GSA Bulletin: Late Cenozoic tectonics of the central and ...courses/c341/Owens_Valley/...ABSTRACT...

ABSTRACT

The central and southern Coast Ranges of California coincide with the broad Pa-cific–North American plate boundary. Theranges formed during the transform regime,but show little direct mechanical relation tostrike-slip faulting. After late Miocene defor-mation, two recent generations of range build-ing occurred: (1) folding and thrusting, begin-ning ca. 3.5 Ma and increasing at 0.4 Ma, and(2) subsequent late Quaternary uplift of theranges. The ranges rose synchronously alongthe central California margin and are still ris-ing; their long axes are quasiparallel to theplate boundary and strike-slip faults. The up-per crustal internal and marginal structures ofthe ranges are contractional, dominated byfolds and thrusts resulting from the convergentcomponent of plate motion. Newly constructedtransects using seismic reflection and refrac-tion, plus gravity and magnetic studies, reveallower crustal basement(s) at depths of 10–22km. The upper surface of the basement andMoho show no effect of the folding and thrust-ing observed in the upper crust. We concludethat horizontal shortening is accommodated atdepth by slip on subhorizontal detachments,and by ductile shear and thickening. Theranges are marked by high heat flow; weakrocks of the Franciscan subduction complex;high fluid pressure; bounding high-angle re-verse, strike-slip, or thrust faults; and uplift at arate of 1 mm/yr beginning about 0.40 Ma.Transverse compression manifested in folding

within the Coast Ranges is ascribed in largepart to the well-established change in plate mo-tions at about 3.5 Ma.

INTRODUCTION

The California Coast Ranges province encom-passes a system of elongate mountains and inter-vening valleys collectively extending southeast-ward from the latitude of Cape Mendocino (orbeyond) to the Transverse Ranges. This paperdeals with the portion of the province that liessoutheast of San Francisco, a subprovince (Fig. 1)that we call the central and southern Coast Ranges,encompassing the Diablo Range, East Bay Hills,Santa Cruz Mountains, Gabilan Range, Santa Lu-cia Range, La Panza Range, San Rafael Range,Sierra Madre Range, Caliente Mountain, andTemblor Range (Fig. 1). Exactly the same area hasbeen called the southern Coast Ranges in the past(e.g., Page, 1981; see also Jennings, 1977).

The central and southern Coast Ranges havebeen studied for more than a century, attentionbeing largely devoted to stratigraphy, petrology,structure, and resources potential. In recent years,geophysical investigations have provided funda-mental subsurface information. In this paper weexamine the main Cenozoic tectonic events andthe culminating rise of the ranges in their presentconfiguration, the timing, manner of uplift, pos-sible causes, and mechanics.

A review of the geology of the central andsouthern Coast Ranges and references to manystudies may be found in Page (1981), and conven-ient summary papers are available in Wahrhaftigand Sloan (1989). Pioneering groundwork wasdone by such notables as A. C. Lawson (e.g.,1893, 1914), R. D. Reed (1933), and N. L.

Taliaferro (e.g., 1943). A prodigious amount ofgeologic mapping by T. W. Dibblee, Jr., pre-sented the areal geology in a form that made gen-eral interpretations possible. E. H. Bailey, W. P.Irwin, D. L. Jones, M. C. Blake, and R. J.McLaughlin of the U.S. Geological Survey andW. R. Dickinson are among many who have con-tributed enormously to the present understandingof the Coast Ranges. Representative referencesby these and many other individuals were cited inPage (1981). Additional significant papers arenoted in this article.

Probably the earliest study that specifically fo-cused on the uplift of the present-day ranges wasthat of Christensen (1965), who recognized thatthe Coast Ranges as we know them are youngerthan 3–4 Ma. Subsequent writings on uplift of theranges include Page (1981, p. 415–416); Zandtand Furlong (1982), Montgomery (1993),Anderson (1994), and Burgmann et al. (1994).Many other authors have contributed importantdata, as noted herein. Our review is an attempt tointegrate an earlier paper (Page, 1981) with newand significant geophysical studies that have en-hanced our knowledge of the crust and mantlewithin the Coast Ranges.

GEOMORPHIC FEATURES OFTECTONIC SIGNIFICANCE

The ranges and major intervening structuralvalleys are subparallel with the continental marginand the Pacific–North America plate boundary.The San Andreas fault is parallel with some of theranges, but is slightly oblique to the Coast Rangesprovince as a whole, and crosses it with a prevail-ing trend of about N41°W. The orientation of theranges with respect to the plate boundary pre-

846

OVERVIEWLate Cenozoic tectonics of the central and southern Coast Ranges of California

Benjamin M. Page* Department of Geological and Environmental Sciences, Stanford University,Stanford, California 94305-2115

George A. Thompson† Department of Geophysics, Stanford University, Stanford, California 94305-2215

Robert G. Coleman Department of Geological and Environmental Sciences, Stanford University,Stanford, California 94305-2115

GSA Bulletin; July 1998; v. 110; no. 7; p. 846–876; 18 figures; 5 tables.

*Deceased.†E-mail: [email protected].

cludes an origin by wrench tectonics (i.e., foldingand faulting produced directly by horizontal shearsuch as in transform interaction), in which casefolds and thrusts would be expected to form at anangle to the boundary, in an en echelon pattern.

The individual ranges are mutually similar in

their trend and in their quasilevel crests, whichhave few conspicuous peaks (Figs. 2 and 3). Theyare 120 km to more than 300 km long, 10 to 50km wide, and generally 400 to 1200 m high. Themodest widths, together with geophysical data,argue against some proposed modes of origin,

such as the presence of low-density roots or localcausative thermal conditions associated with indi-vidual mountain ranges. The scarcity of through-going transverse river valleys, coupled with theweak nature of most Coast Ranges rocks, is sug-gestive of (but not proof of) geologically recent,ongoing uplift.

Except for their mutual parallelism and rela-tively level crests, the overall morphologic as-pects of the several ranges differ widely. Some(e.g., the northern Santa Lucia Range, Fig. 4, andnorthern Gabilan Range) show eroded fault-block features including steep scarp-like flanks,faceted spurs, and V-shaped canyon mouths at theapices of alluvial fans. Several ranges, includingthe two just mentioned, have remnants of uplandpediments or vestiges of gently rolling uplands,presumably formerly near sea level, which aresharply incised by headward-growing moderncanyons. Nearly level upland remnants suggesteven uplift. However, the southern part of theGabilan Range is a tilted fault block that hasdeeply eroded, gently sloping upland surfaceremnants that dip gently toward their juncturewith the floor of Salinas Valley. Some ranges (orfoothill ranges, such as the one visible in Fig. 2bordering the Santa Clara Valley south of SanJose, and the East Bay Hills) look like lineararched ductile welts that have marginal faults butdo not have commanding scarps. In contrast,parts of the Diablo Range and parts of the SantaLucia Range have somewhat depressed widecentral portions bordered on both sides by crestalridges. The northeast flank of the Diablo Rangeis locally accompanied by broad, gently sloping,alluviated, plateau-like tracts that have been onlyslightly uplifted along unspectacular marginalfaults. Each of these diverse morphologic aspectsmust tell us something about the tectonics of thesubprovince.

It is important to note that parts of the flanks ofseveral ranges are deeply embayed by erosion andlack indications of marginal faults (e.g., the north-east flank of the Santa Lucia Range betweenGreenfield and Paso Robles, and much of thesouthwest flank of the Gabilan Range). Hence,fault slip is not always necessary for the rise of theranges, although it has locally played a role.

The principal intermontane valleys, such as theSanta Clara, Salinas, and Cuyama Valleys, andthe Carrizo Plain, are relevant to the tectonics ofthe province, as they are clearly of structural ori-gin and are presumably products of the sameevents that caused uplift of the ranges. Theirfloors are aggraded rather than eroded, and theyare bordered by piedmont fans. The Carrizo Plainlacks a fluvial outlet and contains a saline playa.The longitudinal drainage of the Santa Clara Val-ley reverses its direction of flow near the mid-point of the valley, indicating that the valley prob-

CENTRAL AND SOUTHERN COAST RANGES, CALIFORNIA

Geological Society of America Bulletin, July 1998 847

SF

H

SB

EBH

_ 380

_12

20

_12

20

38 0

SN

D

D

GV

SC

MG

KSV

SLSG

_ 36 0

36 0

C

KH

SAF

PR

PO

SO

L T

CP

SM CV

122

0

_12

00

CR

SR

100 km100 miles

N

Figure 1. Physiographic map of the central and southern Coast Ranges, a subprovince (withinthe white polygon) of the California Coast Ranges. Inset shows location within the State of Cali-fornia. The subprovince approximately coincides with the broad transform plate boundary, char-acterized by active dextral transform faults, three of which are labeled (SAF, SG, and H). Abbre-viations: C—Coalinga; CP—Carrizo Plain; CR—Caliente Range; CV—Cuyama Valley;D—Diablo Range; EBH—East Bay Hills; G—Gabilan Range; GV—Great Valley; H—Haywardfault; K—King City fault; KH—Kettleman Hills; L—La Panza Range; M—Monterey; PO—Pa-cific Ocean; PR—Paso Robles; SAF—San Andreas fault; SB—San Francisco Bay–Santa ClaraValley depression; SC—Santa Cruz Mountains; SF—San Francisco; SG—San Gregorio–Hosgrifault zone; SL—Santa Lucia Range; SM—Sierra Madre Mountains; SN—Sierra Nevada; SR—San Rafael Range; SO—San Luis Obispo; SV—Salinas Valley; T—Temblor Range.

ably existed prior to, and independent of, thepresent drainage system. Most likely, the crustaltracts marked by these valleys were forced downor remained at a neutral level when the adjacentmountains rose.

A few streams have an antecedent relationshipto ranges or parts of ranges. For example, the Pa-jaro River cuts across the south end of the SantaCruz Mountains, flowing out onto a coastal allu-vial plain near Watsonville, and Alameda Creekcrosses the East Bay Hills in a steep-sidedcanyon, debouching onto an alluvial fan that de-scends to San Francisco Bay. Because thesestreams flow upon aggraded surfaces before andafter crossing the mountains, they could not havebeen superposed from an uplifted erosion sur-face. The rise of the transected ranges must havebeen gradual enough, and the stream flow vigor-ous enough, to allow the survival of the trans-verse drainage. On the other hand, most parts ofmost ranges have very few through-going trans-verse streams and appear to have risen too rapidlyto allow antecedent drainage to persist. Strikingexamples of rapid uplift include the East BayHills (even though cut by Alameda Creek) andthe frontal foothill ridge of the Diablo Rangesoutheast of San Jose (Fig. 2).

