Quaternary tectonism in the coastal belt of the Los Angeles urban corridor is diverse. In this paper we report the results of studies of multibeam bathymetry and a network of seismic reflection profiles that have been aimed at deciphering the diverse tectonism and at evaluating the relevance of published explanations of the region's tectonic history. Rapid uplift, subsidence in basins, folds and thrusts, extensional faulting, and strike-slip faulting have all been active at one place or another throughout the Quaternary Period. The tectonic strain is reflected in the modern physiography at all scales. Los Angeles (LA) Basin has filled from a deep submarine basin to its present condition with sediment impounded behind a large sill formed behind uplifts near the present shoreline. Newport trough to the south-southeast of LA Basin also accumulated a large volume of sediment, but remained at midbathyal depths throughout the Period. There is little or no evidence of Quaternary extensional tectonism in either basin although as much as 6 km of subsidence, which mainly occurred by sagging, has been recorded in places since the middle Miocene. The uplifts include folded and thrust faulted terranes in the Palos Verdes Hills and the shelves of Santa Monica and San Pedro Bays. The uplifted areas have been shortened in a southwest-northeast direction by 10% or slightly more, and some folds are reflected in the bathymetry. Two large adjacent midbathyal basins, Santa Monica and San Pedro, show strong evidence of subsidence and slight west-northwest extension (10%) during the same time folding was taking place in the uplifts. The tectonic boundaries between uplifts and basins are folded, normal faulted, reverse-faulted, and strike-slip faulted depending on location. The rapid Quaternary uplift and subsidence, along with the filling of LA Basin, have produced a reversal in the regional physiography. In the early Pliocene, LA Basin was a submarine deep, Palos Verdes and the shelves comprised a northeast basin slope, and the present offshore basins and Catalina Island formed an emergent or shallowly submerged shelf. Since extensional, compressional, and lateral strains are all locally in evidence, simple notions that this part of southern California underwent a change from Miocene transtension to Quaternary transpression fail to explain our observations.
 The modern continental margin in southern California is a broad zone of right shear associated with numerous faults related to the San Andreas Fault system, at least in the broadest sense. However, there is a large magnitude of differential shear within the zone and a high variability in the amount and direction of transport of internally deformed, upper crustal terranes between the faults. This has resulted in local complexities of seismicity, structures, and vertical tectonics that can be hard to understand. The complexities found within the Los Angeles (LA) Basin area and in the region just offshore of it (Figure 1) are typical in this regard. Earthquakes in this area, for example, have a wide range of solutions and many of them have occurred in places where they were least expected beforehand. Many areas where one might expect seismicity seem to be curiously devoid of any, particularly in the offshore region (Figure 2). One of the great challenges in any complex and tectonically active area, especially around a large urban area, is to reach a good enough understanding of the geology to at least predict where earthquakes, surface ruptures, and other geological hazards might be most expected. LA is a highly urbanized area where it is hard to study deeply buried structures, so this goal has not yet been reached, although there has been good progress synthesizing oil industry subsurface data [e.g., Wright, 1991; Yeats and Beall, 1991].
 It is recognized that blind thrust faults are some of the most active structures in the region [e.g., Shaw and Shearer, 1999; Shaw and Suppe, 1996] and many segments of well-mapped strike-slip faults, however youthful in appearance, are aseismic. The most recent thrust earthquakes originated beneath large, mostly buried, anticlines (e.g., the ML 5.9 Whittier Narrows in 1987, the MW 6.7 Northridge in 1994). These occurred unexpectedly on concealed faults in places that previously had exhibited little or no seismicity [Hauksson et al., 1995; Hauksson and Stein, 1989], making it hard to predict where the next earthquake might occur. There have been recent efforts to explain the seismicity, the local uplifts, and most of the structural complexity of the LA region in terms of compressional tectonism rather than strike-slip. The whole area has been modeled as a fold and thrust belt [Namson and Davis, 1991; Shaw and Suppe, 1996]. Folds and thrusts are linked in the deep subsurface, beneath any well control, by a regionally extensive decollement in most of these models. Thus, blind-thrust earthquakes might originate beneath any part of the LA region if these concepts are correct and a compressional earthquake at almost any location can be conveniently explained. The possibility of an extensive thrust-decollement greatly increases the hazard assessment represented by blind-thrust earthquakes over almost all other cases. Thus it is important to evaluate the validity of these models.
