Abstract

The late Cenozoic stratigraphic and tectonic history of the Santa Clara Valley illustrates the dynamic nature of the North American–Pacific plate boundary and its effect on basin and landscape development. Prior to early Miocene time, the area that became Santa Clara Valley consisted of eroding Franciscan complex basement structurally interleaved in places with Coast Range ophiolite and Mesozoic Great Valley sequence, and locally overlapped by Paleogene strata. During early to middle Miocene time, this landscape was flooded by the sea and was deformed locally into deeper depressions such as the Cupertino Basin in the southwestern part of the valley. Marine deposition during the middle and late Miocene laid down thin deposits in shallow water and thick deeper-water deposits in the Cupertino Basin. During this sedimentation, the San Andreas fault system encroached into the valley, with most offset partitioned onto the San Andreas fault southwest of the valley and the southern Calaveras–Silver Creek–Hayward fault system in the northeastern part of the valley. A 6-km-wide right step between the Hayward and Silver Creek faults formed the 40-km-long Evergreen pull-apart basin along the northeastern margin of the valley, leaving a basement ridge between it and the Cupertino Basin. The Silver Creek fault was largely abandoned ca. 2.5 Ma in favor of a compressional left step between the Calaveras and Hayward fault, although some slip continued to at least mid-Quaternary time. Gravity, seismic, stratigraphic, and interferometric synthetic aperture radar (InSAR) data indicate no other major San Andreas system faults within the central block between the present-day range-front faults bounding the valley and the Silver Creek fault. Sometime between 9 and 4 Ma (9 and 1 Ma for the central block), the area rose above sea level, and a regional surface of erosion was carved into the Mesozoic and Tertiary rocks. Alluvial gravels were deposited on this surface along the margins of the valley beginning ca. 4 Ma, but they may not have prograded onto the central block until ca. 1 Ma, because no older equivalents of the Pliocene–Quaternary Santa Clara gravels have been found there. Thus, either the central block was high enough relative to the surrounding areas that Santa Clara gravels were never deposited on it, or any Santa Clara gravels deposited there were stripped away before ca. 1 Ma. Analysis of alluvium on the central block implies a remarkably uniform, piston-like, subsidence of the valley of ∼0.4 mm/yr since ca. 0.8 Ma, possibly extending north to northern San Francisco Bay. Today, the central block continues to subside, the range-front reverse faults are active, and the major active faults of the San Andreas system are mostly outside the valley.

INTRODUCTION

Santa Clara Valley, located within the broad San Andreas fault system in northern California (Figs. 1 and 2), hosts a population of nearly 2,000,000 people (U.S. Census Bureau, 2010). Groundwater from the valley provides this population with nearly half of its water supply. The stratigraphy and structure beneath this highly urbanized valley provide a foundation for defining the boundaries of the groundwater flow system because aquifer systems are a product of the interplay of depositional and deformational processes through time. The stratigraphy and structure of the valley also provide a framework for assessing seismic hazard by mapping locations of faults concealed beneath urban development and young surficial deposits, and by defining concealed basins that can amplify and prolong shaking from local and distant earthquakes.

This paper summarizes the late Cenozoic stratigraphic and tectonic history of Santa Clara Valley as inferred from geologic, stratigraphic, and geophysical data. This summary builds on the more detailed and comprehensive papers in this theme issue and is illustrated by a series of schematic cross sections (Fig. 3). Interpretations of the multiple geologic and geophysical data sets indicate that the history of Santa Clara Valley is a tale of three sedimentary basins. The youngest basin, called the Santa Clara Basin, forms the present-day shape of the valley floor and during the past 1–1.5 m.y. has been accumulating debris from the Santa Cruz Mountains on the west and the Diablo Range on the east. This alluvial basin conceals two deep late Cenozoic basins that are revealed by analysis of geophysical, primarily gravity, data. One is the Cupertino Basin in the southwestern part of the valley, which records transtension associated with the passage of the Mendocino triple junction and development of the San Andreas fault. The other is the Evergreen Basin in the eastern part of the valley, which is the product of a right step within the East Bay fault system. The margins of these concealed basins have been largely overridden by thrust and reverse faults along the western and eastern edges of the valley. The history of these three basins beneath the urbanized Santa Clara Valley provides insight into the way in which the San Andreas fault system has developed in this region, illustrating the dynamic nature of the plate boundary and its effect on basin and landscape development.