Each of the Coast Ranges is flanked by allu-vial fans except where these have been preventedfrom developing, or have been removed by de-structive processes. Many (most?) of the fansshow abrupt changes in sediments or gradientsthat can be related to abnormal rainfall or, veryoften, episodic uplift of the mountainous sourceareas. In some instances, suggestive features ofthe fans correlate with paired stream terraces up-stream, as observed along much of the east sideof the Diablo Range, supporting the likelihood ofa tectonic origin (Bull, 1964). The latter is con-firmed in cases where paired stream terraces pro-ject “into thin air” above an alluviated piedmontplain; this can be a criterion for uplift of a rangealong a marginal fault.

These geomorphic observations indicate thatsome (most?) ranges have risen, in some cases asblocks with concomitant slip on range-frontfaults, but in other cases perhaps by ductilesqueezing up or variable mechanisms that are dif-ficult to categorize. The morphology of some lin-ear foothills suggests ductile arching, in somecases accompanied by faulting at the margins.

ROCKS COMPOSING AND BORDERINGTHE RANGES

Overall Characteristics

Figure 5 is a simplified geologic map of theCoast Ranges. Many, but not all, of the formationsand assemblages in the Coast Ranges are mechan-

PAGE ET AL.

848 Geological Society of America Bulletin, July 1998

Figure 2. View looking south several ranges, from southwest flank of Diablo Range, showingtheir alignment and even crests. Most distant is the Santa Lucia Range (80 km away); just be-yond fog are the Santa Cruz Mountains (32 km away); this side of fog are very recently upliftedfoothills, an outlier of the Diablo Range (16 km away). The active Calaveras fault is along valleyon near side of foothills. Fog occupies Santa Clara (structural) Valley.

Figure 3. Upland surface remnants and concordant ridges collectively representing paleo-topography of low relief, probably formed near sea level. View is to northeast across the DiabloRange north of Mount Hamilton and east of San Jose. The Franciscan Complex, includingmelanges, underlies most of the area. Bare ridge in middle distance is approximately 10 km long.

ically weak and relatively nonresistant to erosionand mass-wasting. There is an abundance ofpoorly consolidated sandstone and shale, and anabundance of rocks that are rendered weak by per-vasive fracturing. These generalizations apply tothe abundant Franciscan (Mesozoic) rocks as wellas the Neogene and Quaternary sedimentaryrocks. The fact that such materials compose topo-graphically high domains implies uplift rates thatcan successfully compete with erosional surficialprocesses.

Mesozoic Complexes

The underlying rocks of the subprovince in-clude four categories of Mesozoic complexes:(1) the Franciscan Complex, (2) the Salinianblock, (3) the Great Valley sedimentary se-quence, and (4) serpentine and ophiolites, in-cluding the Coast Range ophiolite. These com-plexes partially overlap in age, but they differfundamentally and were formed in differentways. Moreover, most of them are multiple; forexample, the rocks that are generally calledFranciscan include diverse subcomplexes,which have little in common except tectonicassociation.

Franciscan Complex. The complex is gener-ally regarded as an accreted subduction zonecomplex, some of which is no longer in its origi-nal latitudinal location. It contains oceanic man-

tle remnants (serpentine), oceanic crustal compo-nents (gabbro, diabase, basaltic greenstone, radi-olarian chert), and terrigenous sedimentary rocks(graywacke, siltstone, and shale), mainly of LateJurassic and Cretaceous age (Bailey et al., 1964).It also contains blueschist facies equivalents ofthe foregoing, metamorphosed under high-pres-sure (P) low-temperature (T) conditions, presum-ably as a result of Mesozoic subduction. Thecomplex includes melanges consisting ofstrongly sheared argillaceous matrix material inwhich blocks of the aforementioned rock typesare more or less chaotically disposed, and largecoherent units of graywacke and metagraywacke,some of which are described as “broken forma-tions.” The melanges are particularly nonresistantto tectonic and erosional processes. We think thatthe large volume and relatively ductile behaviorof much of the Franciscan were important factorsin the tectonics of the central and southern CoastRanges.

Salinian Block. In contrast to the FranciscanComplex, the Salinian block is an allochthonouscomposite of granitic and metamorphic rocksfrom the axial portion of the western Cordilleranplutonic belt, which it resembles petrologically. Itdoubtless came from a region south (perhaps farsouth) of the Sierra Nevada. Hall (1991) gave arecent interpretation of the source and emplace-ment of the allochthon, with comprehensive ref-erences. The Salinian rocks include abundant

granitic plutons (e.g., Ross, 1978) that are mainlyCretaceous and are coeval, but totally incompati-ble, with parts of the Franciscan. Also includedare wall rocks of the Sur Series—largely sedi-mentary rocks metamorphosed under high tem-peratures. The Salinian rocks are inherentlymuch more resistant mechanically than the Fran-ciscan Complex. Locally they are not severelydeformed, but elsewhere they have been perva-sively sheared or crushed. In some localities, thegranitic rocks are topographically prominent, butelsewhere their influence on the morphology ofthe ranges is only moderate to slight.

Serpentine and Ophiolites. Serpentine bodiesderived from peridotite are distributed throughoutFranciscan melanges. Serpentine is also locallypresent at faults (but not the major strike-slipfaults) in or near the Franciscan Complex. Spec-tacularly, serpentine largely composes diapirs suchas the New Idria mass (Coleman, 1996). The NewIdria body has risen through the Franciscan Com-plex and entrains fragments of the latter. It andsome other lesser masses must have come fromwithin or below the Franciscan and may representa mantle component of oceanic lithosphere, whichwas thrust under the Franciscan or incorporatedwithin it during subduction. Serpentinization andmobilization were probably promoted by thetransfer and pressurization of water during sub-duction and during the later transverse compres-sion in the Coast Ranges (Coleman, 1996; see es-pecially his Fig. 4 for a tectonic interpretation).The better understood Coast Range ophiolite(Hopson et al., 1981) is the oceanic floor uponwhich the Great Valley sequence sediments weredeposited. The exposed remnants of the CoastRange ophiolite, including serpentine, have proba-bly been peeled up along the northeast side of theCoast Ranges, bearing superincumbent Great Val-ley sequence sedimentary rocks, from an in situophiolitic basement beneath the Great Valley(sensu stricto; Griscom, 1982; Jachens et al.,1995). Most Coast Range ophiolite occurrencesonly compose the upper parts of an ophiolite se-quence, commonly pillow basalt, but serpen-tinized ultramafic cumulates and mantle peridotite,as well as other ophiolitic members, are locallypreserved. Wherever the Coast Range ophiolite,other ophiolites, or serpentine bodies appear at thesurface, large tectonic displacements can be in-ferred. Figure 5 shows principal occurrences ofsuch materials. It is manifest from the numerous,scattered exposures of these rocks, derived fromvarious deep sources, that the crust has been pro-foundly deformed to a considerable depth.

Great Valley Sequence. Marine clastic sedi-mentary rocks (Bailey et al., 1964) are quasico-eval with both the Franciscan Complex and theSalinian plutons, having been deposited in a fore-arc basin (Dickinson and Rich, 1972) during the



Figure 4. Steep, linear northeast front of the northern Santa Lucia Range, surmounting largealluvial fans that descend toward the Salinas River. These features suggest fault-block uplift ofthis part of the range. Relief on mountain front is approximately 900 m.


��

��

�

��

�

��

��

0 50 mi

0 50 km

121 ˚

38˚

120 ˚

37˚

119 ˚

36˚

35˚

UNDEFORMED:

DEFORMED:

Late Quaternaryalluvium & terracedeposits.

Plio-Pleist. nonmarine sed's

Neogene rocks, mainly marinesediments; some volcanics

U. Jurassic to Paleogene marineclastic sed's, incl. Great Valley Sequence forearc basin deposits

Granitic (Cret.) and metamorphic rocksof the allochthonous Salinian Block

Franciscan Complex: Melanges and largebodies of graywacke, shale, greenstone, and chert.Local blueschist and serpentine. A subductioncomplex assembled in Late Cret. and Paleogene.

Serpentine and ophiolitic rocks (commonly Jurassic)

Scale

N

122 ˚

A

A'

B

B'

St

MF

C

B

PR

SLO

SF

��

SN

GV

PO

SG-HFZ

SAF

same plate convergence cycle that formed thejust-mentioned complexes. (However, these threecomplexes did not necessarily form at the samelatitude.) The Great Valley sequence rocks areabundant, occurring as tectonically emplacedsheets, locally with remnants of the Coast Rangeophiolite at the base, tectonically overlying oragainst the Franciscan Complex. The interveningfault is the regional Coast Range fault (Bailey et al., 1970) or one of many younger successorfaults. The most striking occurrence of the GreatValley sequence is a thick homocline (Fig. 6) aswide as 15 km that flanks the Diablo Range andnorthern ranges and flattens downward to the eastto blend in the subsurface with little-disturbedstrata of the Great Valley. This homocline is thehanging wall of a tectonic wedge (Wentworth etal., 1984) that played an important role in the con-struction of the eastern ranges, as discussed later.

Tertiary Marine Sedimentary Rocks

Tertiary marine sedimentary rocks are char-acteristic of large areas in the Coast Ranges(Fig. 5). Some formations are shallow-waterclastic shelf deposits; others (especially thoseof Miocene age) are deep-water shelf or slope-basin sediments rich in organic matter. The tec-tonic environment is significant in the evolutionof the California margin (e.g., Blake et al.,1978; Crouch et al., 1984), and the widespreadoccurrence of marine middle and upperMiocene deposits in the coastal mountainsplaces a limit on the time of inception of theCoast Ranges. A convenient condensedoverview of Coast Range Cenozoic stratigra-phy was provided by Lindberg (1984).

