We investigate the putative Pliocene–Quaternary removal of mantle lithosphere from beneath the southern Sierra Nevada region using a synthesis of subsidence data from the Great Valley, and geomorphic relations across the Sierra Nevada. These findings are used to test the results and predictions of thermomechanical modeling of the lithosphere removal process that is specific to the Sierra Nevada, as presented in an accompanying paper referenced here as Part I. Our most successful thermomechanical model and the observational data that it explains are further bundled into an integrated physiographic evolution–geodynamic model for the three-dimensional epeirogenic deformation field that has affected mainly the southern Sierra Nevada–San Joaquin Basin region as a result of underlying mantle lithosphere removal.
The coupled Sierra Nevada mountain range and Great Valley basin are recognized as a relatively rigid block (Sierra Nevada microplate) moving within the San Andreas–Walker Lane dextral plate juncture system. Our analysis recognizes that the Sierra Nevada possessed kilometer-scale local and regional paleotopographic relief, and that the Great Valley forearc basin possessed comparable structural relief on its principal stratigraphic horizons, both dating back to the end of Cretaceous time. Such ancient paleorelief must be accounted for in considering late Cenozoic components of uplift and subsidence across the microplate. We further recognize that Cenozoic rock and surface uplift must be considered from the perspectives of both local epeirogeny driven by mantle lithosphere removal, and regional far-field–forced epeirogeny driven by plate tectonics and regional upper-mantle buoyancy structure. Stratigraphic relations of Upper Cretaceous and lower Cenozoic marine strata lying on northern and southern Sierra Nevada basement provide evidence for near kilometer-scale rock uplift in the Cenozoic. Such uplift is likely to have possessed positive, and then superposed negative (subsidence) stages of relief generation, rendering net regional rock and surface uplift. Accounting for ancient paleorelief and far-field–driven regional uplift leaves a residual pattern whereby ∼1200 m of southeastern Sierra crest rock and similar surface uplift, and ∼700 m of spatially and temporally linked tectonic subsidence in the southern Great Valley were required in the late Cenozoic by mantle lithosphere removal. These values are close to the predictions of our modeling, but application of the model results to the observed geology is complicated by spatial and temporal variations in the regional tectonics that probably instigated mantle lithosphere removal, as well as spatial and temporal variations in the observed uplift and subsidence patterns. Considerable focus is given to these spatial-temporal variation patterns, which are interpreted to reflect a complex three-dimensional pattern resulting from the progressive removal of mantle lithosphere from beneath the region, as well as its epeirogenic expressions. The most significant factor is strong evidence that mantle lithosphere removal was first driven by an east-to-west pattern of delamination in late Miocene–Pliocene time, and then rapidly transitioned to a south-to-north pattern of delamination in the Quaternary.
The late Cenozoic removal of mantle lithosphere from beneath the southern Sierra Nevada region has gained attention by virtue of its extraordinary geophysical documentation (cf. Jones et al., 1994, 2004; Ruppert et al., 1998; Zandt et al., 2004; Reeg, 2008; Frassetto et al., 2011; Gilbert et al., 2012), and its diversity of geologic expression (cf. Ducea and Saleeby, 1996, 1998a; Manley et al., 2000; Saleeby and Foster, 2004; Farmer et al., 2002). Geophysical imaging of large-scale structures resulting from this process, and data on its geologic expressions provide valuable constraints for formulation and iteration of thermomechanical models that explore the underlying dynamics. One class of such models was presented in Le Pourhiet et al. (2006) and is refined in the accompanying paper (Saleeby et al., 2012, this themed issue, referred to below as Part I). In these models, arrays of geological and geophysical data are formulated into initial and boundary conditions, and input parameters, and a number of testable geological predictions emerge. With this in mind, the primary aim of this paper is to present new and to integrate existing constraints for late Cenozoic uplift and subsidence of the Sierra Nevada and Great Valley that appear to be related to this mantle lithosphere removal event, and to test the results of the Part I modeling against these observables.
We pose the concept of both anomalous rock uplift and tectonic subsidence applied to vertical displacements that are specific to the southern Sierra region of mantle lithosphere removal. In order to pursue the results of these displacements, it is necessary to develop a working model for the physiographic development of the entire Sierra Nevada and Great Valley. This coupled system is widely recognized as a semicoherent crustal block named the Sierra Nevada microplate, which moves semi-independently within the San Andreas–Walker Lane transform system (Fig. 1) (Argus and Gordon, 1991, 2001; Unruh et al., 2003). As such, the topographic evolution of the Sierra Nevada, and the subsidence history of the Great Valley are commonly assumed to be uniform across the microplate. We depart from this convention and assert that the southern Sierra Nevada and San Joaquin Basin of the southern Great Valley both underwent profound Pliocene–Quaternary rock uplift and tectonic subsidence phases that are missing from the northern reaches of the microplate, and that these phases were genetically related to underlying mantle lithosphere removal. The elucidation of these anomalous vertical displacement components is confounded, however, by paleorelief variations along the Sierra Nevada (House et al., 2001; Wakabayashi and Sawyer, 2001; Stock et al., 2004, 2005; Chapman et al., 2012), and ancient segmentation patterns of the Great Valley inherited from its Cretaceous forearc setting (Repenning, 1960; Harwood and Helley, 1987; Reid, 1988; Imperato, 1995). For this reason, the subject of anomalous uplift and subsidence must be considered from the perspective of the uplift-subsidence history of the entire microplate.
The goals of this paper and of Part I are to address the observed vertical displacements in Earth’s surface resulting from mantle lithosphere removal in the southern Sierra region, and to relate these displacements to the mechanisms governing the removal process. We refer to these displacements as epeirogenic in that they reflect relatively long wavelength to amplitude vertical displacements that arise primarily from radial force components. We posit that the physiographic expression of these displacements has been superposed over paleorelief patterns of the Sierran Nevada microplate that are inherited from the Late Cretaceous convergent margin history of the region, as well as the effects of far-field–driven vertical displacements that operated over Cenozoic time at a regional scale. We further posit that these factors considered together with the flexural-isostatic responses to late Cenozoic drainage basin erosion and related sediment loading of the Great Valley can explain the first-order physiography of the microplate. Here, we pose a working model for the late Cenozoic physiographic evolution of the Sierra Nevada microplate, and link it to a geodynamic analysis of the underlying governing processes as approximated in the thermomechanical models of Le Pourhiet et al. (2006) and Part I. We further iterate between physiographic and dynamic aspects of our integrated model, along with insights derived from volcanism and heat flow patterns, in the derivation of a three-dimensional kinematic model for mantle lithosphere removal. We focus on uplift and subsidence, which we interpret as the imprints of the epeirogenic deformation field resulting from mantle lithosphere removal in the southern Sierra Nevada region. We recognize that this process is most clearly documented in the southern Sierra region, and that it is likely to be concentrated in, or restricted to, this region (Jones et al., 1994; Ruppert et al., 1998; Zandt et al., 2004; Reeg, 2008; Frassetto et al., 2011; Schmandt and Humphreys, 2010; Gilbert et al., 2012). We further note that modern debate on Sierran landscape evolution rarely recognizes the physiographic variation between the southern and northern Sierra Nevada, as was clearly noted in early literature on the subject (Matthes, 1965; Christensen, 1966). Our integrated physiographic evolution–geodynamic model suggests that a significant component of this variation has arisen over the past ∼5 m.y. as a direct result of mantle lithosphere removal that was restricted to, or concentrated in the southern half of the microplate.
TECTONOMORPHOLOGY OF THE SIERRA NEVADA MICROPLATE
At first order, the Sierran Nevada microplate behaves as a W-tilted fault block, the tilt and erosion of which are roughly balanced by westward-increasing subsidence and sedimentation in the Great Valley (Fig. 1). The bounding structures of the microplate are all diffuse zones. To the east, there is the eastern Sierra escarpment system and the Walker Lane–Eastern California shear zone, which, in the current kinematic regime, is a system of dextral and en echelon normal fault segments along which ∼10 mm/yr movement of the Pacific–North American plate motion is partitioned (Unruh et al., 2003). The west margin of the microplate is defined by the San Andreas fault system and Coast Range fold belt (Argus and Gordon, 1991, 2001). The northern end is bounded by the Inks Creek fold belt, which accommodates modest convergence between the microplate and the Klamath Mountains (Harwood and Helley, 1987; Unruh et al., 2003; Fay and Humphreys, 2008). The southern end is bounded by the Garlock fault and the Tehachapi–San Emigdio fold-and-thrust belt (Davis and Burchfiel, 1973; Davis and Lagoe, 1988; Chapman and Saleeby, 2012). The southern ∼100 km segment of the microplate is deforming internally by normal motion on the Kern Canyon fault system and Kern range front–Pond-Poso zone, and along sinistral breaks of the White Wolf zone (Guacci and Purcell, 1978; Clinton et al., 2006; Mahéo et al., 2009; Nadin and Saleeby, 2010; Amos et al., 2010).
Figure 1 also shows the axial trace of the Great Valley syncline, a regional fold in Upper Cretaceous through Cenozoic strata exhibiting a low-dipping broad east limb and a steeper-dipping west limb. Maximum subsidence in the Great Valley is linked to west tilt of the microplate along an axis that runs along the western Foothills ramp (Fig. 1). The tilt axis runs continuously for ∼350 km along the ramp, where, between ∼37°N and ∼36°N, it is deformed and obscured by anomalous subsidence along the eastern margin of the Tulare (sub-) Basin. South of Tulare Basin, the tilt axis is deflected into a more northerly trend along the eastern margin of the Kern arch. Upper Cretaceous and Cenozoic strata with low west dips are exhuming adjacent to the western Foothills tilt axis. The across-strike dimension of the exhuming strata increases significantly along the Kern arch and a zone to the north of Tulare Basin that we designate as the central Foothills swell (Fig. 1).
Regional patterns in topographic relief across the Sierra Nevada are exhibited in the Figure 1 inset by smoothed contours that have been constructed by mechanical contouring across major interfluve surfaces (after Wakabayashi and Sawyer, 2001). The contour pattern removes the topographic effects of major west-flowing rivers and displays the regional geometry of a slowly denuding low-relief landscape surface that was inherited from the western shoulder of the Nevadaplano, a regional orogenic plateau that characterized the U.S. Cordillera in Late Cretaceous through mid-Cenozoic time (De Celles, 2004). The Figure 1 inset mates the smoothed topographic profiles of the Sierran uplands to structure contours on the basement surface beneath the Great Valley (Wentworth et al., 1995; see structure sections later herein). The inset also includes color tones that denote the regions undergoing both anomalous rock uplift (delamination bulge) and tectonic subsidence (Tulare Basin) in Pliocene–Quaternary time (explained further later herein).
The Figure 1 inset shows that, in addition to the regional west tilt of the basement surface, Sierran uplands have more subtle longitudinal regional topographic gradients, which slope to the NW and SE off an ∼175-km-long range crest culmination that is in excess of 4000 m elevation. Geomorphic and isotopic data resolve paleorelief dating back to the end of the Cretaceous (House et al., 2001; Wakabayashi and Sawyer, 2001; Stock et al., 2004, 2005; Sousa et al., 2013). Figure 2 is the same digital elevation model base as in Figure 1, but with ancient relief and tectonic features displayed for the region that was to become the Sierra Nevada microplate. On this figure, the regions of the Sierran uplands that retain evidence for local paleorelief are shaded in tones denoting greater paleorelief southward, culminating in the Mount Whitney region. The areas of paleorelief possessed major canyons that drained off the west flank of the Nevadaplano, prior to the calving off of the Sierra Nevada microplate from the plateau margin at 10 ± 2 Ma (Busby and Putirka, 2009; Saleeby et al., 2009a; Henry, 2012).