REGIONAL GEOLOGIC SETTING

Santa Clara Valley, a broad, mostly flat alluvial plain extending southward from San Francisco Bay, is situated within the San Francisco Bay block (blue shaded region in Fig. 2). We restrict our discussion to that part of Santa Clara Valley northwest of Coyote Narrows (CN in Fig. 2). The San Francisco Bay block is bounded by major right-lateral faults, the San Andreas fault on the southwest and the Hayward and Calaveras faults on the northeast. The San Andreas fault has ∼300–330 km of right-lateral displacement since 23 Ma based on offsets of volcanic fields, shorelines, and deep-sea fans (Hill and Dibblee, 1953; Huffman, 1972; Matthews, 1976; Stanley, 1987a; Graham et al., 1989; Sharman et al., 2013). However, the segment of the San Andreas fault immediately west of Santa Clara Valley (Santa Cruz segment) has significantly less offset, ∼124 km (Jachens et al., 1998), and the Peninsula segment north of the intersection with the Pilarcitos fault has even less displacement, 28–30 km of right-lateral displacement since ca. 3.3 Ma (Dibblee, 1966; McLaughlin et al., 1996, 2007). Bedrock offset determined from analysis of gravity and magnetic anomalies suggests even less offset on the Peninsula segment, ∼22 km (Jachens and Zoback, 2000). The discrepancy between 300 and 330 km of displacement on the San Andreas fault documented along the central part of the San Andreas fault and 124 km of displacement along the Santa Cruz Mountains segment is reconciled by adding displacement taken up on faults east of the San Francisco Bay block. The Hayward and Calaveras faults are part of the East Bay fault system, and they have 160–190 km of cumulative right-lateral displacement (McLaughlin et al., 1996) since ca. 12 Ma (Graymer et al., 2002). This displacement is further refined by offset gravity and magnetic anomalies to ∼174 km (Jachens et al., 1998).

The early development of the San Andreas fault system was accompanied by slab-window volcanism in California and northern Mexico starting about 28–27 Ma that resulted from the deaths of a series of spreading ridge segments during piecewise destruction of an older subduction regime (Atwater, 1970; Wilson et al., 2005; McCrory et al., 2009). In the vicinity of the Santa Clara Valley, the development of discrete faults and northwestward-younging volcanism may have begun about 18–15 Ma in the wake of a slab window that accompanied the northwestward-migrating Mendocino Triple Junction (Dickinson and Snyder, 1979; Johnson and O’Neil, 1984; Fox et al., 1985; Stanley, 1987b; McLaughlin et al., 1996).

The distribution of basement beneath Santa Clara Valley is not just the result of strike-slip faulting and transtension, but also earlier subduction. The basement within the San Francisco Bay block consists of two main, often complexly interleaved coeval Mesozoic packages, the Franciscan complex (a subduction complex) and the Coast Range ophiolite with its overlying sedimentary section, the Great Valley sequence (forearc-basin complex) of Bailey et al. (1964). Along the east side of the Coast Ranges for a distance of 600 km, these two packages are juxtaposed by the Coast Range fault, a subduction boundary (Jayko et al., 1987) highly modified by the roof thrust of a tectonic wedge inboard of the fossil subduction megathrust (Wentworth and Zoback, 1990; Jachens et al., 1995) and/or by attenuation faulting related to collapse of the accretionary prism and unroofing of the low-temperature–high-pressure metamorphic rocks in the Early Tertiary (Harms et al., 1992). About 5 km south of Los Gatos, the Aldercroft fault is the principal boundary between rocks of the Coast Range ophiolite (called at this location the Sierra Azul ophiolite) and the Franciscan complex, occupying the same structural position as the Coast Range fault (McLaughlin and Clark, 2004). In parts of the San Francisco Bay block, Coast Range ophiolite and Great Valley sequence rocks are known to be interleaved with Franciscan complex rocks, suggesting that the simple relations displayed by the Coast Range fault have been modified by subsequent deformation.

Next, we summarize our reconstruction of the tectonic and stratigraphic history of Santa Clara Valley, aided by schematic cross sections shown in Figure 3, which show the geometry of basins and faults at various time intervals and by palinspastically restored map views (Fig. 4).

Pre–18 Ma

In early Miocene time, the area that became Santa Clara Valley (the area currently between the San Andreas fault and the East Bay fault system) consisted of uplands of Franciscan complex basement structurally overlain in places across the Coast Range fault by Coast Range ophiolite and Mesozoic Great Valley sequence and overlying Lower to Middle Eocene strata. Magnetic modeling (Jachens and Griscom, 2004) suggests that slices of the Coast Range ophiolite (simplified on Fig. 3) are the exhumed equivalents of eastward-tapering tectonic wedges that underlie the western parts of the Great Valley sequence and are imaged by seismic-reflection data along the Coast Ranges–Great Valley margin (Wentworth and Zoback, 1990). The Oligocene to early Miocene Temblor Formation, exposed along the southwest margin of the Santa Clara Valley, contains locally derived angular detritus from uplifted Franciscan complex and Coast Range ophiolite and from the Eocene sedimentary sequence in the Sierra Azul block (McLaughlin et al., 1991a, 1991b). Fault reconstructions of the San Francisco Bay block (Jachens et al., 1998) place Santa Clara Valley east of the Vizcaino block before 15 Ma (today the Vizcaino block lies north of Point Arena; Fig. 1) and opposite that part of the Great Valley block to the east that is today between Palo Prieto Pass and Priest Valley (Figs. 1 and 4). Local uplift and erosion during this time may have been associated with low-angle attenuation faulting northeast of the San Andreas fault, followed by reverse faulting along the Sargent fault 10–17 Ma (McLaughlin and Clark, 2004). The topography of the area that now is Santa Clara Valley probably resembled the present-day Santa Cruz Mountains or Diablo Range to the northwest, with broad areas of shallow marine deposition, and local areas of deep marine deposition (Fig. 4B).