Cenozoic Volcanics

Volcanic rocks are locally present in thestratigraphy of the central and southern CoastRanges and are of considerable tectonic interest(e.g., Dickinson and Snyder, 1979). Many aresubmarine extrusives of basaltic compositionindicative of extension. However, most areMiocene in age and thus have little or no directbearing on the Pliocene and Quaternary tectonicevents that concern us.

Pliocene and Pleistocene Sedimentary Rocks:Generally Nonmarine, Commonly Deformed

As emphasized by Christensen (1965), themost significant rocks (or deposits) for under-standing the youthfulness of the present CoastRanges uplift are Pliocene–Pleistocene grav-els, sands, silts, and clays, which are largelynonmarine and are in the proximity of theranges in many localities (Fig. 5). These de-posits are mainly of fluviatile origin, but someare debris flows, some are lacustrine, and someare tongues of shallow-marine or estuarinesediment. It is noteworthy that the basal andnear-basal strata of at least three of the mainlynonmarine formations (Paso Robles, Tulare,and Santa Clara Formations) are shallow-marine or estuarine facies that intertonguewith, or are overlain by, dominant fluviatilesediments. Therefore, the basal beds representa paleo-sea level. The predominantly nonma-rine character of the bulk of these Pliocene–Pleistocene formations denotes the “final”emergence of coastal California. The coarserfacies contain clasts that indicate source areas,rock assemblages, and erosional levels thatwere providing sedimentary debris. In addi-tion, some coarse facies provide current direc-tions and other paleogeographic clues. In a fewinstances, datable tephra layers are interbed-ded with clastic strata.

It is important to note that Quaternary depositsolder than ca. 450 ka are commonly, albeit lo-cally, folded and faulted. The Pliocene–Pleis-tocene sedimentary rocks contain the history ofwidespread uplifts, provide evidence of subse-quent recent deformation, and constrain the tim-ing of the more recent rise of the ranges in theirpresent form.

Among the Pliocene–Pleistocene nonmarineformations that fit the above generalizations arethe following: Santa Clara Formation, IrvingtonGravels, Livermore Gravels, Packwood Gravels,San Benito Gravels, Quatal Formation, MoralesFormation, Tulare Formation, and Paso RoblesFormation. Some salient facts regarding these for-mations are briefly summarized in Table 1. Thereis probably considerable overlap in terminologyand considerable merging and continuity among

sedimentary rocks that have locally been given dif-ferent names. Several of the Coast Ranges of cen-tral California are partially flanked by such depos-its (Fig. 5); in many cases, discordantly. Thedeposits are also generally flanked by younger, un-deformed (or scarcely deformed) fan deposits. TheSanta Clara Formation was interpreted by Vander-hurst et al. (1982), the Paso Robles Formation byGalehouse (1967), the San Benito Gravels by Grif-fin (1967), and the Tulare Formation by Woodringet al. (1940). Other authors are cited in Table 1.

The ages of the deformed nonmarine Pliocene–Pleistocene deposits are critically important for thedating of range uplifts. The available age determi-nations, which generally are between 4 Ma and0.47 Ma, may be grouped in the following cate-gories: (1) ages obtained by correlation of tephralayers (in sedimentary rocks of numerous locali-ties) with isotopically dated ash elsewhere; (2) ra-diometric dates from lavas enclosed in the sedi-mentary rocks, obtained in only one area; and (3) paleontologic ages. A summary of some se-lected age data is given in Table 1. We rely mostheavily on ages in category 2, provided by the im-portant work of Sarna-Wojcicki and associates(e.g., Sarna-Wojcicki, 1976; Sarna-Wojcicki et al.,1985, 1991). Correlations of tephra with the Rock-land ash are particularly important, this ash havingbeen dated at about 0.40–0.47 Ma (Meyer et al.,1991; Sarna-Wojcicki, 1996, personal commun.).For expedience, we have arbitrarily picked 0.45Ma for the age of the Rockland ash and the upperpart of the Santa Clara Formation, which containsthe ash, but we have no scientific reason for pre-ferring 0.45 Ma over any other age between 0.40and 0.47 Ma.

INTERNAL STRUCTURES—LARGELYCONTRACTIONAL

The contractional tectonics have also been em-phasized by others, including Aydin and Page(1984), Namson and Davis (1988), and Jones et al. (1994).

Folds Within the Ranges

All of the mountains except the GabilanRange have numerous pronounced folds. Many



Figure 5. Simplified geologic map of the central and southern Coast Ranges. Note localization of the ranges along the broad transform plateboundary, characterized by large strike-slip faults. Note also the indications of contraction transverse to the plate boundary. Numerous fragmentsof serpentine and ophiolite (both in purple) imply widespread deep deformation bringing pieces of mantle and oceanic crust to the surface. Therecent uplift of the ranges is implied by the relations of deformed Pliocene–Pleistocene formations (brown). Locations of crustal transects A–A′and B–B′ are shown. Abbreviations: B—Bakersfield; C—Coalinga; F—Fresno; GV—Great Valley; M—Monterey; PO—Pacific Ocean; PR—Paso Robles; SAF—San Andreas fault; SF—San Francisco; SG–HFZ—San Gregorio–Hosgri fault zone; SLO—San Luis Obispo; SN—SierraNevada; St—Stockton.

trend parallel (or nearly so) with the San An-dreas fault and the plate boundary, showing agenetic relationship to inferred transverse com-pression. However, some folds are oblique withrespect to the San Andreas fault (Fig. 7), possi-bly reflecting the prevalent horizontal dextralshear of the transform regime. A few obliquefolds such as the Vallecitos syncline in the Dia-blo Range (Fig. 5) have been evolving sincethey were first established as elongate warps,perhaps as early as Eocene time, presumably bydextral transpression during oblique plate con-vergence. A number of folds began to form inlate (or even early) Miocene time. Regardless oftrend and the time of inception, the Cenozoicfolds commonly involve Neogene rocks, and, inmany cases, Pliocene marine strata and/orPliocene–Pleistocene nonmarine formations.

Although the Coast Ranges folds have dif-ferent histories, apparently nearly all grew (ortightened) in late Quaternary time. Remark-ably, most of these young folds are incorpo-rated within the ranges as internal structuresrather than defining the ranges. In other words,an anticline in Neogene rocks usually does notcoincide with an antiformal mountain range.This means that the ranges are even youngerthan most of the Pliocene–Pleistocene folds.However, a few outlying anticlines and syn-clines that might be considered to be satellitesof the Diablo and Temblor Ranges are soyoung that they have direct topographic ex-pression (e.g., Coalinga anticline and Kettle-man Hills, both of which are parallel with theSan Andreas fault). Thus, there has been a suc-cession of folding events, some of which pre-ceded the uplifts of the present ranges andsome of which have been contemporaneouswith the uplifts.

Internal Faults

Several kinds of faults are within the ranges.Like the folds, most are young, as shown by off-sets of upper Miocene or Pliocene–Pleistoceneformations, but few have direct topographic ex-pression. There are low-angle detachments ofunknown origin, and there are low-angle thrusts(e.g., Fig. 8) and high-angle reverse faults.

Some of the thrusts and reverse faults trendwest-northwest or east-west and can be ex-plained by wrench tectonics, like the obliquefolds. However, many internal thrusts and re-verse faults are subparallel with the San An-dreas fault and are best explained as products oftransverse compression.

Small, young reverse faults that do not seem toplay any part in uplift are on the east flank of the

PAGE ET AL.


Figure 6. Homocline of Great Valley sequence (clastic Mesozoic forearc basin sedimentaryrocks) dipping easterly along northeast flank of Diablo Range. Approximately 3 km across thetilted section.

TABLE 1. PLIOCENE–PLEISTOCENE FORMATIONS INVOLVED IN COAST RANGES OROGENY

Name Character Age Associatedrange

Santa Clara Formation Mainly fluviatile clastic rocks. Marine tongues Probably ca. 3.6–0.45 Ma. Lower beds are Pliocene, based on Santa Cruz Mountains,near base. A few lacustrine and rare thin mollusks near Stanford (Addicott, 1969) and K-Ar age (3.6 Ma Diablo Rangeash beds. of interbedded basalt near Gilroy (Sarna-Wojcicki, 1976).

Upper part has 0.47–0.40 Ma tephra on the basis ofcorrelation with Rockland ash (Sarna-Wojciki, 1976).

Irvington Gravels (probably = Fluviatile sand, gravel, silt, and clay. Probably ca. 1.9–0.45 Ma (Lindberg, 1984); derived age of Diablo Rangeupper Livermore Gravels) notable vertebrate fauna, basis of Irvington Stage.

Tulare Formation Mainly fluviatile clastic rocks, some lake Probably ca. 4.0 Ma to late Pleistocene. Pliocene mollusks near Temblor Rangebeds, estuarine at base. base (Addicot and Galehouse,1973). Ash in lower part is ca. Diablo Range

4.0 Ma (Sarna-Wojcicki et al., 1991). Upper part evidently notwell dated.

Paso Robles Formation Mainly fluviatile clastic rocks; marine tongues Probably ca. 4.0 Ma to late Pleistocene. Pliocene mollusks near Gabilan Range,at base. base (Addicot and Galehouse,1973). Ash in lower part is ca. Santa Lucia Range

4.0 Ma (Sarna-Wojcicki et al., 1991). Upper part evidently notwell dated.

Livermore Gravels Mainly fluviatile, sand, gravel, silt, clay, some Pliocene–Pleistocene, based on vertebrates and freshwater Diablo Rangelake beds. mollusks (Griffin, 1967).

San Benito Gravels Mainly fluviatile, sand, gravel, silt, clay, some Pliocene–Pleistocene, based on vertebrates and freshwater Diablo Rangelake beds. mollusks (Griffin, 1967).

Diablo Range between San Luis Reservoir andLittle Panoche Creek (Lettis, 1982). These strikeparallel with the range and appear to be bedding-plane faults in the Upper Cretaceous PanocheFormation, which here generally dips northeast,away from the crest of the range. The sense ofslip on these faults is reverse, the northeast sidemoving up relative to the opposite side. Thus, theslip is opposite to that which would contribute tothe rise of the axial part of the Diablo Range.Holocene movement has not been demonstrated,but the slip (5 to 100 m) is so recent that Pleis-tocene pediment surfaces are offset, remnants ofscarps survive, and drainage seems to be im-peded in some places (Lettis, 1982). Althoughthese rather puzzling faults do not play a criticalpart in the range uplift, perhaps they are “sympa-thetic” to a northeast-dipping thrust fault under-lying the hanging wall of a Wentworth-typewedge (see section on Crustal Structure—TheMiddle Crust).

