the coupled Sierra Nevada–Great Valley region, including uplift of the axial southern Sierra and shallowing of the proximal Great Valley forearc (Bartow, 1992; Sousa et al., 2016).
What was happening in the foothills during this time? We hypothesize that contemporaneous with this regional tectonic event ca. 45–40 Ma, some uplift and exhumation should have occurred in the foothills. Because our He ages are all older than the time of this hypothesized exhumation, we conclude that this exhumation was not of sufficient magnitude to noticeably disturb the Ap-He and Ap-4He/3He data along the bedrock pediment. Based on the QTQt modeling for samples 11SS1 and 11SS6, we estimate that there could not have been more than roughly 500 m of exhumation ca. 45–40 Ma.
The overlying Eocene rocks at the northern and southern termini of our study area closely follow the timing of this event (deposition beginning ca. 40 Ma). Combining the thermal modeling and the evidence from the overlying Eocene deposits, we conclude that a few hundred meters of exhumation could have occurred in the foothills ca. 45–40 Ma in conjunction with shallowing of the proximal Great Valley forearc to the west and axial Sierran fault-controlled uplift to the east (Bartow, 1992; Sousa et al., 2016).
Summary of the Chronology of Southern Sierra Nevada Landscape Evolution and Tectonic Forcing
Integrating our new data with Eocene activity on the WSFS, as well as other previously published data, we piece together a chronology of tectonic and landscape evolution for the southern Sierra Nevada outlined in Figure 12. The first phase of this chronology is the emplacement of the southern Sierra Nevada batholith ca. 115 ± 10 Ma in our study area, and ending at 85 Ma in the eastern part of the range. During the final stages of magmatism (95–85 Ma), the bedrock swath along our study area was rapidly exhumed (shown in red shades in Fig. 12) to ∼3–4 kb levels and ∼55 °C (in the north) and 100 °C (in the south). This exhumation is roughly contemporaneous with, and likely genetically related to, the profound tectonic exhumation and gravitational collapse of the southernmost Sierra–Mojave region to the south (Chapman et al., 2012). After the cessation of magmatism and early rapid exhumation, the entire SSNP slowly cooled at rates roughly the same as those of the axial part of the range from 85 to 40 Ma to near-surface conditions (shown in brown shades on Fig. 12). Combined with igneous barometric emplacement pressures of 3–4 kb (Ague and Brimhall, 1988; Nadin et al., 2016), the thermochronologic data indicate that the early rapid exhumation (95–85 Ma) accounted for ∼8–9 km of exhumation, while the slow erosion from 85 to 40 Ma accounted for the final 2–3 km of exhumation. Previously published thermochronometric and igneous barometric data (House et al., 1997, 1998, 2001; Clark et al., 2005; Ague and Brimhall, 1988) from higher elevations along the axial southern Sierra Nevada further suggest that the early phase of exhumation also included most of the rest of the southern Sierra Nevada batholith (roughly 3–4 km of early exhumation in the axial part of the range).
In the axial Sierra, the extended period of slow erosion (85–40 Ma) resulted in the initial form of the modern Sierra, including the low-relief interfluvial highlands (Clark et al., 2005), and the long-wavelength (>10 km) large-amplitude (>1 km) topographic relief that is visible on digital elevation models (e.g., Fig. 5; House et al., 1998, 2001). In the foothills, this resulted in formation of bedrock pediment morphology as well as the distinctive nickel laterite occurrences discussed earlier in this paper.
Around 45–40 Ma, activity on the WSFS resulted in extension and uplift of the axial southern Sierras and kilometer-scale incision in the major southern Sierran trunk river canyons (shown in green shades on Fig. 12; Sousa et al., 2016). To the west of the WSFS, in the foothills, a small amount of exhumation may have occurred (roughly a few hundred meters) in conjunction with shallowing of the proximal Great Valley forearc (Bartow, 1992; Sousa et al., 2016). From 40 Ma through the late Neogene, slow erosion continued in the axial southern Sierras, and a shallow cover of Cenozoic deposits likely armored the SSNP. Post–10 Ma, as a result of the convective removal of dense sub-batholithic mantle lithosphere, epeirogenic deformation partially re-exposed the SSNP to its current state (shown in yellow shades on Fig. 12). The shallow cover armoring the SSNP eroded, exposing the ancient bedrock landscape, and uplift in the axial southern Sierra resulted in the incision of the inner slot canyons common to the major Sierran trunk rivers (Stock et al., 2004). Active upper-mantle dynamic processes are resulting in Pleistocene to Holocene uplift of the Kern Arch and coupled subsidence of the Tulare Basin (Fig. 1) as the most recent phases of epeirogenic deformation (Zandt et al., 2004; Saleeby et al., 2012, 2013a; Saleeby and Foster, 2004; Cecil et al., 2014).
Multiple types of data including Ap-He, Ap-4He/3He, Z-He, stratigraphic constraints, geomorphic observations, and distinct mineralogical and paleosol occurrences indicate that the bedrock landscape exposed along the southern Sierra Nevada pediment is a Late Cretaceous to early Cenozoic landscape. This landscape evolved during a prolonged period of erosional modification and chemical weathering from ca. 85 Ma to 40 Ma following a phase of rapid, probably tectonic exhumation along the western Sierra Nevada batholith between 95 and 85 Ma. Little to no net erosion has occurred along the length of the pediment over post–40 Ma time.
Incorporating our new data with previously published constraints, we present the following chronology of tectonic and landscape evolution:
(1) Cretaceous batholithic emplacement began in the current Great Valley subsurface at ca. 140 Ma and migrated eastwards ending at ca. 85 Ma along the eastern Sierra Nevada (Saleeby and Sharp, 1980; Chen and Moore, 1982; Saleeby et al., 2010). At ca. 115–100 Ma the principal locus of magmatism corresponded to the area that was subsequently exhumed as the SSNP.
(2) Early batholithic rapid exhumation occurred ca. 95–85 Ma, on the order of 8–9 km along the western foothills and 3–4 km in the axial part of the range. This was likely dynamically linked to the contemporaneous profound tectonic exhumation and gravitational collapse of the southernmost Sierra–Mojave region due to the subduction of a large oceanic plateau immediately south of our study area.
(3) From 85 to 40 Ma, slow erosion and chemical weathering occurred in the foothills, and the axial Sierra low-relief highland plateaus and major trunk river canyons initially formed (e.g., House et al., 1998, 2001; Clark et al., 2005).
(4) From ca. 45 to 40 Ma, extensional tectonics and uplift of the southern Sierra Nevada and Great Valley region resulted in kilometer-scale incision of major Sierran river canyons and high-angle normal faulting on the WSFS (Sousa et al., 2016). At this time, no more than a few hundred meters of exhumation occurred along the foothills.
(5) Post–40 Ma, sediments likely shallowly covered the SSNP, and slow erosion of the axial southern Sierra Nevada batholith continued.
(6) Post–10 Ma, epeirogenic transients caused by mantle lithospheric dynamics resulted in surficial deformation which re-exposed the SSNP, uplifted the modern southern Sierra peaks, and caused subsidence of the Tulare Basin.
We thank Kerry Gallagher for assistance with setting up QTQt runs and Lindsey Hedges for help with sample preparation, analyses, and a life full of friendship. Thanks to Guest Associate Editor Cathy Busby and two anonymous reviewers for constructive reviews of this manuscript. This work was partially supported by the Gordon and Betty Moore Foundation through grant GBMF #423.01 to the Caltech Tectonics Observatory.