The longitudinal topographic gradients of the Sierran uplands also reflect regional paleorelief patterns, as indicated by structural and stratigraphic relations along the northern and southern ends of the microplate. To the north, there is the Late Cretaceous marine Hornbrook Basin, which sits in the subsurface immediately north of the Sierra Nevada (Fig. 2), with its transition into the northern Great Valley marked by the Upper Cretaceous shoreline environment that extended across northern Sierra basement (Nilsen, 1984; Harwood et al., 1981; Harwood and Helley, 1987; Batt et al., 2010). The steeper longitudinal gradient that slopes SE off the range crest culmination represents a paleogradient that developed in response to Late Cretaceous large-magnitude extension of the southern Sierra Nevada–western Mojave region (Wood and Saleeby, 1998; Bartley et al., 2007; Saleeby et al., 2007; Chapman et al., 2010, 2012). Large-magnitude extension of this region was controlled by segmentation patterns in the Late Cretaceous convergent margin (Malin et al., 1995; Saleeby, 2003; Liu et al., 2010), denoted on Figure 2 by the inflection in the Rand (subduction) megathrust lateral ramp. Major crustal structures that developed during this regime are shown on Figure 2 as the Kern Canyon–White Wolf and proto–Owens Valley transfer zones, the Isabella and Kern range front breakaway zones, the Maricopa supradetachment basin, and the southern Sierra detachment system. Figure 2 also shows the presently known extent of the Late Cretaceous southern Sierra–western Mojave extensional terrane and its southwest extension into the displaced Salinia batholithic terrane (after Grove, 1993; Saleeby, 2003; Bartley et al., 2007; Ducea et al., 2009; Saleeby et al., 2009b; Chapman et al., 2010, 2012). The principal tie point between the Salinian basement exposures and those of the western Mojave, across the San Andreas fault, is shown after Huffman (1972). Basement rock exposures of the Late Cretaceous extended terrane are characterized by widespread depositional remnants of Maastrichtian to middle Eocene clastic marine strata (Cox, 1987; Lucas and Reynolds, 1991; Grove, 1993; Wood and Saleeby, 1998; Monastero et al., 2002; Lofgren et al., 2008; Chapman et al., 2012). These depositional remnants sit in structural positions suggestive of formation in supradetachment basins, and/or active graben-horst systems, and they further indicate the development of a continental borderland across the southernmost Sierra Nevada, western Mojave, and restored Salinia in conjunction with extensional tectonism.
Figure 3 is a map of the southern Sierra Nevada region showing selected features that are referred to throughout the text, and Table 1 is an explanation and supporting references for these features. Remnants of Eocene strata interpreted as filling supradetachment basins (Wood and Saleeby, 1998; Chapman et al. 2012) occur along the N-tilted footwall of the proto–Garlock (normal) fault (Fig. 3). These contain marine trace fossils and demonstrate that the low elevations that resulted from Late Cretaceous extension persisted in the southern Sierra region into Tertiary time. A comparison of the positions of the Late Cretaceous extensional terrane and the Hornbrook basin (Fig. 2), and the Upper Cretaceous and Eocene remnants of marine strata lying on northern and southern Sierra Nevada basement (Figs. 2 and 3) with the smoothed topographic contour map of the Figure 1 inset elucidates the importance of regional longitudinal paleorelief variation along the Sierra Nevada.
Regional subbasin domains of the Great Valley consist of the Sacramento and San Joaquin Basins, which are separated by the Stockton arch, a structural high located at ∼38°N that is expressed as a transverse ridge in basement surface structure contours (Fig. 1, inset). The Stockton arch is controlled in the subsurface by fault-controlled relief that is buried by upper Cenozoic strata. It is a polyphase structure, with the (principal) Stockton fault (Fig. 2) originating in the Late Cretaceous as a S-down normal fault, which inverted to reverse motion in the early Cenozoic (Imperato, 1995).
Cenozoic strata of the San Joaquin Basin thicken and reflect progressively greater proportions of marine environments southward from ∼36.5°N as they pass into a deep-marine basin of Neogene age. The approximate positions of the early and late Miocene shoreline environments and the Neogene shelf break are shown on Figure 2. Shallow-marine and adjacent terrestrial environments of the southeastern San Joaquin Basin extended for at least 10 km eastward across the current southern Sierra basement uplift (Fig. 3; Mahéo et al., 2009; Saleeby et al., 2009a, 2013; Saleeby and Saleeby, 2010). In Quaternary time, the eastern San Joaquin Basin was partitioned into the Tulare (sub-) Basin to the north, and the Maricopa (sub-) Basin to the south by uplift of the Kern arch. The southern end of the original San Joaquin Basin has been destroyed by Quaternary uplift driven by N-S shortening of the Tehachapi–San Emigdio fold-and-thrust belt, and the western margin is actively deforming along the Coast Range fold belt (Fig. 1).
A comparison of Figures 1 and 2 shows that active structures of the diffuse southern margin of the microplate developed in crust that had been highly extended in the Late Cretaceous. Major structures of this diffuse margin that appear to be direct precursors to active structures are the proto–Owens Valley and Kern Canyon–White Wolf transfer zones, and the hinge of the Rand megathrust lateral ramp, along which the Garlock fault formed. Seismic imaging, structural mapping, and petrologic studies indicate that the megathrust lateral ramp was responsible for the Late Cretaceous tectonic erosion of the mantle wedge from beneath the southernmost Sierra Nevada batholith and the adjacent western Mojave–Salinia region of the southern California batholith (Cheadle et al., 1986; Jacobson et al., 1988; Li et al., 1992; Malin et al., 1995; Saleeby et al., 2003; Yan et al., 2005; Luffi et al., 2009; Chapman et al., 2010, 2012). Synchronous with this profound tectonic erosion event, the residual mantle wedge that lay north of the lateral ramp was quenched to a lithospheric geotherm (Ducea and Saleeby, 1996, 1998b; Saleeby et al., 2003). It is the geologically recent and ongoing removal of this residual mantle wedge domain that presents the dynamic system for which we pursue the epeirogenic expression.
INTEGRATED PHYSIOGRAPHIC EVOLUTION–GEODYNAMIC MODEL OF MANTLE LITHOSPHERE REMOVAL AND ITS SURFACE EXPRESSIONS
We present an analysis that leverages observational data against thermomechanical modeling results as a means to pursue the underconstrained problem of the late Cenozoic rock and surface uplift history of the Sierra Nevada. The principal geologic imprint of such uplift is erosion, or removal of rock record. We assert that if our modeling is properly posed in terms of initial and boundary conditions (Part I), it follows that valuable insights into the likely uplift history may be gained from model predictions. We recognize that such modeling can only offer approximations compared to the complexity of the geodynamic system modeled. By linking subsidence data constraints for the adjacent Great Valley to our analysis, we can incorporate geologic imprints that produce rock record, subsidence, and sedimentation. This in turn provides model validation tests that help lend confidence to our treatment of the more poorly constrained (uplift) aspects of our analysis.
We hypothesize that resolvable epeirogenic displacements related to mantle lithosphere removal in the southern Sierra Nevada region are superposed over local and regional paleorelief patterns, and potential far-field–induced vertical displacements that may have affected much of, or the entire, Sierra Nevada microplate through Cenozoic time (Suppe et al., 1975; Murphy et al., 1998; Lowry et al., 2000; Pierce et al., 2002; Humphreys, 2008). The region hypothesized to have undergone net rock and surface uplift under the mantle lithosphere removal regime is shown as the delamination bulge on the Figure 1 inset. The bulge dissipates northward across ∼38°N latitude and appears to be bounded to the east by the eastern Sierra escarpment system. South of ∼36°N, it turns southwestward and runs across the Sierra Nevada and into the eastern San Joaquin Basin, where it is bounded by W-side-down normal faults that step down toward the basin along the Kern arch (Fig. 3). The principal zone of delamination, and hence the area of the resulting bulge, was controlled by the lithospheric structure inherited from Cretaceous high-magma-flux growth of the southern Sierra Nevada batholith (Part I, Fig. 4 therein). The delamination bulge is peripheral to the eastern and southern margins of Tulare Basin, which represents focused tectonic subsidence that is dynamically linked to the bulge. The principal area of the delamination bulge underwent anomalous rock uplift and accelerated river incision mainly in the Pliocene across the San Joaquin, Kings, and Kaweah drainages, and in the Quaternary along the lower Kern drainage (Wakabayashi and Sawyer, 2001; Stock et al., 2004, 2005; Saleeby et al., 2009a; Figueroa and Knott, 2010; Nadin and Saleeby, 2010). Stratigraphic, geomorphic, geodetic, and isotopic data (Stock et al., 2004, 2005; Mahéo et al., 2009; Figueroa and Knott, 2010; McPhillips and Brandon, 2010; Nadin and Saleeby, 2010; Hammond et al., 2012; Cecil et al., 2013) indicate that this phase of rock uplift is diminishing in the north (San Joaquin to Kings transition) and accelerating in the south (lower Kern drainage and Kern arch). Here, we interpret this as the mark of mantle lithosphere removal progressing in three dimensions from the northern to the southern region.
Pliocene–Quaternary rock uplift of the delamination bulge was contemporaneous with anomalous subsidence in Tulare Basin (Saleeby and Foster, 2004; see following). The subsidence is anomalous in that for any given longitudinal trace of the axial to eastern Great Valley, significantly more sediment of younger than 7 Ma age has accumulated in the Tulare Basin; also, the eastern margin of Tulare Basin is the only area of the western Sierra Foothills that is embayed by Quaternary sediments that have aggraded up drainages, as opposed to either being actively incised or in regional offlap relationship downslope from the tilt axis (Figs. 1 and 3).
Based on geophysical data, petrogenetic data on volcanic-hosted mantle xenoliths, and patterns in Neogene–Quaternary volcanism, the region between the San Joaquin and Kern River drainages is where geologically recent removal of mantle lithosphere has most clearly occurred (Ducea and Saleeby, 1996, 1998a, 1998b; Ruppert et al., 1998; Farmer et al., 2002; Zandt et al., 2004; Reeg, 2008; Frassetto et al., 2011; Gilbert et al., 2012; C.H. Jones, 2012, written commun.; Part I). Thermomechanical modeling (Le Pourhiet et al., 2006; Part I) suggests that removal was initiated in middle Miocene time by Rayleigh-Taylor (RT) convective mobilization of the cooled mantle wedge that formed beneath the Sierra Nevada batholith. The instability progressed to the state of promoting the delamination of the eclogitic root (arclogite) of the batholith. Arclogite is defined as eclogitic arc cumulates and subordinate interlayered spinel ± garnet (wedge) peridotites (Anderson, 2005), which in this application refers to an ∼40-km-thick cogenetic sequence that developed beneath the southern Sierra Nevada batholith (Ducea and Saleeby, 1996, 1998b; Saleeby et al., 2003; Fig. 4A). Much of the Sierran arclogite is of crustal origin, but its mechanical properties and vertical position place it in the upper level of the mantle lithosphere, where in primary state it sat above ∼40 km of mantle wedge peridotite (Ducea and Saleeby, 1996, 1998b; Ducea, 2001; Saleeby et al., 2003).
The thermomechanical model developed in Le Pourhiet et al. (2006) and Part I used a finite-element code that modeled topography, visco-elasto-plastic temperature-dependent rheologies, and temperature- and composition-dependent densities (after Cundall and Board, 1988; Poliakov et al., 1993; Le Pourhiet et al., 2004). The better-constrained geological parameters such as initial geotherm, structural geometry, and elastic and brittle rheologies were fixed, and the most poorly constrained creep parameters were varied and evaluated in terms of their ability to simulate observed mantle structure and geological resultants. The model trace is along a transverse section centered along the Kings and Kaweah interfluves (Part I, Fig. 1 and Fig 1. herein), and it accounts for initial lithospheric structure, geotherm, and topography. The model does not account for motion across the model trace. Errors induced by not accounting for dextral shear of the San Andreas fault and Walker Lane shear zones (Unruh et al., 2003) are uncertain, but they are provisionally considered to be minimal considering the apparently low degree of coupling between normal and tangential stress and strain components observed along these strike-slip systems (Zoback et al., 1981, 1987; Mount and Suppe, 1987).
In Figure 4A, we present a summary of our preferred model, highlighting the initial state (0 m.y.) and two key steps in forward model time at 10 and 20 m.y. (after Part I; Figs. 5 and 6). The 10 m.y. step is critical because it shows the break-off of the Sierra Nevada microplate by the ascent of asthenosphere to the base of the crust east of the Sierra Nevada batholith. The sequence of events that arises in our preferred model suggests that arclogite root delamination began to progress slowly as microplate break-off occurred (Part I, Fig. 6, 10 m.y. and 14 m.y. steps), accelerated rapidly over ∼5 m.y., and that it is still progressing (Fig. 4A, 20 m.y. step). The Tulare Basin coincides with much of the area where the partly delaminated root remains attached to the lower felsic crust (Zandt et al., 2004; Reeg, 2008; Frassetto et al., 2011; Gilbert et al., 2012; C.H. Jones, 2012, written commun.). Integration of all available seismic data for the region suggests that the partially delaminated root is suspended in the upper mantle steeply eastward and southward from its residual zone of attachment (Le Pourhiet et al., 2006; Part I). Accordingly, we constructed the vertically projected surface trace of the locus of root detachment and defined it as the delamination hinge trace on Figures 1 and 3. This trace approximates the transition between anomalous subsidence of Tulare Basin and the delamination bulge. The delaminated portion of the arclogite root that is suspended in the mantle and the associated thermal anomaly that extends to within the adjacent peridotitic mantle form the high-wavespeed “Isabella anomaly” (after Jones et al., 1994; Part I, Fig. 3 therein).