Circa 18–15 Ma

Exactly when the San Andreas fault reached as far north as the Santa Clara region is only loosely constrained by the age of volcanism and hydrothermal alteration, presumably formed in the wake of the Mendocino triple junction, because the timing of transfer of transform plate-boundary slip inboard to the San Andreas fault is not well known. In the hills east of the San Andreas fault, ca. 16–14 Ma volcanic rocks mark the passage of the triple junction, and possibly the advent of faulting to the west of the valley and displacement of valley rocks relative to rocks west of the San Andreas fault.

During early to middle Miocene time, this landscape further subsided beneath a shallow sea and was deformed locally into deeper depressions such as the Cupertino Basin in the southwestern part of the valley. Miocene strata exposed along the southwest margin of the valley indicate marine deposition at depths of ∼150–1500 m (Stanley et al., 2002), and deepening of that basin migrated northward between Gilroy and Cupertino between the early and middle Miocene (McLaughlin et al., 1999; Fig. 4B).

Multiple lines of evidence point to the existence of the Cupertino Basin. Driller’s logs from a minor oil field in the Los Gatos area indicate a thick (780 m) section of brown shale interpreted to be Monterey Formation; geochemistry from an oil sample from this field indicates derivation from a Monterey Formation source rock generated at a depth of ∼2.1–2.7 km (Stanley et al., 2002). Gravity data indicate that the basin is ∼3 km deep, 30 km long, and ∼10 km wide, with a steep southwest margin and gently sloping northeast margin (Langenheim et al., 1997; Stanley et al., 2002). A seismic-refraction profile (Catchings et al., 2006) confirms the presence of a deep basin in the Los Gatos area, with velocities of <5 km/s extending to depths of 3 km. Although Catchings et al. (2006) interpreted the base of the basin at 1.6–1.8 km depth based on the isovelocity contour of 3.5 km/s, velocity-depth relationships for the various Bay Area rock types indicate that Franciscan complex basement rocks should have velocities of 4.5 km/s or higher at these depths (Brocher, 2008). A north-south–oriented seismic-reflection profile through Santa Clara and Cupertino (Williams et al., 2004) indicates an apparent south dip to the basin sequence, which is consistent with the sense of tilting in the hanging wall of a northeast-dipping master normal fault. The asymmetry of the basin (Fig. 3) suggests that it formed as a half graben.

Initial subsidence and sedimentation in the Cupertino Basin were likely the result of transtension accompanied by northward-younging volcanism and hydrothermal activity. Along the western margin of Santa Clara Valley, there are the 14.8 Ma Page Mill Basalt near Palo Alto, a 15.6 Ma dacite tuff and local intrusive rocks in the New Almaden area, and 17.6 Ma hydrothermal veining near Gilroy (McLaughlin et al., 1996). These ages suggest that the formation of the basin initiated 15–18 Ma. Graymer et al. (2015) favor an alternate interpretation in which the Cupertino Basin formed later after eruption of the dacitic volcanics, based on regional correlation of the Temblor Formation with equivalent rocks to the east after palinspastic restoration and the presence of Monterey petroleum source rocks near the base of the basin fill.

Deposits of the Monterey Formation may have been widespread in Santa Clara Valley. Its deposition also appears to be transgressive, with basal Monterey strata younging to the northwest from Gilroy to Los Gatos. The Cupertino Basin may have extended southeast of its current configuration as defined by its gravity low; Graymer et al. (2015) speculates that a bench in the gravity field southeast of San Jose may reflect Miocene basin deposits that are now structurally concealed (Fig. 4B). The Monterey Formation (and its equivalent, the Claremont Shale, east of the Hayward fault) may also be present elsewhere within the San Francisco Bay block, such as in the east-dipping sedimentary package below an unconformity seen in seismic-reflection profiles in the San Leandro area (Marlow et al., 1999).