SUMMARY OF CENOZOIC TECTONICEVENTS

Figure 8 is a structural section across the SantaCruz Mountains, and shows some ages (denotedby numerals) of certain unconformities, folds, andfaults. It is immediately apparent that more than asingle tectonic episode has transpired, and that thegreater part of the deformation occurred in Ceno-

zoic time, and much of it in Quaternary time.Table 2 summarizes the main events recorded

in the cross section of Figure 8. The late Miocenefolding and faulting are so pronounced that oneschool of thought maintains that the Coast Rangesare largely Miocene, and extremists have sug-gested that their beginning was in Eocene time.We think that the Pliocene and Quaternarydisturbances were genetically related to the rise ofthe ranges with their present outlines, although atleast one Miocene episode (described below) inthe area of the Diablo Range may have had somelong-term influence. Unconformities are numer-ous in the Tertiary records of many of the CoastRanges, as shown in various stratigraphiccolumns (e.g., Lindberg, 1984), and most of theseapparently reflect tectonic warping. They do notcorrelate spatially with particular ranges, and inmost cases their varying chronology from place toplace does not support the notion of far-reachingmountain-building episodes. We briefly summa-rize the tectonic and stratigraphic-sedimentologicrecord, beginning with the latest Cretaceous.

Franciscan Debris in Upper Cretaceous andEocene Sedimentary Rocks

The Maastrichtian (uppermost Cretaceous)Moreno Formation is the oldest sedimentaryrock unit known to contain Franciscan detritus(S. A. Graham, 1996, personal commun.). It is

an upper part of the marine Great Valley se-quence and was deposited during subduction,probably receiving Franciscan debris from anelevated outer arc ridge, which was not directlyrelated to the eventual Coast Ranges. The nextappearance of Franciscan debris in the strati-graphic record is in Eocene sedimentary rocks.The shallow-marine Domengine Formation(ca. 50 Ma) of the central Diablo Range typi-cally overlies unconformably older Paleogenesedimentary rocks and contains Franciscan de-tritus in minor quantities. A marked distur-bance must have elevated a source area, eitherfar beyond (west of) the accumulated Meso-zoic Great Valley forearc basin deposits, orpossibly entailing the stripping of these depos-its to expose underlying Franciscan rocks. Wethink that this event took place during subduc-tion, that it represents an unusual uplift of theouter arc ridge, and that it is not relevant to theorigin of the Coast Ranges.

Late Oligocene Disturbances, 29–25 Ma

Oligocene rocks in the central Coast Ranges in-clude some nonmarine sediments, and Oligocenedeposits are missing in many areas where otherTertiary epochs are well represented. These cir-cumstances are probably related to the changefrom plate convergence to transform interaction atabout 29–25 Ma in areas that were then at the



Cuyama Valley

Temblor Range

Santa MariaValley

RiconadaFault

Santa Lucia Range

San G

reorio - Hosgri F

ault Zone

La Panza Range

10 km

10 mi

Q

Q

Q

Q

Q

Q

KgKg

Kg

Q

SLR

SL

PAC

IFIC O

CEA

N

GREAT VALLEY

35˚120¡

B-B'

Carrizo Plane

San Andreas Fault

Caliente Range

Sierra MadreRange

Figure 7. Map of fold axes and major strike-slip faults in part of the southern Coast Ranges at the south end of Figure 5. The strike-slip faultsapproximate the trend of the plate boundary, and most of the fold axes are nearly parallel with, or at an acute angle to the boundary. Abbrevia-tions: B–B′—line of cross section (Fig. 5); Kg—Cretaceous granitic rocks of the Salinian Block; Q—Quaternary alluvium.

present latitude of southern California. Subse-quent prolonged marine sedimentation occurred,so the 29–25 Ma disturbances did not lead di-rectly to the building of the Coast Ranges.

Early and Middle Miocene Events

Oligocene and older rocks are commonly over-lapped unconformably by shallow-marine “Va-queros-type” sandstones and conglomerates. In ar-eas of the present continental shelf, early Miocenesedimentary basins formed during extensional epi-sodes. Most of these basins and their contents werecompressed later in the same epoch, but their de-formation was not as acute as the later folding andthrusting throughout the site of the Coast Ranges.

Lower and middle Miocene shallow-watermarine sandstones that unconformably overlieolder rocks in west-central California com-monly contain Franciscan detritus as a notice-able minor constituent. This indicates uplift andexposure of the Franciscan marine accretionarycomplex after the cessation of subduction. Ma-rine deposition was widespread thereafter, andthe distribution and character of marine sedi-mentary rocks precludes the existence of exten-sive mountains prior to late Miocene time. Nev-ertheless, local crustal deformation is indicated,especially where marine Miocene rocks rest di-rectly on Mesozoic complexes. This is observedin some areas underlain by Salinian plutonicrocks (e.g., near Carmel and Monterey) as wellas in areas of the Franciscan Complex (e.g., inparts of the San Luis Obispo–Atascadero regionof the Santa Lucia Range).

A possible Coast Range precursor uplift oc-curred within the present confines of the DiabloRange south of Pacheco Pass. There, a thin rem-nant of middle Miocene marine sandstone un-conformably overlies the Franciscan core of therange and underlies a patch of Quien Sabe Vol-canics at lat 36°50′N, long 121°20′W (Fig. 5).The lowest Quien Sabe flows are dated at 11.6Ma (as discussed by McLaughlin et al., 1996).Thus, that particular area was elevated above sealevel in the Miocene, was resubmerged beneaththe sea before 11.6 Ma, and was subsequentlyuplifted again within a central part of the modernDiablo Range. The initial uplift was restricted tothe site of the present range, as shown by the factthat the Franciscan basement rocks were not gen-erally exposed (hence not markedly uplifted)elsewhere. Their cover of Great Valley rocks,which we think had been tectonically emplacedearlier, was not eroded away and is still preservedin extensive remnants around the periphery of themountains (Fig. 5). The spatial coincidence be-tween the Miocene and Quaternary uplifts sug-gests that the former was localized by a deep-seated condition that persisted and influenced thelater event. However, the trend of the Mioceneemergent terrain was somewhat oblique to theaxis of the present mountains, as the uplift appar-ently did not affect the concurrent marine condi-tions in the adjoining Vallecitos syncline area tothe south (Fig. 5). Moreover, there was not a con-tinuous uplift that simply persisted after the earlyMiocene event, as marine conditions returned inareas around, within, and across the site of thepresent mountains. Various parts of the Diablo

Range incorporate marine sedimentary rocks thatwere deposited in late Miocene time and were se-verely folded and faulted prior to the birth of thepresent range.

Late Miocene Tectonism, 11–7 Ma

Marked folding and some faulting occurredoffshore from the present coast late in Miocenetime, and more severe folding and faulting oc-curred in some areas now on land in the presentCoast Ranges. An example may be seen near thesouthwest end of the area shown in Figure 8,where the mildly folded Santa Margarita Forma-tion (uppermost Miocene) unconformably over-lies strongly folded Oligocene strata. In areasnortheast of the San Andreas fault, late Miocenefolding may have been an effect of the passage ofthe Mendocino Triple Junction, because it, wherecurrently situated near Cape Mendocino, is ac-companied by active deformation. The lateMiocene disturbance(s) generally did not coin-cide spatially with individual modern CoastRanges, and where data are available, the foldingwas followed by resumed marine deposition.

Major Tectonism Commencing ca. 3.5 Ma

Figure 5 shows the areal distribution of pre-served Pliocene–Pleistocene nonmarine sedimen-tary rocks such as the Santa Clara, Tulare, andPaso Robles Formations. These rocks record thevarious uplifts that collectively raised western Cal-ifornia above sea level, where it remains today, andthat exposed Cenozoic and Mesozoic marine

PAGE ET AL.


��

��

��

��

��

��

0 10 km0 10 mi

0

5

KJf

?

TKg

Kg

A

��

San

Gre

gorio

f.z.

Santa CruzMudst. 8.7-6.9 Ma

Santa MargaritaFm. 11-9 Ma Bu

tano f

. < 3

Ma

Purisima Fm.6.9-3.0 Ma

Mindego Basalt20.2-23.7 Ma

Lambert Shale

Butano Ss. 52-44 Ma

Folds are Quaternary,involving Purisima Fm.

Salinian basement, incl. Cret. plutons

Detachment

Pila

rcito

s f.

San

And

reas

f.

Santa Clara Fm.ca. 4.0-0.45 Ma

Thrusts younger than 0.45 Ma

Page MillBasalt 14.4 Ma

Santa Clara Fm.

Whiskey HillFm. 52-40 Ma

Franciscan subductioncomplex; Jurassic & Cret.

rocks assembled & accretedin Late Cret.

29-15 Ma

Te

Te

Vaqueros Fm.

Te

San Andreasfault zone;

315 km dextralslip since 23.5 Ma;

20 km since 0.45 Ma

Te

Ladera Ss.15-10 Ma

Major sed'n changeca. 29 Ma

Figure 8. Cross section of Santa Cruz Mountains, and dated formations and structures. This is modified southwest part of line A–A′ (Figs. 5 and10). The section shows dominance of contractional features and youthfulness of deformation. Folds and faults are young (largely Quaternary),but were deeply truncated by erosion prior to uplift of present ranges, commencing ca. 400 ka. Abbreviations: KJf—Franciscan Complex; Kg—Cretaceous granitic rocks of Salinian block; Te—Eocene sedimentary rocks—mainly turbiditic.

formations to erosion. In most Coast Range areaswhere Pliocene–Pleistocene formations are pre-served, these sedimentary rocks are locally foldedand faulted. Thus, there was probably a continuumof tectonic activity, beginning with uplift and ero-sion that fed the Pliocene–Pleistocene nonmarinedeposits, and evolving to a stage of pronouncedfolding and thrusting involving these same depos-its. In many cases, the underlying formations aremore severely deformed, having undergoneMiocene or older deformation as well as this re-cent episode, but in other cases the Pliocene–Pleis-tocene formations are folded almost as much asOligocene and Miocene strata beneath them. Col-lectively, the Pliocene–Pleistocene tectonic eventsin this region were probably the most severe of anysince the Mesozoic. As discussed later herein, wethink that this orogenic sequence was the result ofthe change in plate motions at about 3.5 Ma.