Model predictions (Part I) for the vertical displacements in Earth’s surface along a transverse profile across the microplate arising from root delamination are plotted in Figure 4B. This shows displacement profiles initiating at ∼10 m.y. (forward model time), the time at which lithospheric break-off is predicted, resulting in the inception of the Sierra Nevada microplate (Fig. 4A). Several lines of evidence indicate that microplate inception, as recorded by initiation of the eastern Sierra escarpment system, occurred at 10 ± 2 Ma (Loomis and Burbank, 1988; Henry and Perkins, 2001; Surpless et al., 2002; Mahéo et al., 2004; Busby and Putirka, 2009; Saleeby et al., 2009a). The initial conditions for our model simulate a basal thermal perturbation arising from the opening of the Pacific-Farallon slab window as instigating mantle lithosphere instability. Slab window opening occurred at ca. 23–20 Ma beneath the model trace, but was time transgressive over ∼10 m.y. across the entire southern Sierra Nevada region (Atwater and Stock, 1998; Wilson et al., 2005). This leads to several million years of uncertainty in applying the model results to observed geologic history. The 10 m.y. model prediction for microplate inception and the 10 ± 2 Ma inception from observations are in very close agreement, considering the temporal and spatial relations of slab window opening. The vertical displacement profiles of Figure 4A correspond to minimum rock uplift in upland areas and tectonic subsidence in basinal areas. Added components of subsidence and uplift arising from sediment loading and exhumation are pursued later herein. We interpret the 20–22 m.y. period of model time (Fig. 4A) to correspond to modern geologic time, given the range of uncertainties outlined earlier, and to be applicable to an ∼150-km-wide transverse corridor centered along the Kings and Kaweah drainages (Figs. 1 and 3). Vertical displacement values of Figure 4B are summarized for Tulare Basin and the eastern Sierra crest in Table 2. A comparison of these values to observations is complicated by the limited nature of the observational data, as well as the complexity in temporal and spatial relations. In terms of tectonic subsidence, the best data constraints are for stratigraphic markers in the ≤7 +1/–0 Ma age range (discussed later herein), and thus we adopt the 14–22 m.y. period of subsidence (∼680 m) as most appropriate for comparison. An additional complexity for the rock uplift comparison is the model predictions for rapid eastern crest uplift from 20 m.y. to 22 m.y., followed by rapid crest subsidence. We discuss observational data that perhaps mimic such a pattern across space. For our main comparison with the observational data, we adopt ∼800 m as the best nominal prediction for delamination-driven eastern crest rock uplift, and we discuss geographic variations that could reflect positive growth through the ∼800 m phase (21–22 m.y.), as well as a possible transition into the post–22 m.y. eastern crest subsidence phase.
Model iterations presented in Part I suggest that vertical displacements in Earth’s surface are driven primarily by arclogite root delamination, with RT-driven displacements being of second-order importance. We thus refer primarily to the delamination process here. Delamination-related epeirogenic transients are driven by a combination of flexure of the outer elastic crust related to the changing distributed load geometry of the arclogite root, and to isostatic changes related to upper-mantle–lower-crustal density redistributions. The relative contributions from flexure and isostasy in the production of the delamination bulge cannot be quantitatively extracted from the model, but they can be visualized at first order in Figure 4B, with the flexural component being comparable to the modeled flexural uplift of the Coast Ranges resulting from the root load, and the added isostatic component summing to the form of the delamination bulge. The double peaks for the delamination bulge result from the maximum flexural and isostatic components being slightly out of phase. We provisionally relate our model results quantitatively to observational data on subsidence and uplift, recognizing that the model system is highly idealized relative to the actual physical system that we are modeling. Pysklywec and Cruden (2004) performed analogue experiments investigating surface vertical displacement patterns arising from mantle lithosphere removal, and they found that structural flaws in the crust induce local-scale vertical displacements that are nontrivial with respect to the length scales of the epeirogenic zone. The crust of the southern Sierra Nevada and San Joaquin Basin possessed numerous tectonic structures that predated mantle lithosphere removal that could likewise induce vertical displacements and changes in horizontal length scales that are not accounted for in our numerical modeling approach. We discuss such limitations in our modeling as we proceed.
We now proceed to the observational data on uplift and subsidence, and we compare these data quantitatively to model results along transverse profiles, parallel to the model trace. We then use specific features predicted by the modeling to qualitatively pursue delamination in three dimensions.
LATE CENOZOIC ROCK UPLIFT OF THE SOUTHERN SIERRA NEVADA
Estimates for late Cenozoic rock uplift along the eastern Sierra crest are 2 ± 0.5 km (Huber, 1981; Unruh, 1991; Wakabayashi and Sawyer, 2001; Wakabayashi, 2013) to little or zero (Poage and Chamberlain, 2002; Mulch et al., 2006; Cassel et al., 2009, 2012). The former interpretation is based mainly on the updip projection of upper Cenozoic strata lying along the Foothills ramp, and basement incision beneath Neogene surfaces, both determined north of 37°N. The latter interpretation is based mainly on stable isotope paleo-altimetry. The accuracy of the former procedure is called into question by the modeling of geomorphic forcing (Small and Anderson, 1995; Pelletier, 2007), as well as thermomechanical modeling (Part I), both of which suggest flexure of the Sierra Nevada crust during rock uplift, drainage basin exhumation, and Great Valley sediment loading. Furthermore, the updip projections of some key upper Neogene units along the Foothills ramp yield crest elevations up to ∼400 m too high (Wakabayashi and Sawyer, 2001), and the updip projections of apatite (U-Th)/He isochrone surfaces within the batholith of the Kings and San Joaquin drainages yield crest elevations that are ∼1500 m too high (McPhillips and Brandon, 2010). These findings show that the microplate deformed internally during its west tilt, undermining the precision of the geometric analysis of tilt. Nevertheless, the tilt analysis based on differentially tilted strata of the Foothills ramp (Unruh, 1991) clearly shows that some component of west tilt operated over Pliocene–Quaternary time. Unfortunately, basement incision alone, even below given Neogene surfaces, offers little in terms of the timing of such incision and its application to eastern Sierra uplift. Paleo-altimetric constraints on Sierra Nevada elevation history based on stable isotopes are not robust at up to ∼1000 m resolving power (Galewsky, 2009a, 2009b; Molnar, 2010). Thus, the contrary findings of both procedures for determining Sierra Nevada uplift, or lack thereof, carry comparable uncertainties.
Considering that direct evidence for both geologically recent (Pliocene–Quaternary) delamination, and the timing and amount of rock uplift is restricted to the Sierra Nevada south of 38°N, we focus primarily on Pliocene–Quaternary rock uplift constraints for the region south of 38°N.
Pliocene–Quaternary Rock Uplift Patterns in the Southern Sierra Nevada
Constraints for Pliocene–Quaternary rock uplift are direct for Kings River and lower Kern River drainages, and by inference relatively strong for the Kaweah, San Joaquin, and Stanislaus drainages. Kings River drainage rock uplift is directly constrained by cosmogenic dating of tiered cave sediments in a deeply incised vertical marble unit (Stock et al., 2004, 2005). These studies indicate a total of ∼400 m of incision between 3.0 and 1.5 Ma, with an order of magnitude drop in rates between 1 and 1.5 Ma, interpreted by these workers as a result of protective mantling of the channel bottom by glacially derived debris. Alternatively, the drop in incision rates could result from the upstream migration of a knickpoint past the point where the incision rate was measured. We also note here, and return to under our discussion of Tulare Basin subsidence, that our modeling predicts phases of subsidence across the Sierran uplands, as well as phases of rock uplift (Fig. 4B). Such subsidence phases superposed on the regional tilt pattern could have temporarily dampened river gradients, thereby temporarily slowing incision. Cosmogenic age data for marble units in three forks of the Kaweah drainage (Fig. 1) fall on the Kings River array, suggesting a common incision history. One data point along the lower Stanislaus drainage (Fig. 1) lies at the lower end of the Kings and Kaweah data array, which could be argued to lie on the rapid incision trend, or a slower trend. Proximity of the San Joaquin drainage to the Kings drainage, its intermediate position relative to the Kings and Stanislaus drainages, and the commonly cited evidence for Pliocene west tilt and incision for the San Joaquin drainage (Huber, 1981; Unruh, 1991; Wakabayashi and Sawyer, 2001; Wakabayashi, 2013) indicate a common forcing and incision regime as that which is directly documented for the Kings drainage.
The ∼400 m of late Pliocene incision documented in the Kings drainage was modeled by Stock et al. (2004) as a transient response to ∼1500 m of eastern crest uplift initiating in the late Miocene. Such tectonic forcing is in line with 10 ± 2 Ma inception of the Sierra Nevada microplate, the development of the eastern Sierra escarpment system (Loomis and Burbank, 1988; Henry and Perkins, 2001; Surpless et al., 2002; Mahéo et al., 2004; Busby and Putirka, 2009; Saleeby et al., 2009a), and our model results, which predict the initial stages of delamination bulge growth between 10 and 14 m.y. model time (Fig. 4B).
Quaternary rock uplift along the lower Kern River area is constrained by stratigraphic relations of the eastern San Joaquin Basin, low-temperature thermochronologic data, and geomorphic relations. Map and stratigraphic relations summarized in Figure 3 and in the Figures 5B–5D structure sections show that the upper Pliocene shallow-marine San Joaquin Formation is erosionally truncated updip along the Kern arch. The Kern arch differs from the Foothills ramp north of 37°N in being intensely broken by high-angle normal faults, i.e., far too many to show on Figure 3 (Nugent, 1942; Bartow, 1984; Mahéo et al., 2009; Saleeby et al., 2009a). Normal faulting of the Kern arch extends northwards into Tulare Basin (Figs. 3 and 5; Croft and Gordon, 1968; Saleeby and Foster, 2004, Data Repository map). Regardless of late Cenozoic normal faulting in Tulare Basin and the Kern arch, bedding is regionally homoclinal like that of the main Foothills ramp and eastern Great Valley to the north. Stratal thickness and erosional truncation patterns suggest that in excess of ∼1000 m of Kern arch strata were stripped off the basement along the eastern margin of the arch in the Quaternary (Figs. 5C and 5D). The lower Kern River basement gorge (Fig. 3) was cut subsequent to erosional stripping of the eastern Kern arch strata off the basement, and it represents a superimposed drainage with its large meanders resembling in wavelength and amplitude the meanders that are cut into the Tertiary section downslope on the arch. The lower gorge of the Kern River has incised ∼700 m into basement below the exhumed Tertiary nonconformity surface. Integration of stratigraphic, geomorphic, and apatite He thermochronometric data presented later herein under “Quaternary cryptic subsidence and rock uplift of the Kern Arc” indicate ∼1–1.8 km of Quaternary exhumation and rock uplift for the lower Kern Gorge area, in addition to the ∼700 m of lower gorge basement incision. In terms of local relief variation, the lower Kern River gorge is the most profound river channel to exit the western Sierra Nevada Foothills. Geomorphic studies along the western Foothills between 35.5°N and 37°N suggest that lower channel reaches become progressively younger southward from the Kings to Kern drainages, and that the lower Kern gorge is likely as young as ca. 0.5 Ma (Figueroa and Knott, 2010).