Circa 12 Ma

Marine sedimentation continued even as the faults of the San Andreas system displaced the block containing Santa Clara Valley at ca. 12 Ma. By 12 Ma, the San Andreas fault (sensu stricto) cut through the ophiolite in the Sierra Azul block, displacing to the northwest a slice of the ophiolite that currently resides north of Point Arena in the Vizcaino block (Fig. 1; Jachens et al., 1998) on the west side of the fault. On the east side of the valley, slip on the East Bay fault system began some time later, based on correlation of the 12–10 Ma Quien Sabe volcanic field east of Hollister with equivalent-aged volcanics on Burdell Mountain near Novato (McLaughlin et al., 1996; Graymer et al., 2002; Ford, 2007). At the time of initiation of East Bay faulting, the present Santa Clara Valley region was ∼100–110 km south of the Quien Sabe volcanic field. Slip on the east side of Santa Clara Valley appears to have begun on the Silver Creek fault (also referred to as the proto–Calaveras fault by Graymer et al., 2005), stepping right 6–8 km onto the proto–Hayward fault. The Silver Creek fault, as defined by Wentworth et al. (2010), is the older right-lateral fault that was later locally obscured by the younger Silver Creek thrust in the Yerba Buena Ridge (also known as Coyote Ridge) area. These faults began to dissect a large tabular flat-lying ophiolite, displacing the Priest Valley ophiolite, mostly concealed, but exposed along its southernmost extent near Parkfield, southeastward from its cross-fault counterpart, the ophiolite on Yerba Buena Ridge (R.C. Jachens, 2013, personal commun.). The area within the stepover region began to subside, forming the Evergreen Basin. Although there are no deep wells within the Evergreen Basin, it is likely that marine sediments were deposited in this basin at this time because Miocene sedimentary rocks of this age throughout the California Coast Ranges were mostly deposited in marine environments. Concurrent development of the Cupertino and Evergreen Basins beginning at ca. 12 Ma led to formation of a basement ridge that now bisects the Santa Clara Valley (Fig. 3).

12–9 Ma

During this time, slip across the right step between the Silver Creek–Calaveras and Hayward faults produced ongoing lengthening and subsidence of the Evergreen Basin (Fig. 4B). We suspect that apparent thinning of the ophiolite section and Franciscan complex rocks beneath the Evergreen Basin resulted because of translation of more southerly rocks into the line of section by right-lateral faults (Fig. 3). Concurrently, the Cupertino Basin continued to subside and fill with marine sediments. Marine sedimentation continued until at least 8 Ma, the age of the sediments at the bottom (260 m depth) of the McGlincy drill hole based on diatoms (MGCY, Fig. 2; Lisa White, San Francisco State University, 2004, written commun.). Tilting of the Miocene section in the Cupertino Basin must have ended by 8 Ma, because the sediments found at the bottom of the McGlincy drill hole project onto the flat-lying section imaged in the seismic-reflection profile of Williams et al. (2004).

9–4 Ma

During this time, Salinian basement and its overlying sedimentary cover were translated northwestward to a position adjacent to the San Francisco Bay block by the San Andreas–Pilarcitos fault, according to the fault reconstructions in Jachens et al. (1998). Slip continued along the Silver Creek and proto–Calaveras-Hayward faults in the eastern part of the Santa Clara Valley, with accompanying subsidence in the Evergreen Basin. Some slip may have also been taken up by the Mount Misery (?) fault, a completely concealed fault along the east margin of the Evergreen Basin (Wentworth et al., 2010) that is considered by Graymer et al. (2015) as probable, but not required. Early compression took place along the western range front, based on a significant angular unconformity at the base of the Pliocene–Quaternary Santa Clara gravels (Graymer et al., 2015). Sometime during this time interval, probably toward the end, however, the San Francisco Bay block and its surrounding terrain emerged above sea level. Uplift of the central part of the San Francisco Bay block and erosion of the intervening bedrock ridge between the Cupertino and Evergreen Basins likely began during this time frame, eventually exposing the serpentinite at Yerba Buena Ridge, which provided serpentinite detritus near the base of the Santa Clara Basin fill in the Guadalupe drill hole (GUAD, Fig. 2; Oze et al., 2003). Uplift also continued in the central Diablo Range, stripping Coast Range ophiolite and any overlying Great Valley and Tertiary section that had not already been thinned or removed by attenuation faulting during or before early Miocene time (Harms et al., 1992), while shallow marine and nonmarine deposition continued around the perimeter of the range and in the Vallecitos syncline region (Fig. 4B).

The emergence of the Santa Clara Valley block at this time may imply that a significant volume of the Miocene section outside of the Cupertino Basin (and possibly Evergreen Basin) was also removed by erosion, forming the unconformity that forms the base of the alluvial Santa Clara Basin. The evidence that this section was originally laid down and subsequently eroded can be found in sections that have been displaced out of Santa Clara Valley by the San Andreas fault system. The offset equivalent of the ophiolite on Yerba Buena Ridge (YBR in Fig. 1) east of the San Andreas fault system, near Parkfield, coincides with a large, prominent aeromagnetic anomaly produced by the ophiolite in Priest Valley (PV in Fig. 1; R.C. Jachens, 2013, personal commun.). This ophiolite is buried where it is truncated by the San Andreas fault, and modeling of its resultant magnetic anomaly places the top of the source at ∼2–3 km (Griscom and Jachens, 1990). Seismic velocity models of this area also indicate a strong velocity gradient at about the same depth above a low-velocity zone interpreted to indicate high pore pressures (Eberhart-Phillips, 1989). A drill hole, the Philips Petroleum Company Varian A1, indicates that the source of this anomaly is buried beneath a 1.5-km-thick section of Miocene and younger sedimentary rocks (John Sims, U.S. Geological Survey, 1987, written commun.). The Monterey (middle Miocene) and Temblor Formations in this drill hole were likely also present in Santa Clara Valley, indicating that more than 1.5 km of section may have been removed by erosion outside the Cupertino and Evergreen Basins. The amplitude of the Yerba Buena aeromagnetic anomaly is smaller than at Priest Valley, suggesting that the serpentinite source is thinner in Santa Clara Valley, and thus a deeper level of erosion.