The deformation cited above did not directlyproduce the existing ranges, but evidently “pre-pared the ground.” The individual ranges of thepresent day do not coincide spatially with any par-ticular deformational features of the Pliocene–Pleistocene folding and thrusting, but the entirecentral Coast Ranges subprovince is restricted tothe realm of that deformation. The northeasternboundary of the Coast Ranges tends to coincidewith the northeastern limit of Pliocene–Pleis-tocene folding, thrusting, and wedging. To thesouthwest, Pliocene–Pleistocene folds are ob-served beyond the ranges far offshore on the con-tinental shelf, but the intensity of deformation isrelatively mild. Landward, the shoreline and theedge of the mountain belt are located roughlywhere the Pliocene–Pleistocene deformation be-comes intense.

Although most of west-central California didnot rise above sea level permanently until thePliocene–Pleistocene orogeny, the east SanFrancisco Bay area had emerged earlier. The siteof the present bay was a topographic high in lateMiocene time, shedding sediment eastward(Graham et al., 1984). This points up the youth-fulness of the western California landscape.

Rise of the Present Ranges Since ca. 0.4 Ma

The appreciable separation in time and spacebetween the Pliocene–Pleistocene folding andthe rise of the present Coast Ranges in late Pleis-tocene and Holocene time is shown by the factthat the Pliocene–Pleistocene structures were ex-tensively and deeply eroded prior to the uplift ofgently rolling topographic surfaces, which trun-cate the folds. Moreover, the margins of someranges transect folded Pliocene–Pleistoceneformations, some of which contain beds as youngas 400 000–470 000 yr (Sarna-Wojcicki, 1976).

GEOPHYSICAL DATA AND CRUSTALSTRUCTURE

Two Crustal Transects

In order to understand the character of theCoast Ranges province at depth as well as at thesurface, we have compiled two compositecrustal transects, one passing near San Fran-cisco (Fig. 9, inset), and one passing near SanLuis Obispo (Fig. 10, inset). The compilationsdraw heavily from the work of many geologistsand geophysicists, as noted (incompletely) onthe two figures; we are deeply indebted to all.The two transects show a few significant simi-larities, despite the differences in rocks andstructures. A crustal-scale perspective requiresconsideration of geophysical data, which arediscussed below.

Effects of Transform Movements on Geologyof the Transects. In viewing Figures 9 and 10, thereader must be aware that the southwestern partsof the crust have moved (and continue to move)relatively away from the observer, and the north-eastern parts have moved relatively toward the ob-server. Thus, each transect is an instantaneousframe in a moving picture, and the various rocksand structures did not necessarily originate in theirpresent side-by-side positions. The transform mo-tion began about 29 Ma (Atwater, 1970), is nowproceeding at a rate of about 4.8 cm/yr (DeMets

et al., 1990), and has probably accrued at least 920km of strike separation. This relative motion be-tween the Pacific and North American plates isdistributed in a broad zone, within which the SanAndreas fault accommodates about 3.5 cm/yr(Prescott et al., 1981). Thus any rocks or structuresof Oligocene age or older, if located near thesouthwest end of a transect, may have been dis-placed hundreds of kilometers with respect torocks of comparable age near the northeast end ofthe same transect. However, the younger the rocksand structures, the smaller the total relative dis-placement. Inasmuch as the central and southernCoast Ranges are very young, the transform dis-placements of relevant age must be relatively mod-erate, although still substantial. Pliocene rocks onthe two sides of the San Andreas fault are probablydisplaced less than 175 km, and Pleistoceneformations are displaced less than 55 km.

An acute problem is posed by the ongoingtransform motions vis-a-vis the presence of theseemingly little-disturbed lower crust. This prob-lem is considered in a later section.

Interpretation of Gravity Profiles

Bouguer and isostatic residual anomalies areplotted on the two crustal transects, Figures 9 and10 (foldouts). Isostatic residual anomalies are ob-tained by subtracting from Bouguer gravity thecalculated effect of idealized isostatic compensa-tion for regional elevation (Roberts et al., 1990;Roberts and Jachens, 1993). Because the calcu-lated effect is regional, the result is a high-passfilter that removes the long-wavelength, crustalscale gravity anomalies (wavelength greater thanroughly 50–100 km) while preserving anomalieshaving a shallower source and smaller lateral ex-tent, such as those caused by sedimentary basinsand lithologic variations in the basement.

In section A–A′ (Fig. 9; see Fig. 5 for loca-tion), the Bouguer anomaly is strongly positive(about 25 mGal) at the southwestern end, be-cause the continental crust is unusually thin nearthe coast. On the other hand, the isostatic residual



TABLE 2. MAIN CENOZOIC TECTONIC EVENTS

Time (Ma)(approx.) Event Comments/questions

0.40 Delineation, and beginning of uplift, of present-day Why was there a delay of about 50 000 yr between severecentral and southern Coast Ranges contractional deformation and uplift of ranges?

0.45–0.40 Widespread, deep erosion, drastically modifying folds Enormous scale of erosion facilitated by predominance of nonresistant rocks and pluvial, glacial-stage climate

0.45 Approximate beginning of Coast Ranges folding and Concurrent erosionthrusting, affecting Pliocene–Pleistoceneformations

3.5 “Permanent” uplift of California above sea level, Ascribed to change in plate motions (e.g., Harbert and Cox,initiating Coast Ranges mountain building 1989)

11–7 Folding and thrusting in parts of present Coast Ranges Not directly related to present Coast Ranges, except by province and in areas now offshore thickening of the crust

29–25 Major change in character of sedimentation; extensive Change from subduction to transform regime in domains unconformities formerly farther south

anomaly is compensated for this effect. East ofthe San Andreas fault, the two anomalies trackclose to one another (within 10 mGal), and bothare strongly negative in the northeastern half ofthe transect. As explained more fully in the fol-lowing discussion of section B–B′, the negativeanomalies are closely associated with thick, low-density Cenozoic sedimentary rocks. This corre-lation is clearly demonstrated in the vicinity ofthe Calaveras fault and in the Great Valley; thereis an intervening gravity high over the moredense Cretaceous sedimentary rocks in the north-ern Diablo Range. On a more detailed scale,Cenozoic deposits preserved in faulted synclinesgenerally show associated negative anomalies.The serpentine bodies, which produce strongmagnetic anomalies, have little or no gravity sig-nature at this scale.

In section B–B′ (Fig. 10) the Bouguer andisostatic residual anomalies are virtually coin-cident and near zero about 15 km landwardfrom the coastline. West of this area of coinci-dence, the Bouguer anomaly becomes positive(about 10 mGal at the coast) because of the iso-static effect of the thin continental crust, whilethe isostatic residual anomaly remains nearzero. Northeastward from the coastline, theBouguer anomaly gradually becomes morenegative toward the interior of the continent,and at the northeastern end of the transect, itdiffers from the isostatic residual anomaly byabout –25 mGal. Both anomalies clearly showa correlation with structural and/or lithologicfeatures in the transect and on the geologic map(Fig. 5), especially with low-density sedimen-tary rocks. The isostatic residual anomalies areused here to interpret these features.

Low-density sedimentary rocks, mainly ofCenozoic age but including Cretaceous rocks inthe Great Valley, have a strong negative effect onthe gravity anomaly. Because these rocks havebeen intensively drilled for hydrocarbons, theirthickness, density, and gravity effect can beroughly estimated. The long-dashed line in sec-tion B–B′ (labeled sediment-adjusted gravity)shows the result of removing the effect of theselow-density sedimentary rocks; the gravity curvebecomes almost featureless. The adjusted gravitycurve on this section shows surprisingly littlecontrast between Franciscan rocks (Mesozoicsubduction complex) and Salinian rocks (plu-tonic arc terrane), even where these are juxta-posed along the San Andreas fault. Thus, the twocomplexes do not differ substantially in meandensity. At a smaller scale, some low-amplitudeanomalies can be correlated with local structureand Cenozoic sedimentary rocks preserved insynclines, but the serpentine bodies, which pro-duce the strongest magnetic anomalies, have nogravity signature at the scale of the cross section.

Magnetic Anomalies

As shown on both crustal transects (Figs. 9 and10), the magnetic field varies by several hundrednanoteslas (nt) and is highest in the Great Valley(well beyond the Coast Ranges), where a well-known magnetic high anomaly trends subparallelwith the axis of the valley.

Magnetic Rocks. Generally, rock magnetiza-tion depends upon magnetite content, grain size,and other characteristics of the magnetite, and themagnetization history (e.g., strength and polarityof Earth’s field at the time of magnetization). Onthe regional scale of interest here, remanent mag-netization (often dominant in quickly cooledlavas) can probably be neglected in favor of in-duced magnetization (parallel to the Earth’sfield), which is important in deep-seated igneousand metamorphic rocks. Ophiolitic rocks of theCoast Ranges include (1) peridotite, which is vir-tually nonmagnetic where unserpentinized(Thompson and Robinson, 1975), because theiron is accommodated in olivine; (2) serpentine,usually intensely magnetic because of abundantsecondary magnetite; (3) altered basalt or “green-stone,” which is usually weakly magnetic be-cause of alteration; and (4) gabbro, diabase, andmafic amphibolite, which, if unaltered, tend to bestrongly magnetic. Partial to complete serpen-tinization of peridotite is common and progres-sively increases the magnetization and decreasesthe density, so that the combined analysis ofgravity and magnetic anomalies is meaningful.

Certain Coast Ranges sandstones containingabundant andesite grains (from Sierra Nevada vol-canics) are an additional unusual source of mag-netic anomalies (R. C. Jachens, 1996, personalcommun.). The upper Miocene Neroly Formationis an example (see following). Doell (1956) foundthat the strong magnetism is remanent but was ac-quired subsequent to folding and is approximatelyparallel with the Earth’s field. He suggested thatthe magnetism resides in a bluish iron-rich mont-morillonitic coating on the sand grains, but this in-terpretation awaits further study.