 This paper attempts to document the style of Pliocene to Recent tectonism in an area that we refer to as the coastal belt of the LA urban corridor (Figure 1). This area encompasses two large youthful uplifts on the Palos Verdes Peninsula [Ward and Valensise, 1994] and in the San Joaquin Hills [Grant et al., 1999] (Figure 1). Parts of Los Angeles Basin, that are now subaerial but were mostly covered by midbathyal submarine fans [Redin, 1991] during the early Pliocene to middle Miocene, are also included. Most of the area lies offshore where we were able to study the subsurface structure using marine seismic techniques. Two important blind thrust earthquakes, the 1979 (ML 5.2) and the 1989 (ML 5.0) Malibu earthquakes (Figure 2) [Hauksson and Saldivar, 1989], were recorded deep beneath Santa Monica Bay in the offshore part of the coastal belt. The offshore part of the belt also includes large rapidly filling basins. The coastal belt spans three major structural and physiographic provinces (Figure 1), the western Transverse Ranges in the northwest, the Peninsular Ranges in the southeast, and the southern California Continental Borderland in between [Wright, 1991]. It also crosses the southwest margin of Los Angeles Basin (Figure 1). Therefore structural, depositional, and tectonic elements of each are present in the region. These features make the coastal belt an ideal laboratory for studying young tectonism and we think the observations recorded herein offer important constraints on any “world view” of the seismic setting of this part of the right-shear margin.
 Several data sources were not available to prior workers who attempted to achieve similar goals to ours. Multibeam seafloor imagery [Gardner et al., 1999, 2003; Gardner and Mayer, 1998] provides a new and highly accurate way to study the submarine geomorphology. The Los Angeles Regional Seismic Experiment (LARSE) study furnished seismic [Brocher et al., 1995] and potential field [Langenheim and Jachens, 1996] data that help to define the overall crustal thickness and geometry, the depth to basement, and the mid-crustal structural geometry. High-resolution multichannel seismic data and Huntec deep-tow boomer profiles collected on a grid of track lines spaced at 2 to 4 km [Normark et al., 1999a, 1999b] help constrain our interpretations of the upper crustal and near-surface structural geometry over most of the area.
2.1. Multibeam Data
 The bathymetry and acoustic backscatter of the area were mapped in 1996, 1998, and 1999 with high-resolution multibeam echo sounder (MBES) systems [Gardner et al., 1999; Gardner and Mayer, 1998]. The shelves, basin slopes, and proximal basin offshore of Santa Monica were mapped with a 95-kHz MBES system (Kongsberg Simrad EM1000) [Gardner et al., 2003]. The San Pedro shelf was mapped with a 300-kHz MBES system (Kongsberg Simrad EM3000) [Gardner et al., 1999] and the adjacent slope and proximal basin were mapped with a 30-kHz MBES system (Kongsberg Simrad EM300) [Gardner and Mayer, 1998]. These data provide geodetic-quality bathymetry and calibrated, geo-referenced, and coregistered acoustic backscatter data for the entire area. Navigation for the surveys used differential GPS-aided inertial navigation systems. Positions are accurate to ±1 m or less. All soundings have been corrected to mean lower low water using measured tides. The spatial resolution of the MBES data is <4 m for shelf depths and 8 m for slope/basin depths. The mapping of the combined area generated more than 500 million depth soundings and a similar number of acoustic backscatter measurements. See Gardner et al. [1999, 2003, 1997] for details.