Another inference for erosion and removal of part of the Neogene sedimentary section comes from southern Santa Clara Valley near Hollister and highlights the lateral variation of deposition within the San Francisco Bay block. A drill hole in Hollister Valley (HV in Fig. 1), west of the Calaveras fault (Texaco Nutting No. 1, API number 069000288; http://owr.conservation.ca.gov/WellSearch/WellSearch.aspx, accessed June 2014), penetrates more than 2.3 km of late Miocene (?) and Pliocene Purisima Formation (California Division of Oil and Gas, 1982; Robbins, 1982; Rogers, 1993). Other deep drill holes west of the Calaveras fault in Hollister Valley (Texaco Recht No. 1, API 06900290; Trico Breen #1, 06900062; Chevron O’Connell B-2, 06900061; Chevron O’Connell B 1, 069000289; http://owr.conservation.ca.gov/WellSearch/WellSearch.aspx, accessed June 2014) encounter as much as 2.2 km of Purisima Formation overlying Mesozoic basement rocks that include serpentinite. Note that all of these wells lie south of or near the projection of the Sargent fault (Figs. 2 and 4B), which has an unknown amount of strike slip during this time frame but is likely to be on the order of several kilometers (which restores ophiolitic and Franciscan complex rocks into reasonable proximity; McLaughlin et al., 1999). Robbins (1982) modeled the gravity and magnetic anomalies in this area as reflecting a substantial section of late Miocene (?) to Pliocene sedimentary rocks overlying basement that must include serpentinite and related ophiolitic rocks. This magnetic basement produces a magnetic anomaly that is truncated at the Calaveras fault along a reach of ∼25 km (R.C. Jachens, 2013, personal commun.) and can be followed northwestward into the Santa Cruz Mountains. Restoring 174 km of displacement on the Calaveras and central San Andreas faults places the Hollister magnetic anomaly against a higher-amplitude magnetic anomaly near Palo Prieto Pass (PPP on Fig. 1). This magnetic anomaly, truncated along a reach of 25 km along the San Andreas fault, reflects a large concealed magnetic body, likely composed at least partly of serpentinite (R.C. Jachens, 2013, personal commun.; Hanna et al., 1972). Because Mesozoic rocks crop out above the Palo Prieto Pass body, this body has remained buried since separation; Hanna et al. (1972) modeled the top of the magnetic body at 4–4.5 km depth. The Hollister Valley magnetic body, on the other hand, probably forms the floor of the mostly Pliocene deposits that fill Hollister Valley (Robbins, 1982), indicating that the serpentinite body was exposed after separation from the Palo Prieto Pass body and subject to erosion before being reburied beneath Hollister Valley. Note also that the amplitude of the magnetic anomaly in the Palo Prieto Pass area is much higher than that in the Hollister area, suggesting uplift, exposure, and erosion of the serpentinite beneath Hollister Valley as it was translated right-laterally by the Calaveras–Silver Creek–Hayward faults (Fig. 4B).

The erosional surface at the base of the Quaternary deposits documented by seismic-reflection profiles and various drill holes in Santa Clara Valley may extend at least as far north as the San Leandro synform (Fig. 2; Wentworth et al., 2010), a structure revealed by seismic-reflection and gravity data (Marlow et al., 1999). The western, east-dipping limb of the synform consists of ∼1–1.5 km of Tertiary(?) marine (?) sedimentary rocks truncated by a horizontal, angular unconformity overlain by ∼300 m of flat-lying strata. The projection of the east-dipping limb to the west lies above the angular unconformity imaged in seismic-reflection data, suggesting that ∼1–1.5 km of section were stripped or eroded away (Marlow et al., 1999). The age of the development of this unconformity is poorly constrained; the development of the surface may have initiated during this time frame or initiated later (as discussed by Wentworth et al., 2010).