Interpretation of Magnetic Profile in Transect of Figure 9. Beginning at about kilome-ter 20 from the southwest part of the transect, ananomaly of 50–75 nT west of the San Andreasfault is associated with the Mindego basalt of earlyMiocene age. These rocks include flows and shal-low intrusions of fresh basalt. Northeast of the SanAndreas fault a weak multipeaked anomaly ofabout 25 nT may be attributed to Franciscangreenstones, which are abundant at the surface.Farther northeast, a strong anomaly of amplitudeabout 150 nT near the edge of San Francisco Bay,and a similar anomaly on the northeast side of thebay, are most likely caused by serpentine in shearzones or melanges in the Franciscan Complex.

Serpentine, probably related to the northeasternanomaly, is exposed at one end of Coyote Hills. Atthe Hayward fault zone, a sharply peaked anomalyof amplitude about 200 nT is ascribed to serpen-tine bodies like those that are extensively exposedfarther northwest along the fault in associationwith slivers of various lithologic units of the CoastRange ophiolite.

East of the Calaveras fault zone and on strikewith the Mount Diablo diapiric structure to thenorthwest is a broad magnetic high with super-imposed peaks. The striking peaks are caused byvolcanic sandstone of the Neroly Formation.Note that the gravity low associated with thebroad magnetic high is caused by low-densityCenozoic sedimentary rocks more than 5 kmthick. An exposure of Coast Range ophiolite inpart of the diapiric core of Mount Diablo (13–18km northwest of our transect) suggests that thehigh magnetic and gravity anomalies there stemfrom a substantial subsurface body of ophiolitethat includes both magnetic serpentine and densemafic and/or ultramafic rock, perhaps obductedinto the body of the range, or perhaps lifted fromthe Great Valley ophiolitic basement by an east-ward-advancing wedge of Franciscan material.This latter interpretation is adopted (with somemisgivings) in Figure 9, in which we extend theRiggs Canyon fault (striking toward the MountDiablo diapir) downward into the subsurface asthe hanging wall of a Franciscan wedge.

From about kilometer 95, the magnetic fielddeclines eastward toward the Great Valley, butbegins to rise near the valley edge as part of themajor Great Valley magnetic and gravity highassociated with an ophiolitic basement. Unpub-lished maps of magnetic potential and high-pass-filtered magnetic potential, provided byR. C. Jachens, indicate a general westwardtransgression of the Great Valley magneticanomaly into the Coast Ranges province, as in-terpreted by Jachens et al. (1995). Large massesof ophiolite and/or displaced ophiolitic serpen-tine are inferred to underlie the Diablo Range,on the basis of magnetic and gravity anomalies.

Interpretation of Magnetic Profile in Transect of Figure 10. Beginning in the south-west, most of the prominent anomalies in thetransect can be associated with serpentine bodies,either exposed or concealed. Note that corre-sponding positive gravity anomalies are gener-ally lacking, as expected for serpentine. The large(300 nt) sharp-peaked anomaly in the Santa Lu-cia Range is associated with the Cuesta Ridgeophiolite (Page, 1972), a fragment of the CoastRange ophiolite. The broader anomaly of similaramplitude near the western edge of the Salinianblock has no obvious source. It has no coincidinggravity anomaly and may be caused by an unusu-ally magnetic facies of quartz diorite or by meta-

PAGE ET AL.


morphic wall rocks. In the Diablo Range north-east of the San Andreas fault, a magnetic peak ison strike with the Table Mountain serpentine tothe north. A smaller high at 132 km lies over up-turned Cenozoic sedimentary strata and is in-ferred to be caused by “blue” volcanic sandstones(see Magnetic Rocks section). The huge GreatValley anomaly is attributed to ophiolitic GreatValley basement (Jachens et al., 1995), as previ-ously discussed.

Anomalous Heat Flow and Elevated CrustalTemperatures

Heat flow in the central and southern CoastRanges is anomalously high, averaging about 83mW/m2 (Lachenbruch and Sass, 1973), comparedto the stable United States continental interior(about 42–63 mW/m2; Lachenbruch and Sass,1977). The high heat flow drops markedly fromthe Coast Ranges province to the adjacent GreatValley. On our crustal transects (Figs. 9 and 10),the values in the Coast Ranges vary from 63 to 100mW/m2, whereas, in the Great Valley, the valuesare 40 to 64 mW/m2. Thus, there is likely a rela-tionship between Coast Ranges tectonics and theelevated heat flow. However, we are uncertainwhich came first, the high heat flow or the intensedeformation, which may have been one of thecauses of the anomaly. In interpreting the heatflow, one needs to bear in mind that tectonic ex-tension tends to increase observed heat flow bynecking and convergence of isotherms, whereastectonic contraction (typical of the Coast Ranges)has the opposite effect. Erosion tends to increase,and deposition to decrease, the observed heat flow.Advection by magma or other fluids, and frictionalheat in deforming rocks, can disrupt simple con-ductive heat flow. Moreover, thermal changes inthe lower crust or mantle take millions of years tobe felt in the near surface where measurements aremade. Nevertheless, thermal data supply impor-tant constraints for tectonic history.

As pointed out by A. Lachenbruch and C. Williams (1994, personal commun.), in thecentral and southern Coast Ranges, heat flowdoes not decrease southward as might be ex-pected if the cause of the anomalous values wererelated to the northward passage of the Mendo-cino Triple Junction, that is, if it were caused bythe upwelling of hot asthenosphere into a north-ward-advancing slab window such as that pro-posed by Dickinson and Snyder (1979). The highheat flow appears to extend offshore, but the marine measurements are somewhat suspect (A. Lachenbruch, 1995, personal commun.). Ifan anomalous flux exists offshore, there may be aspatial coincidence between it and the lateMiocene folds and reverse faults that are ob-served on the continental shelf. These folds and

faults, although of moderate intensity, are themost pronounced Cenozoic structures recognizedoffshore. Late Miocene folds and faults are notedin parts of the Coast Ranges, but there they areovershadowed by the effects of Pliocene–Pleis-tocene deformation.

Significantly, the San Andreas strike-slip faultin west-central California produces no observedlocal effects, consistent with low friction and lit-tle heat generation on this fault (Lachenbruchand Sass, 1973, 1980; J. H. Sass, 1995, personalcommun.).

The high heat flow predicts high crustal tem-peratures. Assuming steady-state flow, Sass et al.(1995, personal commun.) calculated 700 °C ormore at the base of the crust and about 400 °C at15 km, the approximate base of the seismogeniczone (where ductile behavior of rocks becomesmore pronounced as a function of depth and tem-perature) in the Parkfield area in the central CoastRanges. Such temperatures require metamor-phism of particular rocks at particular depths, andthis must be taken into account in interpretingvariations of seismic velocities at increasingdepths. Williams et al. (1994) analyzed thecauses of the Coast Ranges high heat flow andconcluded that the source of the heat was mostlikely in the deep crust and/or uppermost mantleand that there is probably no simple cause of theobserved anomaly, but rather, a combination ofcontributing causes such as asthenospheric up-welling, magmatism, and mechanical heatingduring deformation.

In the larger geologic perspective, Mesozoicand early Cenozoic subduction of an old oceanicplate is thought to have “refrigerated” the GreatValley and Sierra Nevada mantle and to have pro-duced the low reduced heat flow (heat flow cor-rected for upper crust heat production) still ob-served there (Blackwell, 1971; Roy et al., 1972).The northwestward advance of the transform sys-tem, replacing subduction, would have spared thesite of the central and southern Coast Rangesfrom this type of “refrigeration” progressively;i.e., after about 25 Ma in the south and after about7 Ma at the site of San Francisco.

At the latitude of southern California, subduc-tion of a platelet (the Monterey plate) of young,initially relatively warm oceanic crust occurred inOligocene and Miocene time (Lonsdale, 1991).This subduction ceased at about 19 Ma, but appar-ently the platelet underlying the continental mar-gin, but now attached to the Pacific plate, rodenorthwestward with the latter as dictated by thePacific–North American transform plate interac-tion (Nicholson et al., 1994; Bohannon and Parsons, 1995). Part of the Coast Ranges systemevolved in the overlying crust that rode piggybackon this relatively young platelet. The platelet mayhave served as a (waning) heat source or as a heat

sink, depending on how deep it subsided. The re-mainder of the central Coast Ranges, although be-yond the limits of the Monterey plate, is also prob-ably underlain by oceanic crust (e.g., Fig. 9). Theorigin and age of this presumed oceanic materialare unknown. Part (or all) may be a leftover slab ofFarallon oceanic crust, in which case it would betoo old (Mesozoic and/or Paleogene) to be a likelysource of heat. It may be part of an unidentifiedcounterpart of the warm Monterey plate. There isa possibility that the edge of the continent wasthrust over the oceanic plate(s) of the Pacific basinduring the Neogene-Quaternary transverse com-pression at the site of the Coast Ranges after truesubduction had ceased (Page and Brocher, 1994).If this happened, the oceanic crust beneath westernCalifornia could be of almost any age, and wouldnot necessarily be a source of heat.

Because the excess heat source seems to havebeen deep in the crust or in the upper mantle, andbecause it doesn’t seem to have required a slabwindow and isn’t restricted to the area of theMonterey plate, perhaps one likely source is shearalong subhorizontal surfaces in the deep crust, forexample, shear along the upper surface of the sup-posed oceanic crustal slab beneath the west part ofthe Coast Ranges province, and the upper surfaceof the Great Valley basement beneath the east part.Lachenbruch and Sass (1973) reasoned that be-cause the San Andreas fault lacks the local heatflow anomaly that would be predicted for a high-friction fault, the resistance to plate motion maynot come from strike-slip friction, but instead maystem from a broad shear zone below the seismo-genic layer. Such a broadly generated thermalanomaly would be consistent with the heat flow ofthe Coast Ranges. Slip along the subhorizontalupper surface(s) of the seismic basement is onelikely source. By seismic basement, we mean thesupposed oceanic crustal slab(s) beneath the westpart of the Coast Ranges province, and the GreatValley basement under the east part. Anothersource might be the intense late Miocene folding.This deformation apparently has about the sameareal extent, offshore and onshore, as the anom-alous heat flow, whereas intense Pliocene–Qua-ternary deformation is largely confined to the on-shore, coinciding with the Coast Ranges province.