 The general submarine geomorphology offshore of LA has been known since the works of Shepard and Emery  and Emery . These, and most subsequent, studies relied on contoured bathymetry interpolated between sparsely spaced single-beam echo sounder survey lines with poor navigation. Although the general bathymetry served well for years, the submarine landscape was far less known than its counterpart on land. With the advent of precisely located multibeam data and shaded relief maps based on high-resolution digital elevation models, Gardner et al.  were able to provide a description of the geomorphology of the shelves, slopes, and proximal basin of Santa Monica Bay that rivals any on land. Multibeam coverage is now continuous from the Malibu coastline to south of Data Point and Figure 3 shows the bathymetry of this area overprinted by the same type of physiographic provinces established by Gardner et al. .
2.2. Seismic Sources
 The Los Angeles Regional Seismic Experiment (LARSE) included multichannel seismic reflection profiling across part of the shelf and slope south of the greater Los Angeles area [Brocher et al., 1995]. Data were collected using a 137.7-L (8470 cubic inches) tuned-air-gun array and a 160-channel, 4.2-km digital streamer with a group interval of 25 m and digitizers located every 100 m. Locations of the segments of these lines within the coastal belt are shown in Figure 4.
 High-resolution, multichannel seismic reflection profiles were collected in a 2- to 4-km grid offshore of the Los Angeles and San Diego metropolitan areas (Figure 4) in 1997, 1998 and 1999 [Normark et al., 1999a, 1999b]. Data quality are good for the1998 and 1999 surveys where the acoustic source was a 35/35 cubic inch double-chamber, gas-injection air-gun and the receiver was a 24-channel streamer with 10-m-long groups of three phones each. Locations were determined by GPS input every 6 s, 24 hours a day. Differential GPS was used when available. Problems with the compressor system and the source/receiver array occurred in the 1997 survey that resulted in a limited number of lines in which data are noisy. We report on the parts of these data sets that are coincident with the area bounded by Lasuen Knoll in the south, Laguna Beach in the east, Long Beach Harbor in the north, and Palos Verdes Hills/San Pedro Sea Valley in the west (Figure 3). Track lines within that area are oriented northeast, roughly perpendicular to the coastline and most structures, or east- west, slightly oblique to most features. The grid covers the shelf and slope between the 3-mile limit and the San Pedro Basin floor. It comprises 700 line km of seismic profiles.
 A basic processing sequence was applied to all of the high-resolution multichannel data in order to obtain stack sections for interpretation. In addition the data were all migrated at 85% of stacking velocity and a 500-ms AGC was applied. All data are displayed in the time domain.
 No special attempt was made to attenuate multiples, as moveout differences were slight due to the short length of the streamer. The water-bottom multiple is a strong return, but is not a factor in the deeper water where it appears within the acoustic basement beneath most reflectors. Multiples are a problem in the midslope regions where they obliterate most of the useful data beneath them. The return from the seafloor on the shelf is indistinct; consequently, the corresponding multiple is weak beneath the shelf. Other multiples that originate in the reflective part of the section represent a bigger problem beneath the shelf.
 Several regional stratigraphic relations have a bearing on a Miocene high area southwest of LA Basin and are important to our study. The cross sections of Wright [1991, Figures 7, 8, and 9] and Yeats and Beall [1991, Figure 2] show that the entire stratigraphic section beneath the LA area thins to the southwest. Sediment accumulated almost continually in the central part of LA Basin from the early Miocene to the present. Individual rock units and time stages get consistently thinner and older rock units successively pinch out on basement in a southwest direction. The early and middle Miocene Topanga Group and the Miocene San Onofre Breccia, which form a thick basal section beneath large parts of LA Basin, are thin or absent in the coastal zone where Monterey Formation rests on basement. The San Onofre Breccia is a coarse-grained orogenic deposit that was derived from a highland of Catalina Schist that probably existed in the present location of the coastal belt of this study [Stuart, 1979]. The breccia is well represented in the San Joaquin Hills [Vedder, 1971] and in the subsurface of southeastern parts of LA Basin [Wright, 1991]. Dibblee  tentatively referred to the schist-bearing basal conglomerate within the Monterey Formation on Palos Verdes Peninsula as San Onofre Breccia (?), but that deposit is younger than schist-bearing breccia in the San Joaquin Hills. Similar basal conglomerate has been described from wells in many other parts of the coastal zone and it clearly is a time-transgressive rock unit that gets younger to the southwest [Yeats and Beall, 1991].