Geomorphic and stratigraphic relations along the Kern arch segment of the Foothills ramp are of little use in constraining eastern Sierra crest rock uplift. Pliocene–Quaternary internal faulting of the Kern arch and the southern Sierra south of ∼36°N clearly negates the updip projection of bedding surfaces as a constraint for eastern crest rock uplift (Mahéo et al., 2009; Saleeby et al., 2009a). The west tilt of the Kern arch and the updip exhumation of its Tertiary strata appear to be linked to late Cenozoic W-side-up normal faulting along the Greenhorn and Breckenridge faults, and late Quaternary breaks of the Kern Canyon system (Fig. 3; Mahéo et al., 2009; Saleeby et al., 2009a; Amos et al., 2010; Nadin and Saleeby, 2010). The region east of the Kern Valley graben (Fig. 3) is characterized by the relict regional south slope inherited from the Late Cretaceous (Fig. 2), which to the south has been back tilted ∼3°N by Neogene extensional faulting (Fig. 3; Mahéo et al., 2009).
The southward progressive youthfulness of lower channel reaches from the Kings to Kern drainages is paralleled by other constraints on geologically recent rock uplift patterns. The comparison of basement and river channel detrital apatite He age arrays from the San Joaquin and Kings River drainages indicates that relief is decreasing in the San Joaquin drainage relative to the Kings drainage (McPhillips and Brandon, 2010). As discussed already, the Kings and Kaweah drainages have slowed considerably in basement incision over the past 1.5 m.y., while the lower Kern channel has undergone accelerated incision through the Quaternary. Geodetic data for the Sierra Nevada indicate a similar pattern for contemporary vertical displacements in Earth’s surface (Fay et al., 2008; Bennett et al., 2009; Nadin and Saleeby, 2010; Hammond et al., 2012). Figure 6 shows a synthesis of contemporary vertical displacement data based on global positioning system (GPS) monuments and interferometric synthetic aperture radar (InSAR) data for much of the Sierra Nevada, as well as stratigraphic and level line data for the Kern arch. This synthesis shows that the area shown as the delamination bulge on the Figure 1 inset is undergoing active rock uplift, with the main area of the bulge undergoing ∼1 mm/yr, increasing to ∼2 mm/yr into the Kern arch area. The rapid vertical displacement pattern of the Kern arch area is truncated to the west by W-side-down normal faulting (Fig. 3). West of the Kern arch, the San Joaquin Basin is in subsidence, although absolute vertical motion is obscured by groundwater removal subsidence (Lofgren and Klausing, 1969; Castle et al., 1983). Figure 6 also shows that the central Foothills swell is actively ascending relative to the northeast edge of Tulare Basin, as well as the ramp to the north of ∼38°N. As discussed herein, we interpret this swell as part of the delamination epeirogenic signal as well.
Accelerated Rock Uplift and Active Thermal Transient
The most actively rising region of the southern Sierra Nevada, as delineated here (Fig. 6), also coincides with an active thermal anomaly defined here as the spatial overlap of numerous warm and hot springs and wells with regionally low basement heat flow. Measured heat flow in the Sierra Nevada batholith is amongst the lowest known on continents, particularly along the western Foothills (cf. Saltus and Lachenbruch, 1991). Study of Sierran lower-crustal exposures and mantle xenoliths show that this low-heat-flow signal is inherited from the conductive cooling of the Cretaceous batholith and its mantle wedge from beneath by slab flattening (Ducea and Saleeby, 1996; Lee et al., 2000; Saleeby et al., 2003, 2007; Brady et al., 2006; and modified after Dumitru, 1990). In Figure 7, we plot the locations of warm and hot springs and wells, and “hot oil fields” of the Kern arch (after Laney and Brizzee, 2003; Saleeby and Saleeby, 2009; Cecil et al., 2013). In the Figure 7 inset, we show comparative down-hole temperature data for a sample of hot oil fields on the Kern arch in comparison to values typical of the basin directly west of the arch. These data demonstrate a profound thermal anomaly beneath the Kern arch, which is modeled and discussed in more detail along with down-hole thermochronologic data in Cecil et al. (2013). On Figure 7, we also plot the basement heat-flow data (after Saltus and Lachenbruch, 1991). To this, we overlay a color tone denoting the areas containing younger than 4 Ma volcanic rocks interpreted to have erupted in conjunction with delamination (Ducea and Saleeby, 1998a; Manley et al., 2000; Farmer et al., 2002; Part I). The springs and wells are co-extensive with the volcanics along the eastern Sierra Nevada region, but they also extend southwestward along the most rapidly rising domain of the delamination bulge (Figs. 6 and 7), which has been nonvolcanic since early Miocene slab window–related volcanism.
We interpret the hydrothermal systems that are delivering the heated waters to springs and wells in the otherwise anomalously cool batholithic host rocks as the “head wave” of an active thermal transient that is being driven by the youngest phases of root delamination. The area of the anomalous thermal transient corresponds to the footwall tilt-block of the Greenhorn–Breckenridge–Kern Canyon normal fault system (Fig. 3), which is the fastest-rising area of the entire Sierra Nevada (Fig. 6). Scattered delamination volcanics and elevated basement heat-flow values to the east of this normal fault system (Fig. 7) represent a more mature thermal state of the crust arising from the late Miocene–Pliocene phases of east-to-west delamination. Accordingly, this eastern, partially equilibrated domain merges northward with the region characterized by more copious delamination volcanism. The western anomalous province overlies prominent midcrustal and upper-mantle negative conversions in seismic receiver functions that appear to terminate adjacent to the delamination hinge trace (Frassetto et al., 2011; Part I, Figs. 3B and 3C therein). The negative conversions merge northeastward with the negative conversions that lie beneath the area of widespread delamination volcanism. Seismic tomography and refraction data show low-velocity upper mantle extending to the base of the crust beneath the anomalous province as well (Mereu, 1987; Ruppert et al., 1998; Schmandt and Humphreys, 2010; Reeg, 2008; Part I, Figs. 3B and 3C therein). We interpret the shallow, low-velocity mantle as ascended asthenosphere that has replaced the delaminated root, and the negative conversions beneath the anomalous province as domains of partial melt that have yet to render surface eruptions.
The correspondence of the anomalous thermal province with the area of most rapid active rock uplift, their mutual position relative to the delamination hinge trace, and the diminished rock uplift rates in regions to the north are in line with the interpretation that root delamination initiated from east to west in late Miocene–Pliocene time along the southern Sierra, and then transitioned to a south-to-north pattern in Quaternary time in the region of the Kern arch and lower Kern River basement gorge. The resulting delamination hinge trace correspondingly forms an ∼90° sector along the southern perimeter of Tulare Basin.
SUBSIDENCE PATTERNS ALONG THE AXIAL TO EASTERN SAN JOAQUIN BASIN
In this section, we pursue anomalous subsidence in Tulare Basin by the analysis of a series of (longitudinal) structure sections drawn parallel to the axis of the southern Great Valley (Fig. 8). We extend this analysis across the Kern arch in pursuit of related subsidence across the arch that has been rendered cryptic due to Quaternary rock uplift and erosion.
Anomalous Subsidence of Tulare Basin
Three evenly spaced longitudinal structure sections are presented along the San Joaquin Basin in Figure 8. The section traces are shown on the Figure 5 inset. Well files used to generate the sections are tabulated in Supplemental File 33. Regionally continuous stratigraphic markers along the eastern to axial San Joaquin Basin consist only of the upper Miocene Santa Margarita Formation, an extensive, mainly littoral sand sheet, Upper Cretaceous forearc basin strata, and the basement nonconformity. The range of Santa Margarita deposition is regionally constrained to 12–6.5 Ma (after Bartow and McDougall, 1984; Olson, 1988; Goodman and Malin, 1992; Hosford Scheirer and Magoon, 2003). In our analysis, we generalize the end of deposition as 7 Ma. In the southwestern part of the basin, facies transitions into the upper Miocene Reef Ridge shale are used in some wells as Santa Margarita age equivalents. Northward along the eastern section (Fig. 8A), Santa Margarita marine sands are swamped out by volcaniclastic debris of the Zilch Formation. Additionally, key Pliocene–Pleistocene stratigraphic horizons are included in Figure 8 above the Santa Margarita unit, as data are available. Shallower levels of the basin typically are not cored or logged, and a predominance of terrestrial deposits inhibits the correlation of what data is available. Local horizons for the top of the Upper Pliocene shallow-marine San Joaquin Formation and its southeastward facies transition into the Kern River Formation (ca. 2.5 Ma) and the Corcoran Clay E-member (ca. 615 ka) are shown, where found. Age constraints for the San Joaquin Formation and E-clay were reviewed in Hosford Scheirer and Magoon (2003).
The Figure 8 structure sections were constructed as a search for the deepest part of Tulare Basin, in longitudinal profile, in order to constrain late Cenozoic subsidence that is anomalous relative to regional subsidence, which is relatively uniform along any given longitudinal trace of the Great Valley. Anomalous subsidence is defined here as the difference in subsidence between the deepest part of Tulare Basin, in longitudinal profile, and subsidence that is expressed at the northwestern ends of each section. For Miocene and younger time, the northern ends of each section are considered typical of regional Great Valley patterns, lacking Tulare-related anomalous subsidence. This analysis also includes determination of structural relief of key stratigraphic horizons between Tulare Basin and the crest area of the Kern arch. The Kern arch constitutes an internally faulted NW-tilted fault block proximal to the western Sierra (Fig. 8A), which diminishes in structural relief progressively westward into the deeper Great Valley basin (Figs. 8B and 8C). In sections 8A and 8B, the Kern arch combined with the Edison basement high form a buried late Cenozoic horst and graben system. Measurement locations for the anomalous subsidence residuals and Tulare Basin–Kern arch structural relief are shown for each section on Figure 8, and the data are compiled in Table 3.
Inspection of Figure 8 leads to designating section 8B as most appropriate for basin center depth relations. This section trace runs ∼30 km east of the synclinal axis of the Great Valley, while section 8C is adjacent to the synclinal axis, making a much larger component of its subsidence history related to the regional asymmetry of the Great Valley trough. Section 8A is more proximal to the western Sierra Foothills, which makes the younger critical horizons more susceptible to local facies variations resulting from Pliocene–Quaternary alluvial fans of the western Foothills (cf. Atwater et al., 1986), and the southern reaches of the voluminous Neogene terrestrial volcaniclastic tongue that filled the Sacramento and northern San Joaquin Basins from the northeast (cf. Busby and Putirka, 2009). The Salyer well yields useful data most proximal to the deepest part of section 8B, and unit thicknesses are used from this well for the Table 3 measurements. Log data from the Salyer well, as well as our stratigraphic picks, are presented in Supplemental File 44. From these, we constructed total and tectonic subsidence curves (after Allen and Allen, 1990; Watts, 2001) (Supplemental File 55), and we add these to a synthesis of pertinent published subsidence curves (Moxon, 1987; Moxon and Graham, 1987) for the eastern Great Valley region (Fig. 9; well locations in Fig. 6). Note that on Figure 9, the two wells from the eastern Sacramento Basin (A and B) lack the distinct latest Miocene–Quaternary subsidence phase that is present in the three wells from Tulare basin (C, D, and F). This is in line with our concept of anomalous subsidence in Tulare Basin, and the use of subsidence residuals. Note also that latest Miocene–early Quaternary anomalous subsidence of Tulare Basin continued across the Kern arch (Fig. 9G). This is discussed later herein.
Structural relations along section 8B show that 1150 m of total anomalous subsidence occurred in the central part of Tulare Basin, post–7 Ma (Table 3). This translates into 625 m of anomalous tectonic subsidence (Supplemental File 5 [see footnote 5]). Part I model results for delamination-related tectonic subsidence in Tulare Basin for the 14–22 m.y. interval of model time are ∼680 m (Fig. 4B; Table 1). We consider this as good agreement between model results and observation. The Figure 4B displacement curves also make predictions concerning early-stage anomalous subsidence affecting the western Foothills region prior to westward migration of the principal depocenter. Unfortunately, facies relations along the eastern margin of Tulare Basin are such that this cannot be tested by stratigraphic means. Under our paleogeographic synthesis, herein, we present sedimentological data that may record this early-stage subsidence migration pattern.
Structural relief measurements between Tulare Basin center and the crest of the Kern arch for the 7 Ma horizon along sections 8A and B (Table 3) are 1260 and 1400 m, respectively. This is in line with the excess of ∼1000 m of strata eroded off the exhumed basement nonconformity along the eastern edge of the Kern arch that was approximated by the projection of bedding relations.