A puzzling aspect of the heat flow–tectonicproblem is the paradox that the central CoastRanges, characterized by high heat flow, are alsocharacterized by large volumes of the FranciscanComplex, some of which includes blueschist fa-cies metamorphic rocks. The blueschist type ofmetamorphism requires high pressure, 3 to 8 kbar(300–800 Mp), equivalent to depths of 10–25km, and relatively low temperature, 150–300 °C(Ernst, 1965); therefore it is believed to be a prod-uct of subduction. In the case of Franciscanblueschists, the metamorphism (hence subduc-



tion) is of various ages, younging up to 90 Ma orpossibly 70 Ma. The mineral assemblages,formed at depths of 10–25 km, were somehowprotected from heating for 30 m.y. or more, andto receive this protection, must have been broughtup to shallow levels quickly. They remained shel-tered in the upper crust notwithstanding the ac-tive deformation during Cenozoic time, and musthave avoided the influences at depth that pro-duced today’s high heat flow.

Information from Seismic Reflection and Refraction

Knowledge of the Coast Ranges crust in thesubsurface has been provided largely by seismicreflection and refraction imaging. In our crustaltransects (Figs. 9 and 10), most of the featuresshown in the middle and lower crust, and theMoho, are derived from seismic traverses by oth-ers. Readers may obtain more details fromsources cited in the transects. Particularly usefulpapers include those by Walter and Mooney(1982), Fuis and Mooney (1990), Howie et al.(1993), and Brocher et al. (1994).

Seismic refraction (e.g., Walter and Mooney,1982) typically shows that rocks and facies seenat the surface may not be present below depths of8–15 km, depending on the location. This is usu-ally based on P-wave velocities, which commonlyincrease markedly at certain subhorizontal bound-aries or zones. The question arises, What is the na-ture of these subhorizontal features? Those thatare reflective and which separate units withmarkedly different velocities are presumed to bepetrologic and perhaps tectonic; depending on thecircumstances, some reflective zones could be de-tachments or zones of intense ductile shear such

as those envisioned at the upper surfaces of “seis-mic basements” (higher velocity crystalline crust;see crustal transects, Figs. 9 and 10). Less-defini-tive velocity transitions may mark differences infacies or different degrees of metamorphismwithin a single rock assemblage such as the Fran-ciscan Complex, or may denote the contact be-tween two very different assemblages having onlyslightly different elastic properties. In such caseswe have made interpretive judgments based onlikely lithologic variability within rock complexes(e.g., Salinian magmatic arc rocks or the Francis-can Complex) and consideration of known re-gional tectonic relations and other factors. Seis-mic reflection imaging has been a powerful toolfor recognizing the presence and character of sub-surface structural features in the Coast Rangesand adjacent provinces (e.g., Brocher et al., 1994),and combined reflection and refraction experi-ments have been very fruitful.

Earthquake Sources and Focal Mechanisms

Maps showing the distribution of earthquakeepicenters (e.g., Goter et al., 1994; U.S. Geologi-cal Survey, 1990) indicate some of the sites of on-going crustal activity. Seismicity is dominated byparts of the San Andreas fault (see articles in Wallace, 1990) and other large dextral strike-slipfaults of the transform system. These are promi-nently marked by linear concentrations of epicen-ters; in addition, apparently random, sparsely scat-tered epicenters of small-magnitude earthquakesappear within the individual Coast Ranges. Thisdenotes broadly distributed strain involving occa-sional slip on variously oriented small- to medium-sized faults, and is clearly a symptom of persistentstress and probably ongoing uplift.

Focal mechanisms in the central Coast Rangessubprovince predominantly indicate dextralstrike-slip on northwest-trending near-verticalfaults of the transform system. However, thrustmechanisms are occasionally evident, notably atthe site of the main shock of the 1983 M 6.7Coalinga earthquake (Rymer and Ellsworth,1990). The causal fault, which is near the CoastRanges–Great Valley boundary, strikes N53°Wand dips 23° southwest at the hypocenter, whichwas at a depth of 10 km. The focal mechanismshows that contractional strain transverse to theplate boundary occurs today as in the recent geo-logic past. This interpretation is strengthened bythe thrust mechanism of the 1985 Kettleman Hillsearthquake (Ekstrom et al., 1992), which was cen-tered a few kilometers northeast of the margin ofthe Coast Ranges (see Fig. 10). The M 7.1 LomaPrieta earthquake of 1989 originated on a San An-dreas fault segment, which (atypically) dips 70°southwest and which, at the time of the earth-quake, showed a reverse component as well asdextral strike slip (Plafker and Galloway, 1989).Active thrust faults along the northeast flank ofthe Santa Cruz Mountains west of San Jose strikemore or less parallel with the plate boundary anddip beneath the range (McLaughlin et al., 1991).Along the same flank of the Santa Cruz Moun-tains, on the San Francisco Peninsula, small earth-quakes show various focal mechanisms, in someof which the interpreted fault planes strike paral-lel with the San Andreas and dip southwest to-ward the latter (Kovach and Beroza, 1993; Olsonand Zoback, 1995; Kovach and Page, 1995).These and other seismological observations indi-cate present-day continuation of transverse con-tractional strain accompanying the concurrentstrike-slip activity.

PAGE ET AL.


TABLE 3. RATES OF UPLIFT

Type of data Range Uplift rate Reference Remarks(mm/yr)

Stratigraphic/structural relations Santa Cruz Mountains ca. 1.4 This paper Based on relations and age (ca.3–0.45 Ma) of Santa Clara Formation

Stratigraphic/structural relations Diablo Range 1.4–2 This paper Based on relations and age (ca. 2.2–0.5 Ma)

Differential uplift of erosion Santa Cruz Mountains 0.8–0.36 This paper Uplift alongside new, dated segment ofsurface on either side of 1.25 San Andreas faultMa fault

Uplifted marine terraces Santa Cruz Mountains near 0.16–0.26 Bradley and Griggs (1976) Uplifted flight of upper Pleistocenebase of range terraces near base of mountains

Uplifted marine terraces Santa Cruz Mountains near 0.17–0.41 Lajoie et al. (1991) Approximately same territory; improvedbase of range dates

Uplifted marine terraces Santa Cruz Mountains near 0.13–0.35 Valensise and Ward (1991) Used Santa Cruz territory ca. 1.25 kabase of range

Uplifted marine terraces Santa Lucia Range, northwest 0.16 McKittrick (1988) Mean for past 200 k.y.base

Uplifted marine terraces San Luis–Pismo block 0.11–0.22 Pacific Gas and Electric Territory (ca. 83 ka) uplifted between Company (1988) reverse faults

Theoretical model Santa Cruz Mountains near ca. 1.4–1.8 Anderson(1994) Realistic input, sophisticated modelLoma Prieta

Fission tracks Santa Cruz Mountains near ca. 0.8 Burgmann et al. (1992) Approximate mean, past 4.6 m.y.Loma Prieta

Releveling East Bay Hills and Diablo ca. 1.4–1.8 Gilmore(1993) Unrefined but suggestive dataRange

PARTICULAR FEATURES OF THE CRUST

Upper Crust

Cenozoic Folds and Thrusts Viewed as aWhole. Intense deformation at shallow depths iswidespread except where strong, rigid Salinianbasement rocks have resisted regional stresses, asin the central part of the area shown in Figure 10.(The Salinian rocks have not survived as well inthe area shown by the Fig. 9 transect, as dis-cussed in another section). Layered formations,mostly sedimentary, are extensive and are com-monly folded and cut by faults. The folding isgenerally not directly reflected in the morphol-ogy of the ranges; for example, anticlines aregenerally not topographic highs. This is com-monplace in mountain belts all over the world,but it is surprising here because the folds are soyoung. Although folds are not directly expressedin the topography, the Coast Ranges province as awhole is strikingly coincident with the belt of in-tense Quaternary folding (Figs. 9 and 10). To thewest of the Coast Ranges in the offshore region,the numerous anticlines and synclines are com-paratively mild; the transition from mild defor-mation to intense deformation seems to mark(imprecisely) the location of the shoreline alongmost of the coast. There are, however, outliers ofstrong deformation both at sea and in the marginof the Great Valley. As emphasized throughoutthis article, many of the folded and faulted rocksare Neogene.

Amount of Upper Crustal Shortening. Wehave measured shortening in the upper 10 km ofcrust implicit in the crustal transects of Figures 9and 10 and in a section across the Santa Cruz

Mountains (Fig. 8), and we show the results inTable 3. The table includes some estimates ofother authors for comparison. We conclude thatshortening across the entire span of the centraland southern Coast Ranges is between 15% and40% of the original width of a hypothetical hori-zontal stratum that extended from one side of thesubprovince to the other. Our preferred estimateis 20%–40%. We think that the entire lithospherehas undergone comparable amounts of shorten-ing, albeit by different mechanisms at differentdepths; if this is so, the transverse strain denotedby folds and thrusts in the upper crust is an indi-cator of the amount of shortening of the entirelithosphere. Note the “pseudo-subduction” oflower crust that is postulated in Figures 9 and 10in order to accommodate shortening equivalent tothat observed in the upper crust. It is of interest tosee what rates of deformation would be requiredin order to achieve various postulated amounts ofshortening within permissible time spans. Asshown in Table 4, most of the calculated rates areon the order of millimeters, rather than centime-ters, per year. This is compatible with calculatedrates of uplift of the present-day ranges.

With regard to folding and horizontal short-ening in the Coast Ranges, we have concludedthat (1) the uplifts that elevated the westernfringe of California above sea level, where it hasremained to the present, commenced at about3.5 Ma and were followed by folding and thrust-ing, but (2) the rise of the individual ranges ascurrently delineated was delayed until about0.40 m.y. after most of the deformation had oc-curred. The transverse shortening that is relatedto the latter episode is accordingly only a smallfraction of the total amount that can be surmisedfrom most cross sections showing folded and

thrust-faulted rocks. In areas where Plioceneformations have been folded and thrust faulted,and especially where deformed beds have beendated as 450 000 yr or younger, we can makefairly reliable estimates of the amount of Qua-ternary shortening (e.g., Fig. 8),but clearly mostof this, notwithstanding its recency, occurredbefore the present ranges were uplifted.