4–1 Ma

Deposition of alluvial gravels began on both sides of Santa Clara Valley. At ca. 4 Ma, the margins of the Cupertino and Evergreen Basins began to receive nonmarine sediments, such as the Silver Creek, Santa Clara, and Irvington/Packwood gravels (Page, 1992). During this time interval, the range-front thrust systems along both margins of the valley were active (Figs. 3 and 4B), overriding the Miocene–Pliocene (?) margins of the Cupertino and Evergreen Basins. Fission-track data indicate that rocks in the Santa Cruz Mountains in the area of Loma Prieta (LP on Fig. 2) have been uplifted 3.5–4 km since ca. 4 Ma (Bürgmann et al., 1994). Gravels of the Santa Clara Formation were folded, uplifted, and faulted in the hanging-wall blocks of the Monte Vista and Berrocal thrust faults. Average shortening rates of these faults during the past 5 m.y. are similar to those accommodated by folding, 0.5–0.6 mm/yr (McLaughlin et al., 1999). During this time interval, the San Andreas fault reorganized, abandoning the Pilarcitos strand and straightening its trace (McLaughlin et al., 2007). As a result of this reorganization, a slice of Permanente terrane bounded by the Pilarcitos and San Andreas faults was offset 22 km to the northwest relative to the belt of the Permanente terrane along the southwestern margin of Santa Clara Valley (Jachens and Zoback, 2000). Timing of this slight reorganization in the San Andreas fault is provided by Pliocene–Pleistocene gravels of the Santa Clara Formation that are offset from a distinctive clast source across the San Andreas fault by ∼28–30 km (Cummings, 1968; Dibblee, 1966). The discrepancy between the bedrock offset based on matching magnetic anomalies and offset of the Santa Clara Formation can be resolved if the offset of the Santa Clara gravels did not solely occur on the Peninsula segment of the San Andreas fault, but also partly occurred on the Pilarcitos fault (McLaughlin et al., 2007). This would argue for a period of time during which slip occurred on both the Pilarcitos fault and the Peninsula segment.

The distribution of Santa Clara gravels in and around Santa Clara Valley is puzzling. Santa Clara gravels crop out on the hanging-wall block of the Monte Vista fault, which is at a higher elevation than the valley floor, where at least the older part of the section (>1 Ma) has not been identified in any of the drill holes (Wentworth and Tinsley, 2005; Andersen et al., 2005). This suggests one of two possibilities. (1) The basin block may have been at a higher elevation than that of the hanging-wall block of the Monte Vista fault during the deposition of these gravels; thus, gravels were never deposited on the valley floor. (2) Santa Clara and other gravels may have been deposited out in the valley, but then they were eroded to remove all trace of the Santa Clara gravels. Subsequently, other gravels were deposited across the valley floor. Some support for the second scenario comes from distinctive chert clasts derived from the Claremont Shale (Andersen et al., 2005) and “exotic” siliceous felsite and volcanic porphyry clasts (Vanderhurst et al., 1982) in the oldest Santa Clara gravels exposed in the hanging-wall block of the range-front system near Saratoga. The most likely source of these clasts is to the east and south of the valley; paleocurrent indicators suggest that the older part of the section apparently was deposited to the north across the valley ca. 2 Ma. The mechanism to accomplish this history of deposition, erosion, and deposition of gravels in and around the Santa Clara Valley may have involved two periods of slip on the Monte Vista fault system or other unknown basin-bounding faults. Alternatively, this history could have resulted from changes in base level that were not structurally controlled.

On the east side of Santa Clara Valley, a reorganization of the East Bay fault system led to abandonment of the right step between the proto–Calaveras-Hayward faults, displacement on the Mount Misery (?) fault, and development of a restraining bend between the active Calaveras and Hayward faults. The Silver Creek fault (or central proto–Calaveras fault of Graymer et al., 2005) no longer transferred right-lateral slip to the Calaveras-Hayward faults. Instead, Wentworth et al. (2010) have argued that the pull-apart basin between the Silver Creek and proto–Hayward faults was bisected by the completely concealed Mount Misery fault. Also during this time, the southern part of the Silver Creek fault was modified by reverse and thrust faults (including the Silver Creek thrust of Wentworth et al., 2010) that placed serpentinite over the 2–4 Ma Silver Creek gravels. Strike slip on the central Calaveras fault took over at ca. 3.5 Ma (Page, 1992), or even later at ca. 2 Ma (Wentworth et al., 2010; Graymer et al., 2015). On the eastern side of the Evergreen Basin, the basin margin formed by the southern Hayward fault is also overprinted by a stacked sequence of thrust faults. A flap of Cretaceous Great Valley sequence is thrust over the basin as a result of oblique right slip within the restraining bend between the Calaveras and Hayward faults. This fault reorganization on the east side of the valley postdated the eruption of the basalts of Anderson–Coyote Lake Reservoir (Graymer et al., 2015) dated at 2.5–4 Ma (Nakata et al., 1993).