Pliocene–Quaternary Cryptic Subsidence and Rock Uplift of the Kern Arch
Stratigraphic relations along the northwest flank of the Kern arch strongly suggest that the upper Pliocene shallow-marine San Joaquin Formation and its facies transition into the fluvial-deltaic Kern River Formation extended across much, or all of the area of the arch (Figs. 3, 5, and 8). As illustrated in Figures 3, 5C, and D5D, erosion of the crest area of the arch has exhumed the Tertiary section down to lower Miocene levels over a broad area, and thus a substantial thickness of strata may have rested on the area of the arch prior to its Quaternary growth. Assumption of a constant thickness for the upper Neogene strata along sections 8A and 8B across the southern Tulare Basin and the arch suggests that ∼1250–1400 m of sediment overburden have been removed from the crest area of the arch (Figs. 5 and 8; Table 3). In order to gain additional constraints on this exhumation, we examined a number of cores from oil fields on the Kern arch. We found a variety of features that constrain the latest Miocene to early Quaternary (ca. 6–1 Ma) depth of burial of the arch, which in conjunction with stratigraphic, geomorphic, and geodetic data summarized herein, further constrain the Quaternary rock uplift and exhumation of the arch. These features include mechanical granulation textures from rapid sediment loading, low-grade metamorphic mineral growth, vitrinite reflectance, (U-Th)/He thermochronometry of detrital apatite grains, and thermal modeling of down-hole temperature determinations (Cecil et al., 2013). Some of the results from this large and diverse data set are summarized in Figure 10.
Figure 10A shows a simplified version of the Figure 8B longitudinal structure section with constraints plotted for post–1 Ma exhumation across the medial area of the Kern arch based on findings from oil fields that are located along the trace of the section (Cecil et al., 2013). Local structural relief on the basement nonconformity has been smoothed across the arch in order to minimize the effect of complex Neogene and possible Late Cretaceous faulting. Significant faults that appear to have been involved in generating the arch uplift are retained. The depth of the ca. 1 Ma depositional surface is approximated across the Tulare and Maricopa Basins, assuming constant subsidence, sedimentation, and compaction rates between 2.5 Ma (top of San Joaquin marker) and the modern surface. In Figure 10B, the ca. 1 Ma depositional surfaces in Tulare and Maricopa Basins, and the ca. 1 Ma surface determined from exhumation across the Kern arch are retrodeformed to a flat profile that is intended to approximate the ca. 1 Ma sediment surface prior to both the rise of the arch and post–1 Ma subsidence and sedimentation in the adjacent subbasins. Mismatches in the arch-bounding faults are shown, which require either reversals in displacement patterns, or the involvement of additional structures in the uplift of the arch that have yet to be resolved.
The Figure 10B reconstruction shows a regionally continuous southward-facing slope for the eastern San Joaquin Basin at ca. 1 Ma, along the trace of the Figure 8B structure section. This is interpreted to approximate the N-S profile of the axial to eastern San Joaquin Basin prior to the rapid uplift and exhumation of the Kern arch. This is significant because it shows that the area of the arch shared a common Pliocene–early Quaternary anomalous subsidence history with both Tulare Basin and Maricopa Basin, prior to its uplift and exhumation. This reconstruction also shows that latest Miocene to early Quaternary total subsidence increases southward along the basin, due primarily to sediment loading (Fig. 9; also Goodman and Malin, 1992). This loading component increases as the delta and delta front submarine fan system of the Caliente River is approached (Fig. 3). Exhumation and reburial relations along the northern slope of the arch also indicate recent aggradation of Tulare Basin sediments across the previously exhumed margin of the arch. Such aggradation is suggested to be a result of sediment supplied by higher degrees of exhumation across the higher levels of the arch to the east (Cecil et al., 2013), in conjunction with a rise in base level across Tulare Basin by the building out of the Kings River alluvial fan (Atwater et al., 1986).
The Quaternary exhumation of ∼1100–1300 m of upper Miocene to lower Quaternary strata off the Kern arch as shown on Figure 10A implies a comparable amount of cryptic subsidence in the region. Fuhrman well offers the widest range of burial indices, including depth-dependent mechanical granulation textures, low-grade metamorphic mineral growth, disturbed (U-Th)/He apatite age arrays, and down-hole temperature constraints (Saleeby and Saleeby, 2009; Cecil et al., 2013). We thus use the Fuhrman well relations here to generalize the cryptic subsidence-superposed rock uplift relations for the medial area of the Kern arch. Oil fields from the eastern part of the arch yield higher post–1 Ma exhumation values of up to ∼1800 m (Cecil et al., 2013), but these are more difficult to interpret in terms of cryptic subsidence relations due to the growth of syndepositional grabens along the range front, and also a poorer resolution as to the nature of the exhumed strata (i.e., possible range front proximal alluvial fans). We thus focus on the medial area of the arch and present log data and stratigraphic picks for Fuhrman well in Supplemental File 66. We impose ∼1350 m of post–Santa Margarita cryptic subsidence from Cecil et al. (2013) in producing the subsidence plot of Figure 9F (Supplemental File 77). There are no direct constraints on the form of the cryptic subsidence–rock uplift phases of the plot, other than ∼1350 m cryptic subsidence, mainly post–2.5 Ma, followed by rock uplift and exhumation in conjunction with the rise of the arch starting at ca. 1 Ma. We designate 1 Ma as the time of maximum sediment loading and subsequent time as the period of rock uplift and exhumation (Fig. 9F). This satisfies Figueroa and Knott’s (2010) interpretation for incision of the lower Kern gorge since 0.5 Ma, also considering that before the ∼700 m of lower gorge basement incision, ∼1–1.4 km of strata were eroded off the basement nonconformity—more than 1 km based on stratigraphic projections (Figs. 5B and 5C; Mahéo et al., 2009) and ∼1.4 km from Cecil et al. (2013). The subsidence to uplift transition at ca. 1 Ma is also consistent with seismic-reflection data in Tulare Basin, suggesting that the arch began to rise as a sediment source area at ca. 1 Ma (Miller, 1999), and is consistent with the vertical rate of 2.3 ± 0.5 mm/yr measured on the GPS monument that is located along the southeast margin of the arch (Fig. 6), assuming that this rate became operational at, or shortly after, ca. 1 Ma. A critical feature of the Figure 10 relationships, and of the Figure 9 subsidence curves, shows that “Tulare Basin” anomalous subsidence affected the area of the Kern arch prior to the Quaternary rise of the arch.
LATE CENOZOIC EPEIROGENY AND REGIONAL RELIEF EVOLUTION ALONG THE SIERRA NEVADA MICROPLATE
We now pursue the temporal and spatial relations of epeirogenic deformation across the southern Sierra Nevada region arising from root delamination by an analysis of stratigraphic and geomorphic data along two roughly orthogonal transects: (1) a NE-SW (transverse) transect that crosses the eastern Sierra crest at ∼37°N, which is considered applicable to an ∼150-km-wide corridor that is centered over the transect; and (2) a NW-SE (longitudinal) transect that synthesizes relations along the eastern San Joaquin Basin–western Foothills transition. The position of the delamination hinge trace is used as a reference for the geodynamic interpretation of the observed epeirogenic displacements, and in this context the results of our Part I modeling are tested against the data synthesis. The uncertainty in correlation of geologic and model time is applied spatially in our analysis. Considering that the San Joaquin drainage appears to be losing relief relative to the Kings drainage (McPhillips and Brandon, 2010), and the Kings drainage has slowed in its recent incision history over the past 1.5 m.y. (Stock et al., 2004, 2005), we correlate modern geologic time for the transverse trace with the 22–23 m.y. interval of model time (Fig. 4B), since modeled eastern crest rock uplift related to root delamination culminates and then ceases over this time interval. Since the Kaweah and Kern drainages are in more youthful stages of geologically recent uplift, we correlate the 20–22 m.y. interval of model time to recent time in these areas. Finally, we add an analysis of the first-order effects of geomorphic forcing and late Cenozoic far-field–induced epeirogeny to the delamination related epeirogenic effects for a more complete analysis of late Cenozoic epeirogeny across the entire microplate.
Transverse Vertical Displacement Profiles
In this section, we discuss Pliocene–Quaternary rock uplift of the southern Sierra Nevada that resulted from the tectonic forcing of root delamination, compounded by the flexural-isostatic response (geomorphic forcing) to accelerated drainage basin erosion in the Sierran uplands and sediment loading in the San Joaquin Basin (i.e., Small and Anderson, 1995; Pelletier, 2007). The geomorphic forcing studies are pertinent to the application of the Part I modeling to eastern Sierra crest rock uplift because such forcing is not accounted for in the modeling. Here, we merge our tectonic forcing results with the geomorphic forcing analyses of Small and Anderson (1995), recognizing that the modeling of Pelletier (2007) suggests that for the magnitude of tectonic forcing that our modeling predicts, a nontrivial component of geomorphic forcing arises. Small and Anderson (1995) used a flexural-isostatic model with distributed erosional unloading across the Sierran uplands, distributed depositional loading across the San Joaquin Basin, and the position of the Foothills tilt axis fixed.
A critical factor to the geomorphic forcing analysis is an approximation for the effective elastic thickness (Te) of the Sierra Nevada microplate crust. In Small and Anderson (1995), a 10–50 km range of Te was investigated over mean erosion rates (<ε>) of ≤0.3 mm/yr. A current Te of ∼5–15 km is derived for the Sierra Nevada and its transition into the Great Valley from a combination of geophysical and geomorphic approaches (Lowry et al., 2000; Granger and Stock 2004; Lowry and Perez-Gussinye, 2011). For our first-order analysis, we use a value of 10 ± 5 km. Small and Anderson’s model is also dependent on <ε>. Wakabayashi and Sawyer (2001) suggested a Pliocene range-wide value of ∼0.15 mm/yr, although this value is shown to be most applicable to main river channel basement incision between the Stanislaus and Kaweah drainages during the late Pliocene (Stock et al., 2004, 2005). The much faster Quaternary <ε> for the lower Kern gorge area and Kern arch, as discussed earlier herein, is not considered pertinent to this analysis and is discussed further later. Long-term <ε> of the Sierra Nevada from end of Cretaceous through at least early Neogene time is ∼0.05 mm/yr (Clark et al., 2005; Cecil et al., 2006; Clark and Farley, 2007; Mahéo et al., 2009). Considering these range of rates, and arguments for lower river gradients and erosion rates for the Sierra Nevada through early Cenozoic time (Wakabayashi and Sawyer, 2001; McPhillips and Brandon, 2012; Wakabayashi, 2013), we use <ε> of ∼0.1 ± 0.02 mm/yr for Pliocene–Quaternary time. Using these Te and <ε> values in Small and Anderson’s (1995, Fig. 4C) model produces a geomorphic rock uplift response of ∼400 ± 100 m since 5 Ma for the southeastern Sierra crest. Added to the ∼800 m of delamination-driven uplift (Fig. 4B), we find ∼1200 m of rock uplift by compounded delamination and geomorphic forcing.
We posit that topographic relief of the delamination bulge has been superimposed over regional relief patterns of the Sierran uplands, similar to Tulare Basin subsidence being superposed over regional subsidence patterns of the Great Valley. We designate a transect that crosses the medial area of Tulare Basin and the headwaters of the Kings River as a representative profile across the delamination bulge and dynamically linked zone of subsidence. In comparison, we consider a transect centered adjacent to the Mokelumne River drainage (Fig. 1) as being representative of microplate regional physiographic patterns where the effects of root delamination are lacking, or trivial. We base this on: (1) seismic imaging, indicating that the area of this drainage is north of where root delamination is occurring (Reeg, 2008; Schmandt and Humphreys, 2010; Frassetto et al., 2011; Gilbert et al., 2012; C.H. Jones, 2012, written commun.); (2) the pattern of Neogene–Quaternary volcanism (compare Busby and Putirka  and Farmer et al. ); and (3) its position to the north of the principal range crest topographic inflection, as well as the area shown by geodetic data to be undergoing resolvable contemporary uplift (Figs. 1 and 6). We interpret the Mokelumne headwater region to be located immediately north of the delamination bulge. In parallel, tectonic subsidence curves for late Cenozoic time along the Great Valley segment of the Mokelumne transect show no sign of Tulare Basin–like anomalous subsidence (Figs. 6 and 9). In Figure 11A, we show comparative smoothed topographic and basement depth profiles for the Kings and Mokelumne transects (Fig. 1 inset), and from these we derive a basement surface profile residual plot (Fig. 11B). If we assume that short-term (<1 m.y.) summit flat erosion rates of 0.012–0.075 mm/yr (Small et al., 1997; Riebe et al., 2000; Stock et al., 2004, 2005) are comparably applicable to the Kings and Mokelumne headwater regions since 5 Ma, the error induced by summit erosion for the crest elevation residual is trivial.