Relation Between Upper Crustal Structureand Areal Extent of Central and SouthernCoast Ranges Subprovince. Figures 9 and 10show an unmistakable spatial relation betweenthe Coast Ranges and the broad belt of intenseNeogene and Quaternary folding. Although indi-vidual folds within the ranges have virtually nodirect effect on the topography, mountain rangesare not present beyond the domain of intensefolding. To the east, Cenozoic strata of the mainpart of the Great Valley are not folded or arefolded only slightly, except for recent outliers ofthe Coast Ranges such as Kettleman Hills (Figs.5 and 10). These outliers are especially numerousin the southwest part of the Great Valley, wherethey are so young that some are directly ex-pressed by topographic highs and lows of minorrelief. They are not yet integral parts of the CoastRanges. To the west of the province, offshore,there are many Neogene, and some Quaternary,folds that have been mapped by seismic reflec-tion. However, when shown in cross sectionwithout vertical exaggeration, most of thesestructures are strikingly mild compared with theircounterparts in the Coast Ranges (e.g., Fig. 10).Incidentally, we suggest that the location of theshoreline, which imperfectly coincides with theboundary of the Coast Ranges province, wassomehow determined by the western limit of in-tense folding. Another feature that approximately



TABLE 4. TRANSVERSE SHORTENING

Range or locality Transverse Reference Remarksshortening

Santa Cruz Mountains approximate full ~21 km = 38% of original width, This paper Collective folding and thrusting sincewidth; Neogene and Quaternary 56 km middle Miocenestructures

Santa Cruz Mountains between San ~12 km = 32% of original This paper Collective folding and thrusting since Gregorio and San Andreas faults; distance, about 37 km middle MioceneNeogene and Quaternary structures

Santa Cruz Mountains, folds in ~0.28 km = 4.5% of original This paper Late Pliocene–Quaternary component remnants of Purisma Fm. (Pliocene) length, 6.22 km of deformation only

Santa Cruz Mountains, vicnity of ~2.3 km = 19% of original This paper Late Pliocene–Quaternary component Butano thrust distance, 12.1 km of deformation only

Santa Cruz Mountains, between San ~7.7 km = 43% of original This paper Quaternary component onlyAndreas fault and Stanford University distance, 17.7 km

Santa Cruz Mountains, Sierra Azul ~4 km = 100% in belt of thrust McLaughlin et al., (1991) Local shortening by thrusting sincearea southwest of San Jose; faults Miocene, mostly since about 3 Maimbricated

Diablo Range, entire width north of ~6 km = 14% of pre-Tulare (pre- Namson and Davis (1988) Pliocene–Quaternary deformation,Coalinga 2.2 Ma) width based on stylized and balanced cross

sectionCentral Coast Ranges, entire width of ~33 km = 23% Namson and Davis (1988) Quaternary component could be less

subprovincePyramid Hills and Kettleman south 3.7 km = 17% measured at five Bloch et al. (1993) Meticulous analysis not wholly within

dome, vicinity of east side of Diablo different horizons, 70–14 Ma the Coast Ranges sensu strictoRange in age; 2.1 km or 9% at

horizon 2.5 Ma in age

coincides with the southwest boundary of thesubprovince is the San Gregorio–Hosgri faultzone (Hall, 1975; Silver and Normark, 1978).This zone has undergone large amounts of dex-tral strike slip, probably as much as 150 km in itsnorthern reaches (Clark et al., 1984), and most ofits principal faults are presumably subvertical.We do not know the reason for its near coinci-dence with the southwest boundary of the centraland southern Coast Ranges. The northeast mar-gin of the subprovince is marked by the north-east-dipping homocline of the Great Valley se-quence (Figs. 9 and 10), which owes its genesisto the progressive easterly insertion of a subsur-face wedge of mainly Franciscan rocks (see thefollowing section). Clearly the homocline and thewedge are relevant to the origin of the easternCoast Ranges.

Middle Crust

We consider the middle crust in provincialterms that apply to the central and southern CoastRanges subprovince. It cannot be defined interms of composition, which is largely unknown.We regard it as the crustal domain from a depthof about 8–9 km down to the upper boundary of(likely mafic) material characterized by P-wavevelocities of 6.6 to 7.1 km/s. On the basis of thesecriteria, the middle crust is thickest (10–15 km)along the axial region of the subprovince. It gen-erally transmits P-waves at velocities from 5.6 to6.2 km/s, but this range is locally slightly ex-ceeded. The lower velocities overlap those of theupper crust, where velocities are commonly 3.5to 5.8 km/s (or even higher in Salinian rocks).

The composition and structure of the middlecrust throughout most of the central and southernCoast Ranges are not known with certainty. Thegenerally higher P-wave velocities, as comparedwith the upper crust, suggest any of the followingexplanations: the closure of pores and cracks; thepresence of somewhat different facies of thesame rocks that are seen at the surface; the transi-tion of certain minerals to denser polymorphs;the presence of rock assemblages that are funda-mentally different from those at the surface; thepresence of low-angle contacts, perhaps grada-tional; the presence of low-angle tectonic con-tacts; and so forth.

The middle crust beneath most of the rangesprobably consists of Franciscan rocks. This ideais partly based on the near certainty that the Fran-ciscan is an accreted subduction complex of largelateral and vertical extent. Parts of the complexhave undergone conditions at great depths andhigh pressure, as indicated by blueschist faciesmineral assemblages. Characteristic variations inpetrology and degree of metamorphism can read-ily account for differences in seismic velocities.

Where the upper crust is predominantly Francis-can rocks, the middle crust with higher Vp mayalso be Franciscan, perhaps with more abundantgreenstone or high-P metamorphic rocks. (Fran-ciscan rocks may be absent and other assem-blages may be present.)

At the surface, the core of the Diablo Rangeconsists of Franciscan rocks. These are exten-sively exposed, but their downward limit is notknown with certainty. Although subhorizontal el-ements are rarely seen at the surface, puzzlingdiscontinuous low-angle seismic reflectors areimaged locally at depth, and seismic refractionshows one or more subhorizontal velocityboundaries that apparently occur within the Fran-ciscan Complex (Walter and Mooney, 1982). Arather continuous P-wave velocity boundary(about 5.9 vs. 6.7 km/s) occurs in the DiabloRange at a depth of about 16 km, probably mark-ing the contact between Franciscan rocks and un-derlying mafic basement (Walter and Mooney,1982). This boundary is adopted in our Figure 10.

Of paramount importance is the presence ofwedge(s) of upper or midcrustal Franciscan rocksinserted laterally beneath Great Valley sedimen-tary rocks along the northeast flank of the CoastRanges (e.g., Wentworth et al., 1984). Suchwedges of middle crustal material are evidence ofhorizontal compressional strain at least as fardown as the top of the mafic lower crust, and theiremplacement must have required uplift of theoverlying material.

Our ignorance concerning midcrustal materialbeneath known Salinian magmatic arc rocks (in-cluding granitic types) is particularly frustrating.Vp values are high enough to accommodate thehypothesis that the Salinian block is a “flake” thathas overridden unrelated, unseen assemblage(s);on the other hand, the observed velocities couldbe explained by the presence of Saliniananisotropic gneisses and/or schists with high hor-izontal velocities. The Gabilan Range is one areawhere seismic evidence favors an interpreted thin(about 8–9 km) Salinian granitic upper crustlacking reflectors, below which a reflective unitextends to a depth of at least 13 km (see Hale andThompson, 1982). We have adopted an inferredboundary between granite and underlying schistor gneiss in our crustal transect, Figure 10. Justbeneath the granite there may be schist compara-ble to the Pelona Schist of southern California(see Ross, 1976). Between this and the top of themafic lower crust at about 21 km, the nature ofthe crust is even less certain. In any case, foliationalone probably does not account for the reflectiv-ity, which Beaudoin (1994) suggested is tectonicbanding resulting from ductile deformation dur-ing large-scale transport of the Salinian block.

It is important to note in our crustal transects(Figs. 9 and 10) that the combined upper and

middle crust are appreciably thicker in the centraland southern Coast Ranges than in adjoining do-mains to the east and west, suggesting a relation-ship between the thickening and the existence ofthe subprovince.

Lower Crust

The lower crust is taken to be a deep layer hav-ing P-wave velocities generally ranging from 6.6to 7.1 km/s. Beneath the western half of the cen-tral and southern Coast Ranges subprovince, thislower crust apparently consists of oceanic maficrocks, which can be tracked with a fair degree ofconfidence by means of seismic reflection and re-fraction from the crust of the Pacific basin inland.The relevant research has been concentrated in alarge offshore-onshore region around San LuisObispo and a comparable large offshore/onshoreregion near San Francisco. The results in the SanLuis Obispo region were summarized by Howieet al. (1993) and were tested and refined by Lafond and Levander (1995); the results near SanFrancisco were presented by Brocher et al.(1994) and refined by Holbrook et al. (1996). Inboth regions a slab-like (presumably mafic) layerdescends at a low angle beneath the edge of thecontinent and is nearly level under the medianaxis of the central and southern Coast Rangessubprovince (Figs. 9 and 10).

The current relative plate motion, involvingdextral horizontal shear, dominates the regionaltectonics with respect to total accrued displace-ments and present-day strain rates. It must in-volve the entire lithosphere, and the presence ofan apparently unbroken subhorizontal lowercrust seems incompatible with the large-scalestrike-slip faulting observed at the surface. Do thelarge strike-slip faults, particularly the San An-dreas fault, cut and offset the lower crust in someundetected manner? Is the slip on the near-verti-cal San Andreas fault parlayed into slip on thesubhorizontal upper surface of the mafic lowercrustal slab (e.g., Brocher et al., 1994; Jones et al., 1994)? This question is still not answeredwith certainty, but teleseismic shear wave split-ting indicates that the upper mantle is highlyanisotropic in a broad ba

GSA Bulletin: Late Cenozoic tectonics of the central and ...courses/c341/Owens_Valley/...ABSTRACT...

Documents

Transcript of GSA Bulletin: Late Cenozoic tectonics of the central and ...courses/c341/Owens_Valley/...ABSTRACT...