The reorganization of the fault systems that bound Santa Clara Valley within this time interval may have resulted from a change in plate motion that produced as much as 10 mm/yr plate-normal convergence beginning between 3.9 and 3.4 Ma (Harbert, 1991), timing that overlaps with the onset of uplift and transpression throughout the central and southern Coast Ranges (Page et al., 1998). Later plate reconstructions, however, indicate significantly less plate-normal convergence (DeMets et al., 1994) or plate-normal convergence initiating much earlier at about 8 Ma (Atwater and Stock, 1998). Furthermore, refined estimates of the onset of transpression along the margins of the Santa Clara Valley suggest that transpression may not be synchronous (Graymer et al., 2015), as would be expected if the plate-motion changes are solely responsible for onset of uplift and transpression. Alternatively, the reorganization of these faults may be related to more local fault interactions and strain partitioning.

1–0 Ma

After widespread uplift and erosion across Santa Clara Valley, the area began to subside and receive sediment to form the Santa Clara Basin ca. 1–1.5 Ma, while adjacent uplands continued to rise. Analysis of the sediment from the various deep drill holes in Santa Clara Valley indicates a uniform subsidence rate of ∼0.4 mm/yr during the past 750 k.y. (Wentworth and Tinsley, 2005). Sedimentation rates, on average, apparently were equal to the subsidence rates, as sedimentary environments within Santa Clara Valley have not migrated much in time and space during the past 1–1.5 m.y. Clast compositions throughout the coreholes indicate that the major present-day alluvial fans on the west side of the valley were also prevalent in the past (Andersen et al., 2005; Locke, 2011), and estuarine sediments are not found far from the current shore of San Francisco Bay, despite sea-level highstands during the past 1–1.5 m.y.

Seismic-reflection data in Santa Clara Valley indicate that the strata within this <1–1.5 Ma alluvial basin are generally flat-lying and parallel to the ground surface (Williams et al., 2004). The alluvial strata are only locally deformed, for example, by the Silver Creek fault (Wentworth et al., 2010) and prominently at the eastern side of the Evergreen Basin. In the seismic-reflection profiles, the alluvial section is at most 400 m thick (at least 468 m thick according to a well on the west rim of the Evergreen Basin; Crittenden, 1951) and lies on top of an irregular Mesozoic basement or over a horizontal Miocene (Cupertino) or Pliocene (Evergreen) sedimentary section. In the Cupertino Basin, the Miocene sediment in turn overlies a dipping, presumably Miocene sedimentary section. The extent of subsidence and sedimentation of this young basin may have been widespread, if these basin sediments are the same as those imaged in the seismic-reflection profiles across the San Leandro synform, as suggested by Wentworth et al. (2010). Only in scattered areas, such as Coyote Hills and Coyote Point in the north, and Oak Hill (also known as Communications Hill) and Yerba Buena Ridge in the south, is the relief on Mesozoic bedrock large enough that bedrock crops out above the alluvial deposits and San Francisco Bay mud. It is this alluvial basin that hosts the groundwater for the valley.

DISCUSSION AND CONCLUSIONS

Several models have been proposed for the formation of Santa Clara Valley and can be classified in terms of extensional, compressional, or strike-slip tectonics (Sedlock, 1995). Extensional tectonics predict thick, late Cenozoic basin fill in the present-day San Francisco Bay depression. The tectonic and stratigraphic history summarized here suggests that the extensional model is applicable to the San Francisco Bay block locally during the Miocene, but certainly not during the late Pliocene. Overprinted on this Miocene history is basin formation caused by stepovers within the East Bay strike-slip fault system. Compressional tectonic mechanisms, expressed both in terms of folding and thrust faults, clearly were predominant during the Pliocene and Quaternary. However, no one tectonic model completely explains the history of Santa Clara Valley, which illustrates the dynamic and complex nature of deformation here. It is not clear which model applies to the formation of the Santa Clara Valley alluvial basin, a very widespread depression with only 150–500 m of sediment. Clearly, the formation of the depression within the San Francisco Bay block is not the result of a huge pull-apart basin under San Francisco Bay, given that the margins of the valley are being overriden by reverse faults.