The Figure 11B plot yields an ∼1200 m topographic residual for the eastern Sierra crest, and ∼1400 m total subsidence residual for the Great Valley basement. The residual profile is to be compared to the 22 m.y. modeled surface displacement profile of Figure 4B. We plot the 22 m.y. profile in Figure 11C, and we adjust it for ∼400 m of geomorphically forced rock uplift of the Sierran crest (discussed earlier), and for the ∼525 m sediment loading residual determined for the “center” of Tulare Basin (Table 3; Supplemental File 3 [see footnote 3]). The adjusted 22 m.y. model profile is plotted on Figure 11D along with the residual profile, and they are shown to be quite comparable. This is consistent with our hypothesis that the principal differences in regional relief and late Cenozoic subsidence patterns between the northern and southern segments of the microplate result from the superposing of the delamination bulge and its dynamically linked subsidence zone over regional microplate physiographic patterns. The Figure 11D comparison should be considered semiquantitative in that the modeled rock uplift and geomorphic-forced components are both approximations. Nevertheless, the close correspondence of the Figure 11D profiles lends confidence to this analysis considering the regional setting of the Mokelumne profile relative to the region to the south, where numerous features point to ongoing delamination and related surface displacements.
Longitudinal Vertical Displacement Profiles along the Eastern San Joaquin Basin and Western Sierra Nevada Foothills
In Figure 12, we plot a series of vertical displacement profiles constructed along a longitudinal corridor that covers the western Foothills to axial San Joaquin Basin area (Fig. 5 inset). We register the profiles to an underlying lithospheric structure section synthesized from seismic data (Part I, Fig. 3C therein). Vertical displacement profiles based on total thicknesses for the top of the Santa Margarita Formation are constructed from stratigraphic relations of the Figure 8A and 8B structure sections, and from cryptic subsidence data from the Kern arch (Figs. 9F and 10; Saleeby and Saleeby, 2009; Cecil et al., 2013). The top of the Santa Margarita Formation is defined as sea level at ca. 7 Ma, and depth control for subsequent geologic time is provided by the ca. 2.5 Ma top of the San Joaquin Formation, the ca. 1 Ma maximum burial surface of the Kern arch, and modern depths. Also shown is a topographic profile along the western Foothills that runs parallel to the lithospheric structure section. The profile extends along the eastern margin of Tulare Basin. The topographic profile and lithospheric structure section both cross the upper exhumed levels of the Kern arch along the footwall of the Greenhorn–Breckenridge–Kern Canyon fault system, and the two Santa Margarita displacement profiles run from the center of Tulare Basin across the Kern arch and terminate in Maricopa Basin. Next, we relate the first-order features of the Figure 12 profiles to structural and kinematic relations of the underlying lithospheric structure section.
The Figure 12B profiles show that the axial to eastern San Joaquin Basin, south of ∼37°N, went into accelerated subsidence after ca. 7 Ma, and that after ca. 1 Ma, the Kern arch began to emerge, partitioning off the Tulare and Maricopa subbasins. The post–7 Ma accelerated subsidence is interpreted to have been forced by east-to-west–directed root delamination. The initial delamination-related accommodation space formed along the entire eastern to axial San Joaquin Basin in latest Miocene–early Quaternary time, paralleling the initial distribution of the in-place arclogite root (Part I, Fig. 4 therein). Then, as delamination transitioned from east-to-west to south-to-north during Quaternary time, the residual Tulare space continued to subside over the residual root load, while the Kern arch emerged over the area of active delamination. Growth of the arch is dampened westward into the basin (Figs. 8B and 8C), sympathetic with the westward thinning out of the initial arclogite root (Part I, Fig. 4 therein). Eastward from the basin edge, the upper slope of the Kern arch is exhumed of its Tertiary cover strata as it climbs toward the proximal footwall area of the Greenhorn–Breckenridge–Kern Canyon fault system.
Along any given profile running normal to the southern curved segment of the delamination hinge trace (Figs. 1, 3, and 7), the footwall block of the Greenhorn–Breckenridge–Kern Canyon system and the Kern arch form a fault block that is tilted into Tulare Basin (Figs. 8A and 8B), similar to the southern Sierra Nevada west tilt into Tulare Basin in transverse profiles to the north. Our modeling of root delamination predicts crustal extension off the flank of the delaminating root, represented in transverse profile by the eastern Sierra–Death Valley extensional province (Fig. 7). As delamination has transitioned into its south-to-north trajectory, it appears that delamination-driven extension has migrated to the Greenhorn–Breckenridge–Kern Canyon system, and to within Maricopa Basin (Figs. 3 and 8). The Maricopa Basin is commonly referred to as a foredeep to the Tehachapi–San Emigdio fold-and-thrust belt. However, this “foredeep” spatial relationship only developed during the Quaternary rise of the fold-and-thrust belt. Maricopa Basin is clearly controlled by high-angle normal and sinistral faults along the section traces of Figures 8A and 8B, and while the geometry of the White Wolf fault is poorly imaged along the Figure 8C section, where it appears to have inverted to active reverse motion (Bawden et al., 1997), the shallow levels determined for the E-clay proximal to the fault along this section clearly show that the optimal area for active thrust loading of the basin is not under the expected subsidence. Profound sedimentation in Maricopa Basin was controlled by extensional faulting in the early Neogene during slab window opening (Hirst, 1986; Goodman and Malin, 1992; Mahéo et al., 2009), and then in late Pliocene–Quaternary time, interpreted here as the transition into the south-to-north phase of root delamination. These two tectonic forcing regimes provided an accommodation space that underwent profound sediment loading off the delta front of the Caliente River (Figs. 3, 8, 9F, and 10; see also Maricopa Basin subsidence curve presented in Goodman and Malin, 1992).
Rock and surface uplift of the delamination bulge as expressed across the Kern arch and lower Kern gorge area reaches a maximum value of up to ∼2000 m (Mahéo et al., 2009; Cecil et al., 2013). This is considerably more than the ∼1200 m combined delamination-geomorphic forced uplift deduced previously for the bulge along the eastern Sierra. Our transverse model is not applicable to the acute three-dimensional conditions that characterize the southern curved segment of the delamination hinge (Part I, Figs. 3 and 15 therein; Fig. 13 herein). Focusing of the root load along the curved trace of the hinge, and then its rapid release during the south-to-north phase of delamination could account for a greater tectonic component of subsidence across the arch, followed by proportionally greater epeirogenic uplift. The form of the Santa Margarita displacement profiles on Figure 12 suggests a rapid stationary uplift above the delamination hinge, although the limited resolving power of the profiles could miss a rapid migration pattern to the bulge, as predicted in Figure 4B.
Study of the Figure 12 western Sierra Nevada topographic profile and lithospheric structure section reveals that in addition to the Kern arch sitting in the position of the delamination bulge, the central Foothills swell sits in a position analogous to the sympathetic flexural bulge (Fig. 4B). In the (transverse) model profile, the flexural half-node spacing between the center of Tulare Basin and the peak of the sympathetic bulge is ∼135 km. A similar distance is scaled off on Figure 12A northwestward from the center of Tulare Basin. This distance roughly corresponds to the crest area of the central Foothills swell. We posit that as delamination progressed to its Quaternary south-to-north phase, a coupled NE-directed migration of the sympathetic bulge swept beneath the western Foothills and broadly warped the Foothills ramp upwards between 37°N and 38.5°N, forming the central Foothills swell. This broad zone of upward flexure coalesced with the northern shoulder of the principal delamination bulge, rendering the northern wrap pattern of the bulge around northeastern Tulare Basin (Fig. 1 inset; Fig. 6). Propagation of the sympathetic bulge from the Coast Ranges to the central Foothills area during the Quaternary provides a mechanism for the temporary partitioning off of the Great Valley, and the formation of the Corcoran Lake internal basin. The map distribution of Corcoran Lake based on abundant subsurface data from the 710–615 ka Corcoran E-clay (Frink and Kues, 1954; Lettis, 1982, 1988) is shown in Figure 6, and its known subsurface distribution is shown in the structure sections in Figures 5 and 8. The northern limit of the lake parallels the broad culmination of the central Foothills swell, and the age of the lake corresponds in time to constraints presented earlier for growth of the Kern arch. Our hypothesis predicts the contemporaneous rise of the Kern arch and the central Foothills swell, both dynamically linked to the south-to-north phase of delamination. The origin of Corcoran Lake basin has remained enigmatic, with the most widely cited possibilities, i.e., glacial outwash fans or sea-level change controls, encountering problems in timing (Lettis and Unruh, 1991). These authors cite a tectonic control by uplift in the area of the Stockton fault (central Foothills swell) as a little-explored viable possibility, in line with our hypothesis.
Elevation History of the Sierra Nevada
Herein, we have presented an analysis that recognizes regional topographic and bathymetric variation of the Sierra Nevada and Great Valley that dates back to end of Cretaceous time, and to this we have layered on a model for late Miocene to Holocene epeirogenic deformation related to the delamination of the arclogite root of the southern Sierra Nevada batholith. This analysis accounts for many of the first-order physiographic features of the Sierra Nevada microplate, but it is not complete. Supplemental to the Pliocene–Quaternary tilt analysis of the western Foothills ramp, Unruh (1991) suggested that the uplift dynamics of the Sierra Nevada microplate are embedded within a regional epeirogenic field that in general is producing widespread rock uplift distributed across much of the U.S. Cordillera. Such far-field forcing arises from dynamic effects of regional mantle buoyancy structure (Suppe et al., 1975; Lowry et al., 2000). Early Cenozoic plate-margin processes capable of driving regional epeirogeny across the entire microplate include changes in the subduction trajectory of the Farallon plate (Dickinson, 2006; Liu et al., 2010), and the opening and coalescence of multiple slab windows (Atwater and Stock, 1998; Wilson et al., 2005; Humphreys, 2008; Muller et al. 2008). Regional thermomagmatic systems that more profoundly affected the northern Sierra Nevada region include the Yellowstone plume swell, and the ancestral Cascades arc (Suppe et al., 1975; Murphy et al., 1998; Lowry et al., 2000; Pierce et al., 2002; Busby and Putirka, 2009). Deconvolving these potential regional epeirogenic forcing systems, including their decaying phases, is beyond the scope of our analysis, but consideration of their first-order aggregate effect on microplate elevation history is required for our analysis not to be grossly inadequate.
The presence of Upper Cretaceous and Lower Paleogene marine strata in elevated positions at the northern and southern ends of the Sierra Nevada records possible 1000-m-scale rock and surface uplift of at least the southern and northern ends of the microplate over Cenozoic time. These depositional outliers arise as a result of the regional longitudinal paleotopographic gradients that graded to sea level at the two extreme ends of the Sierra Nevada (Fig. 2). The settings of these outliers in relation to stratigraphic relations in adjacent related basins carry implications on such uplift potentially affecting the entire microplate. As much as ∼1200 m of Paleocene–Eocene marine fauna–bearing clastic strata sit nonconformably on axial southern Sierra basement with basal elevations spanning ∼1400 m to ∼1700 m (Fig. 3). There is >1 km of Neogene strata unconformably situated on these strata, much of which has been eroded and transferred into the southeastern San Joaquin Basin (Michael, 1960; Dibblee and Louke, 1970; Wood and Saleeby, 1998; Saleeby and Saleeby, 2010; Table 1 herein). Roughly 1000 m of this uplift can be attributed to early Neogene footwall uplift along the proto–Garlock fault (Fig. 3; and our structural manipulation of Blyth et al.  and Blyth and Longinotti  data). This leaves ∼700 m of mid-to-late Cenozoic surface uplift from another mechanism(s). The area of the elevated Eocene marine strata corresponds to the low-lying Late Cretaceous extensional terrane that continued southward into the western Mojave plateau region. Presently, the plateau floor lies at ∼800 m elevation. We interpret the plateau to have gained elevation in conjunction with the additional rise of the southern Sierra marine strata. Far-field forcing for this uplift regime may have arisen from shallow low-velocity mantle beneath the region (Reeg, 2008; Schmandt and Humphreys, 2010; C.H. Jones, 2012, written commun.), reflecting upwelling along the southern margin of the Isabella anomaly, and/or the lingering of shallow asthenosphere emplaced into the Pacific-Farallon slab window.