From the data presented here, it appears that Santa Clara Valley is relatively free of young intravalley faults. Seismic profiling (Williams et al., 2004) does not image significant structures projecting up into the uppermost 100 m other than the Hayward fault, Hayward-Calaveras connecting thrusts, and the Silver Creek fault. Williams et al. (2004) did not find evidence of the Cascade and Shannon faults, faults that have been previously postulated to lie basinward of the Monte Vista fault, along their Cupertino–Santa Clara seismic-reflection profile. The Cascade, Shannon (north of Los Gatos), and Santa Clara (also called the Stanford) faults were located on the basis of truncated stream channel sands inferred from interpretations of driller’s logs (California Department of Water Resources, 1975), a finding inconsistent with the analysis of driller’s logs and stratigraphy by Wentworth et al. (2010, their p. 8), although the spatial resolution of their data may not be sufficient to rule out faults with small displacement. The Cascade and Shannon faults, however, coincide with changes in stream sinuosity and stream-channel gradients interpreted to result from zones of localized uplift (Hitchcock and Kelson, 1999). Furthermore, the Cascade fault, as defined, relocated, and renamed as New Cascade fault by Hanson et al. (2004), appears to form a groundwater barrier, except where traversed by coarse-grained channels, within the basin. The location of this barrier is constrained by a pair of wells and is drawn to merge into the range-front faults to the northwest and southeast. Perhaps these faults, if they are faults, are so young or have such low slip rates that they do not have enough displacement (<20 m according to William et al., 2004) to be imaged by seismic-reflection or stratigraphic data. In contrast, Catchings et al. (2006) interpreted extensive faulting and folding along a profile ∼10 km to the southeast of the Williams et al. (2004) profile (Fig. 2). Faults interpreted by Catchings et al. (2006, their fig. 11) do not appear to offset reflectors in the older section, whereas their interpreted folds do not extend into the young Santa Clara Basin section. The faults inferred by Catchings et al. (2006) east of the range-front fault system do not appear to be seismogenic structures because (1) they cannot be extended to the northwest to the longitude of the Williams et al. (2004) profile, (2) many of the reflectors are not consistently offset, and (3) the faults do not appear to extend below 500 m of the ground surface.

The stratigraphic and tectonic history of Santa Clara Valley has implications for seismic hazard and management of groundwater resources. The old basin-bounding faults of the Cupertino and Evergreen Basins may not be slipping at rates as high, or in the same manner, as they did during basin formation, but there is evidence for reactivation and Quaternary movement. The Monte Vista fault zone bounding the Cupertino Basin and the Silver Creek fault bounding the Evergreen Basin may be the sources of the M6.5 October 1865 earthquake and the two 1903 M6 earthquakes, respectively. The locations for these events based on analysis of historical accounts of damage (Bakun, 1999) are tantalizingly close to the Monte Vista fault zone and the Silver Creek fault, respectively. Damage with evidence of contraction was concentrated along the inferred extent of the Monte Vista fault zone (as well as along the Berrocal and Shannon faults) during the 1989 Loma Prieta earthquake (Schmidt et al., 1995, 2014; Langenheim et al., 1997). Geodetic and surveying data suggest that northeast-directed contraction continued after the earthquake along the southwest margin of Santa Clara Valley (Bürgmann et al., 1997; Schmidt et al., 2014). The Silver Creek fault (proto–central Calaveras fault of Graymer et al., 2005) beneath Santa Clara Valley coincides with a sharp deformation gradient in interferometric synthetic aperture radar (InSAR) data (Galloway et al., 2000; Schmidt and Bürgmann, 2003). The deformation results from partitioning of the basin aquifer by the fault, rather than by tectonic slip (Schmidt, 2002; Schmidt and Bürgmann, 2003) and is accentuated by coarser-grained facies present on the west side of the fault (Hanson, 2015). Hydrologic modeling indicates that the fault acts as a groundwater barrier, except where traversed by coarse-grained, permeable stream channels (Hanson et al., 2004; Hanson, 2015). The documented latest movement on the Silver Creek fault is 140 ka, the maximum age of the sediments within the groundwater basin that show evidence of deformation on a seismic-reflection profile (Williams et al., 2004; Wentworth et al., 2010). Holocene displacement is possible, but not conclusively demonstrated, by the arching and topographic steps along Coyote Creek across the fault (Wentworth et al., 2010). Minor fault offset was also suggested by the stream gradient and fluvial terrace analysis of Hitchcock and Brankman (2002), which indicated some broad Holocene deformation above the concealed fault.

Characterization of the older basin architecture also aids in predicting ground motions from future earthquakes. Simulations using the Cupertino and Evergreen Basins to assign seismic velocities beneath Santa Clara Valley predict enhanced shaking along the basin margins (Harmsen et al., 2008). Recent simulations of the 1906 San Francisco and 1989 Loma Prieta earthquakes also show elevated ground motions within Cupertino Basin (Aagaard et al., 2008; Graves and Pitarka, 2010). The deep, V-shaped geometry of Evergreen Basin can also amplify ground motions by as much as a factor of 3 by trapping long-period waves from sources along the long axis of the basin (Hartzell et al., 2010). Although most of the formation of deep sedimentary basins occurred before 4 Ma, the superposition of thrusting on the basin margins and the geometry of these old basins emphasize the ongoing relevance of the tectonic and stratigraphic history beneath Santa Clara Valley in terms of seismic hazard and groundwater resources to the inhabitants and infrastructure that reside within the valley.

This work would not have been possible without funding from the National Cooperative Geologic Mapping and Earthquake Hazards Programs of the U.S. Geological Survey and from the Santa Clara Valley Water District. Reviews by Joseph Clark, Andrei Sarna-Wojcicki, and Richard Sedlock improved the manuscript. We also appreciate the comments by guest editor Randy Hanson.