Upper Cretaceous marine strata, typically deltaic facies, are nonconformable across northwestern Sierra Nevada basement, with basal elevations ∼100–400 m (Repenning, 1960; Jenkins, 1977; Dickinson et al., 1979; Harwood et al., 1981). Isopach patterns along the northeast Sacramento Basin indicate that the Upper Cretaceous strata were up to ∼1000 m thick, indicating 1000-m-scale regional uplift of the entire northern microplate in Cenozoic time. The Upper Cretaceous strata are conformably overlain by relatively thin Paleocene to Upper Eocene terrestrial and marine strata in proximal parts of the basin, above which there is a latest Eocene to middle Miocene hiatus across much of the basin, and which along much of the Foothills ramp completely removed the Upper Cretaceous strata. The position and timing of this hiatus are such that it likely represents an epeirogenic deformation progression forced in sequence by: (1) accretion of the Siletzia large igneous province; (2) opening of the Siletzia/Farallon-Kula slab window; and then (3) initial uplift along the southern margin of the Yellowstone plume swell (Wells et al., 1984; Engebretson et al., 1985; Murphy et al., 1998; Lowry et al., 2000; Pierce et al., 2002; Muller et al., 2008; Humphreys, 2008; McLaughlin et al., 2009). Immediately following this regime, there was an ∼13 m.y. period of ancestral Cascade arc volcanism, starting at ca. 16 Ma, with eruptive centers clustered along the eastern Sierra crest as far south as ∼38°N (Fig. 2; Busby and Putirka, 2009). The volcanic stratigraphy records multiple pulses of rock uplift and west tilt initiating in middle Miocene time (Busby and Putirka, 2009), indicating that some component of west tilt initiated prior to the more widely recognized Pliocene–Quaternary phase (Unruh, 1991). Thermal expansion of the eastern Sierra region related to ancestral Cascades arc volcanism north of 38°N is another potential source for late Cenozoic transient uplift of the region (Busby and Putirka, 2009). The possibilities for multiple uplift transients having migrated through the northern Sierra Nevada region in Cenozoic time, including their negative decaying phases, and the resulting complexity on regional elevation patterns or river incision imprints are little explored.
A critical factor to this analysis is that evidence for the existence or survival of the arclogite root for the Sierra Nevada batholith is restricted to the region south of ∼38°N, as is evidence for its Pliocene–Quaternary removal (Ducea and Saleeby, 1998a, 1998b; Manley et al., 2000; Farmer et al., 2002; Saleeby et al., 2003; Part I, Figs. 3 and 4 therein). In contrast, the Neogene ancestral Cascades arc terminates southward at ∼38°N (Fig. 2), underscoring the geodynamic as well as the physiographic contrasts between the northern and southern Sierra Nevada. This suggests that the southern Sierra Nevada arclogite root and its substrate of wedge peridotite, which together had been quenched to a lithospheric geotherm in the Late Cretaceous (Ducea and Saleeby, 1996; Saleeby et al., 2003), inhibited the southwest sweep of the Neogene arc from Nevada into the eastern Sierra Nevada region south of 38°N (after Dickinson, 2006; Busby and Putirka, 2009). Compounded on this longitudinal contrast along the microplate, there are the effects of the ≥16 Ma initiation of the Yellowstone plume in a position proximal enough to the northern Sierra Nevada to place much of the region within the 1000 m plume swell isopleth (Lowry et al., 2000; Pierce et al., 2002). The ca. 16 Ma Lovejoy (plume) basalt erupted in the northern Sierra Nevada region further indicates proximity to the plume head (Garrison et al., 2008). Subsequent migration of the plume in theory has resulted in ∼300 m of subsidence of the northern Sierra Nevada region off the southwest margin of the plume swell.
The survey of far-field forcing mechanisms presented here suggests two principal components: (1) a SW Cordilleran-wide system that operated on the entire microplate arising from Farallon slab subduction dynamics, and the opening and coalescence of multiple slab windows; and (2) thermal-magmatic forcing of the Yellowstone plume swell and ancestral Cascades arc, which was focused in the northern Sierra Nevada region. It seems likely that epeirogenic forcing in the Sierra Nevada region by the southward opening of the Siletzia/Farallon-Kula slab window had decayed to zero by Neogene time (Humphreys, 2008). This suggests that the principal drivers of late Cenozoic far-field epeirogeny across the Sierra Nevada microplate were the Yellowstone plume swell in the north and the Pacific-Farallon slab window in the south, with the coalescence and overlap of these two systems initiating at ca. 10 Ma at ∼38°N, and progressing northward to the modern Mendocino triple junction (Atwater and Stock, 1998; Pierce et al., 2002; Wilson et al., 2005). These two systems together are suggested to have raised the entire axial to eastern Sierra Nevada region at 500–1000 m scale and the western Foothills region at 100–200 m scale. Focused uplift arising from the ancestral Cascades arc is likely to have been additive to the plume swell uplift, and then to have thermally decayed as the arc progressively shut off northward with the migration of the triple junction, just as the plume swell effect was decaying as the plume migrated eastward (Lowry et al., 2000; Pierce et al., 2002; Busby and Putirka, 2009).
A more subtle but potentially profound effect that the ancestral Cascades arc had on Sierra Nevada physiographic evolution was to blanket the northern Sierra Nevada with Neogene volcanic strata, and raise the Neogene base level of the Sacramento Basin by the building out of a terrestrial volcaniclastic apron (Busby and Putirka, 2009). This preserved early Cenozoic paleolandscape features in the northern Sierra Nevada, accentuating physiographic contrasts with the southern Sierra Nevada, where early Cenozoic landscape features lay bare through the Neogene, rendering them susceptible to continued erosional modification. In Figure 2, local paleorelief that dates back to end of Cretaceous time is color coded across the Sierra Nevada. These could represent minimum values, particularly in the northern Sierra Nevada. River canyons north of the Stanislaus drainage that incise into basement below Neogene reference surfaces (Wakabayashi and Sawyer, 2001; Wakabayashi, 2013) are not well constrained on the timing(s) of such basement incision, and thus some nontrivial component of the observed basement incision could also date back to the paleorelief regime that is coded on Figure 2. Deep basement channels inherited from either the margin of the Nevadaplano from end of Cretaceous time and/or forced by the early Cenozoic opening of the Siletzia/Farallon-Kula slab window could have laid buried beneath the ancestral Cascades volcaniclastic apron, to be reincised into and likely deepened sometime following the ca. 5 Ma disruption of the volcanic source by westward encroachment of the eastern escarpment system, and northward migration of the Mendocino triple junction. In this context, the Neogene reference surfaces used to document basement incision could be remnants of volcanic strata perched on ancient strath terraces cut into basement. That is to say, the topographically lowest preserved Neogene, and locally Oligocene, volcaniclastic rocks may not mark the deepest point in observed paleochannels, but rather higher levels of the paleochannel fills as well as interchannel cover. In this scenario, the deepest parts of the canyon fills were completely stripped of their Neogene deposits in late Pliocene–Pleistocene time. This underscores the fact that there are currently no direct constraints on the age of basement incision north of the Stanislaus River.
The analysis presented here recognizes the importance of the physiographic contrasts between the northern and southern Sierra Nevada. It also points out the likelihood of fundamental lithospheric structural contrasts between the two regions that date back to end of Cretaceous time, and which strongly controlled the Neogene–Quaternary physiographic development of the two regions. Nevertheless, the entire microplate appears to have responded comparably in terms of regional uplift to the integrated far-field forcing mechanisms recognized. Layered on top of this regional epeirogeny, the ancestral Cascades arc has rendered a strong contrast in the preservation state of paleolandscape features, and the delamination of the southern Sierra Nevada batholith arclogite root has driven profound focused epeirogeny across the southern microplate. These two regimes underscore fundamental ancient lithospheric contrasts between the northern and southern Sierra Nevada (Part I, Fig. 4 therein) and account for distinct physiographic differences between the two regions that were pointed out by early investigators (Matthes, 1965; Christensen, 1966).
PALEOGEOGRAPHIC EXPRESSIONS OF MANTLE LITHOSPHERE REMOVAL
The late Cenozoic paleogeography of the Sierra Nevada microplate is accented by a number of features that appear to be direct expressions of the progression of underlying mantle lithosphere removal. These are summarized here. Special attention is placed on the tectonics and paleogeography that preceded and influenced root delamination and its surface expressions, as well as our kinematic model for three-dimensional delamination in the southern Sierra region since ca. 5 Ma (Fig. 13).
Early and middle Miocene extensional tectonism, deepening of the San Joaquin Basin, and formation of the Walker graben occurred in crust that had undergone extension and tectonic erosion of its underlying mantle lithosphere during the Late Cretaceous, above the lateral ramp in the Rand subduction megathrust (Figs. 1, 2, and 3; Part I, Fig. 4 therein). The Miocene phase of extension was forced by the northward opening of the Pacific-Farallon slab window (Atwater and Stock, 1998; Wilson et al., 2005), which, according to our model, also instigated mantle lithosphere mobilization beneath the southern Sierra Nevada region. The intensity of Miocene extension dropped significantly northward across the ∼35.5°N limit of Late Cretaceous large-magnitude extension (Figs. 2 and 3), mimicking lithospheric structure that had been established in the Late Cretaceous. Pliocene (?)–Quaternary extension affected the same area as the Miocene extension, but it also extended northward through the area of the northern Kern arch and Tulare Basin, as recoded by numerous high-angle normal faults (Nugent, 1942; Croft and Gordon, 1968; Saleeby and Foster, 2004, their Data Repository Map; Figs. 3, 5, and 8). The lateral ramp in the Rand subduction megathrust left the arclogite root of the Sierra Nevada batholith truncated and liable for delamination. As discussed in Part I, the added gravitational potential of the high-standing Nevadaplano foreland to the Sierra Nevada batholith, as opposed to the lower potential of the southern Sierra–Mojave extended terrane (Part I, Fig. 4 therein), favored the east-to-west over south-to-north initial pattern in delamination, once the slab window perturbation was regionally imposed. As discussed in Part I, and herein, this required special circumstances to promote the current south-to-north delamination regime.
The 10 ± 2 Ma (late Miocene) mantle lithosphere separation, and derivative inception of the Sierra Nevada microplate (Fig. 4A), is expressed by stratigraphic relations along much of the Great Valley, as well as across the southern end of the microplate. This event is interpreted to mark the initiation of regional west tilt to the microplate, which accelerated in Pliocene time (Unruh, 1991). Sediment provenance, dispersal, and strandline relations changed abruptly across the entire San Joaquin Basin and into the southern Sacramento Basin at the time of microplate inception, with a marine incursion northward along the western margin of the microplate, and the deposition of the extensive “Santa Margarita” sand sheet(s) (Fig. 2). The granitic provenance of the Santa Margarita detritus signals the rejuvenation of major Sierra Nevada drainages in progression to the widely recognized Pliocene phase of river incision. In the southernmost Sierra Nevada, rise of the eastern crest region is recorded by the south and eastward delivery of copious Sierran detritus into the El Paso Basin (Fig. 3; and Loomis and Burbank, 1988), and the westward redistribution of the Walker graben fill into the southern San Joaquin Basin by the Caliente River (Fig. 3; Saleeby and Saleeby, 2010; Saleeby et al., 2013). Voluminous sediment loading from the redistributed Walker graben fill drove profound latest Miocene to early Quaternary subsidence in the Maricopa Basin, and the adjacent area that became the Kern arch, in conjunction with N-side-down normal displacement on the White Wolf fault (Figs. 8B and 8C). The Caliente River delta front issued submarine fans into the San Joaquin Basin, which served as important petroleum traps (MacPherson, 1978; Harrison and Graham, 1999). The results of our thermomechanical modeling (Fig. 4B; Part I, Figs. 11 and 13 therein) further indicate that during the structural progression to lithosphere separation, the load of the arclogite root drove ∼300 m of subsidence along the eastern Sierra Nevada–Basin and Range transition. This provided accommodation space for the development of late Miocene–Pliocene lake formation along the transition zone (Bachman, 1978; Bacon et al., 1982), which was subsequently disrupted by extensional faulting and eastern Sierra crest uplift as root delamination accelerated through Pliocene time.
Figure 13 diagrammatically shows renderings of our model of arclogite root delamination in three dimensions since ca. 5 Ma, in relation to regional paleogeographic patterns. The top of the root is viewed along Moho levels from the southeast. Surface-level views are lifted off the Moho view in order to render delamination-related physiographic changes through time. In Figures 13A to 13C (5 Ma to 2 Ma), delamination is shown to have driven Pliocene rock uplift of the southern Sierra Nevada, as is clearly expressed by accelerated river incision and west tilting (Huber, 1981; Unruh, 1991; Wakabayashi and Sawyer, 2001; Stock et al., 2004, 2005), and by the exhumation of the Walker graben fill (Fig. 3). Accelerated uplift of the Sierra Nevada over this time period is paralleled by accelerated or anomalous subsidence in the San Joaquin Basin.
The results of our thermomechanical modeling suggest that a total of ∼500 m of anomalous subsidence was first focused along the western Sierra Foothills, and then subsidence migrated westward to the center of the basin as the delamination bulge grew and expanded into the Foothills region (Fig. 4B). The model predicts a rapid reversal from subsidence to rock uplift within the model time period that we correlate to the late Pliocene (∼17–20 m.y.). The apparent ease with which kilometer-scale exhumation of the Pliocene–lower Quaternary sediments of the Kern arch has occurred since 1 Ma (Fig. 10) suggests a low preservation potential of correlative sediments that may have aggraded across the Foothills during the initial subsidence phase. The position of the top of the upper Pliocene San Joaquin Formation, relative to the position of the exhumed nonconformity/basement pediment surface to the east (Fig. 4), indicates that the shallow-marine conditions represented by the formation could have easily spread eastward across the entire low-relief surface as a result of the (modeled) early phase of subsidence of the Foothills. This is in line with the unique sedimentology and geomorphic setting of the eastern reaches of the San Joaquin Formation. Upper Miocene–lower Pliocene strata of the eastern San Joaquin Basin are dominantly sand rich, representing shallow-marine and deltaic facies. During late Pliocene time, this changed dramatically to fine-grained green siltstones and mudstones of the San Joaquin Formation, signaling sediment starvation. This sedimentological change to relatively low-energy conditions immediately adjacent to the western Sierra occurred precisely when nearby documented river incision rates were greatest (Stock et al., 2004, 2005). The forms of the 14–20 m.y. modeled displacement curves (Fig. 4B) are such that river gradients could have increased along the axial to eastern Sierra, accelerating main channel incision, while the Foothills region was under subsidence and very likely marine conditions. Our petrographic studies show that the eastern San Joaquin Formation is characterized by tonalitic gruss layers within a green siltstone matrix. A simple explanation for this distinct detrital association is that marine conditions with limited sediment transport capability spread across the mainly tonalitic basement pediment that lies along the east margin of the basin. The larger sediment dispersal events of this unique ephemeral environment correspond to the spreading of proximal basement-derived gruss into the shallow basin. We posit that the temporary low-energy conditions that developed so proximal to the uplands resulted from the initial subsidence signal suppressing western Sierra river gradients as subsidence temporarily embayed into the uplands. This may have resulted in a brief period of profound aggradation, the record of which was erosionally erased as the delamination bulge subsequently expanded into the western Sierra region. Alternatively, the predicted subsidence-uplift transient for the Foothills area may have been dampened by the necking off of an adjacent root fragment at ca. 3.5 Ma, as discussed next.
In Figure 13, it is further hypothesized that the necking off and foundering of the northeast portion of the actively delaminating root profoundly affected subsequent kinematic patterns of delamination, as well as its surface expressions. The physical conditions that favored the initial east-to-west phases of delamination, as opposed to south to north, and the physical rationale and evidence for hypothesizing the necking off and foundering of a fragment of the root are covered in Part I. Possibly the clearest surface expression of the foundering of this fragment is the overlying pulse of ca. 3.5 Ma delamination volcanism that uniquely affected the corresponding area of the Sierra Nevada (Fig. 13B; Manley et al., 2000; Farmer et al., 2002). Following the detachment of the foundered root fragment, the residual root was configured in such a way as to transition into the south-to-north delamination of its southern end. The prior history of east-to-west delamination had, according to our model tests (Part I, Fig. 10 therein), established the lower-crustal channel across the entire arclogite root–felsic batholith interface, whereupon the release of the partially delaminated limb, which also inhibited south-to-north delamination, and the resulting concentration of the root load to the south instigated the south-to-north delamination phase. This transition caused W-side-up normal faulting of the eastern escarpment system to step over to the Greenhorn–Breckenridge–Kern Canyon system, which followed the Late Cretaceous basement damage zone of the Kern Canyon fault (Mahéo et al., 2009; Saleeby et al., 2009a; Amos et al., 2010; Nadin and Saleeby, 2010). This stepover in W-side-up normal faulting developed in conjunction with the west tilt and uplift pattern of the Kern arch (Figs. 3 and 13D).
The Quaternary rise of the Kern arch profoundly affected depositional and geographic patterns of the eastern San Joaquin Basin (Fig. 13D). The Tulare and Maricopa subbasins were partitioned off from the principal San Joaquin Basin depocenter at this time. The distribution of the Corcoran E-clay in the subsurface (Figs. 6 and 8) suggests that growth of the Kern arch pushed the southeastern shore of Corcoran Lake westward, leaving the southern part of Corcoran Lake embayed eastward in Tulare Basin. The analysis presented here for Figure 12 longitudinal relationships posits that similar dynamics were at work driving late Quaternary crustal deformation along the eastern San Joaquin Basin–western Foothills transition as a result of south-to-north delamination as those that operated during east-to-west delamination. Specifically, the Kern arch corresponds to the delamination bulge, and the central foothills swell corresponds to its sympathetic bulge (Fig. 4B). It follows that a late Quaternary axis of maximum delamination-driven subsidence should follow a transverse trend through Tulare Basin. The area of this axis could correspond to the maximum depth of the Corcoran E-clay on the Figure 8 structure sections, as well as the veneer of Quaternary fluvial sediments that have aggraded across the exhumed ancient pediment surface and up into the lower Kaweah drainage (Figs. 3 and 5A; Saleeby et al., 2013). Rise of the central Foothills swell is hypothesized to have temporarily partitioned off the southern Great Valley as the internal Corcoran Lake basin (Fig. 6), and ultimately promoted the current broadened zone of incision of Pliocene and Quaternary strata along the central segment of the Foothills ramp. Growth of the Kern arch and the building out of the Kings River alluvial fan (Atwater et al., 1986) have left the residual Corcoran Lake basin as the partitioned Tulare and Buena Vista and Kern lakebeds (Fig. 6).
Uplift of the southern Coast Ranges is also suggested to be in part a result of mantle lithosphere removal dynamics. The central segment of the San Andreas fault, in the southern Coast Ranges, is shown to lie on the relative plate-motion trajectory between the Pacific and North American plates (Argus and Gordon, 2001), calling into question the common interpretation that recent phases of the Coast Range uplift result from plate-margin–driven transpression. The results of our modeling indicate that up to ∼500 m of total rock uplift of the southern Coast Ranges results from regional flexure that is sympathetic to the Sierran delamination bulge (Fig. 4B). Furthermore, excess gravitational potential created by the Sierran uplift as well as extension along the eastern Sierra–Death Valley province are sufficient to drive crustal shortening across the western margin of the microplate (Jones et al., 1996, 2004; Lowry et al., 2000), expressed as blind thrusts rooted westward beneath the Coast Ranges (Fig. 1).
SUMMARY AND CONCLUSIONS
The historical debate as to whether the high elevations of the Sierra Nevada are an ancient feature inherited from the Late Cretaceous or result from geologically recent uplift has continued into the current research era (cf. House et al., 1998, 2001; Poage and Chamberlain, 2002; Wakabayashi and Sawyer, 2001; Jones et al., 2004; Mulch et al., 2006; Busby and Putirka, 2009; Cassel et al., 2009, 2012; Molnar, 2010; Wakabayashi, 2013). Rarely is the question of transient vertical displacements considered in this debate. This is not surprising in that such transients by nature include potentially significant cryptic signals. Furthermore, a dynamic framework in which to consider such transients has not been widely recognized. We have attempted to leverage a diversity of constraints on this problem, including stratigraphy and geomorphology across the Sierra Nevada microplate, geophysical imaging of the underlying upper mantle, and dynamic modeling of the partial removal of its mantle lithosphere, in order to better resolve the principal components of Sierra Nevada elevation history. We find considerable evidence for significant regional and local paleorelief throughout the microplate dating back to end of Cretaceous time, with such paleorelief still strongly expressed in microplate physiography and basin structure. We also find evidence for far-field tectonic forcing that elevated the entire microplate at kilometer scale in the Cenozoic and contributed to its regional west tilt. The far-field forcing of the Yellowstone plume head and multiple slab window encounters, as well as the thermal swell of the ephemeral ancestral Cascades arc, probably carried both positive and negative vertical displacement transients, confounding their elevation change and river incision imprints on the Sierra Nevada. Furthermore, numerical modeling of low-temperature thermochronometric and depth of basement exhumation data suggests kilometer-scale east-central Sierra Nevada region elevation decay during early to mid-Cenozoic time, followed by similar-scale late Cenozoic rock and surface uplift that we derive from combined delamination, geomorphic, and far-field forcing mechanisms (McPhillips and Brandon, 2012). These profound model predictions are not at odds with our integrated physiographic-geodynamic model, nor are they at odds with the primary observations cited for the contemporary dominance of paleorelief versus the dominance of geologically recently generated relief.
Resolution of the far-field–forced and paleorelief components of Sierra Nevada microplate physiography facilitates the resolution of more local epeirogenic transients related to mantle lithosphere removal. Our procedure for the resolution of these local components is driven by iterations in thermomechanical modeling and comparison of model predictions for surface responses with stratigraphic and geomorphic data. Model iterations encompassing lithospheric structure and composition and geologic history have led to the bundling of model constraints and predictions into consistency with the principal surface-level observations. Our preferred model in terms of lithospheric structure and compositional constraints predicts ∼800 m of Pliocene–Quaternary eastern Sierra rock uplift, driven by the delamination of the arclogite root that formed beneath the southern Sierra Nevada batholith, and ∼680 m of contemporaneous tectonic subsidence in the Great Valley where the mantle lithosphere root remains attached to the crust. The predicted tectonic subsidence compares well with the ∼625 m of anomalous tectonic subsidence resolved stratigraphically in the Tulare Basin of the Great Valley. Comparisons between model predictions and observations for eastern Sierra crest rock uplift are more complex due to exhumation during uplift. We derive a geomorphically forced rock uplift component of ∼400 m (after Small and Anderson, 1995; Pelletier, 2007), stimulated by delamination forcing, summing to ∼1200 m of southeastern crest uplift. This alone is short of the estimated 2000 ± 500 m of Pliocene–Quaternary rock uplift of the eastern crest based on stratigraphic-geomorphic geometric constructions (Huber, 1981; Unruh, 1991; Wakabayashi and Sawyer, 2001), although adding in kilometer-scale, far-field–forced regional uplift brings these into agreement, assuming some component of west tilt accompanied the regional uplift pattern. In addition to these reasonably favorable quantitative comparisons between model predictions and stratigraphic-geomorphic data, we find that distinct variations in the regional morphology of the Sierra Nevada–Great Valley basement surface are favorably predicted by our modeling as well. This lends further support to our conceptual model that the principal physiographic differences between the northern and southern Sierra Nevada microplate arise from the superposing of delamination-related epeirogeny primarily across the southern reaches of the microplate. These results, backed by close attention to geologic structure and history in the formulation of our thermomechanical models, lend confidence that this modeling is reasonably approximating the underlying dynamics for the southern Sierra Nevada case of mantle lithosphere removal.
This research was supported by National Science Foundation grants EAR-0230383 and EAR-0606903, a grant from the George and Betty Moore Foundation, and Faculty Research Funds from the Université Pierre et Marie Curie, France. This is Caltech Tectonics Observatory Contribution 142. Conversations and field excursions with C.J. Busby, J. Gillespe, S.A. Graham, J.S. Lewis, D.D. Miller, K. Putirka, A.G. Sylvester, and J. Wakabayashi helped enrich this work. We also acknowledge the editorial efforts of J.R. Unruh, C.H. Jones, and D.L. Harry in greatly improving the clarity of this publication.