 The coastal belt lies mostly offshore where the structure and geologic history is best understood with subsurface data. Interpretations of seismic data fall victim to the technique-dependent differences in the way strata are named, subdivided, and described. Oil industry drill hole interpretations are weighted heavily on biostratigraphic criteria [Blake, 1991; Wright, 1991]. Geologic map units [e.g., Dibblee, 1999] are chiefly subdivided and named based on obvious lithologic differences. Offshore bottom samples yield very general lithologic and age constraints [e.g., Vedder, 1990]. The only reliably logged drill holes in the offshore part of the coastal belt are in the Beta field (Figure 4), so most of our interpretations are based solely on differences in seismic reflectivity. Seismic reflection profiles depict subtle differences in acoustic character that may not relate at all to lithologic subdivisions based on other criteria. Thus we could only differentiate a minimum of units and in many profiles the seismic character of mapped units is similar. Unconformities between units aided in differentiation in many instances. Figure 5 shows the reflective character of the simplified age and lithologic-unit subdivisions employed in this paper. Formation names and major age boundaries primarily follow Wright  and Blake  from their Los Angeles Basin studies.
3.1. Reflective Character of Basement Seismic Interval
 Everything beneath the Miocene Monterey Formation is considered part of the basement interval in this report. As so defined, basement is distinctly less reflective than intervals that overly it, although it is not all acoustically unstratified. Acoustically stratified parts are identified as reflective basement, as opposed to true acoustic basement. Acoustic basement is characterized by discontinuous reflections that are typically shorter than a few hundred meters in length. The reflective basement has reflections that exhibit highly variable wavelength and amplitude and they are wavy, nonparallel, and form irregular patterns of truncation. The contact between acoustic and reflective basement is indistinct or is obscured by multiples almost everywhere, so we made no attempt to define it.
 Acoustic basement is probably Catalina Schist in most places. The schist might be intruded by plutons of early Miocene age similar to the Miocene pluton on Catalina Island [Boundy-Sanders et al., 1993; Sorensen, 1985]. Reflective basement is most likely composed of Miocene volcanic rocks, Miocene San Onofre Breccia, or possibly rocks similar to those in the middle Miocene Topanga Group.
3.2. Reflective Character of Monterey, Puente, and Repetto Formations
 The Monterey, Puente and Repetto Formations are highly reflective in the seismic data. Commonly mapped biostratigraphic markers, lithofacies patterns and major formational boundaries cannot be located with any degree of confidence given the absence of drilling data in the coastal belt. Thus these formations are informally referred to as the Monterey-Puente-Repetto (MPR) interval. The Monterey Formation is present beneath most of the coastal belt, not its lateral equivalent the Puente Formation [Wright, 1991]. The younger Repetto Formation is known to be discontinuously exposed southwest of the Palos Verdes Fault and it is probably present beneath most areas of the coastal belt on the northeast side of the fault [Wright, 1991].
Figure 5 shows two examples that are typical of the seismic character of the MPR interval in an area where it can be positively identified from studied seafloor outcrops. The interval is typified by numerous parallel reflections, almost all of which are coherent over lateral distances of several kilometers except where they are disrupted by faults. The reflections have a very uniform vertical spacing and exhibit a narrow range of amplitudes both laterally and vertically, giving the interval an easily recognizable seismic signature. Reflections beneath the Santa Monica and San Pedro shelves have shorter wavelengths and lower amplitudes than they do elsewhere, giving the interval a slightly finer-textured appearance in those areas. The fine texture probably reflects thinner bedding caused by slower rates of accumulation of the hemi pelagic/biogenic sediments over the local highs. Local unconformities within the MPR interval are also common near faults, knolls and areas of Pliocene and Miocene deformation.
3.3. Reflective Character of Pico Formation and Temporally Equivalent Intervals
 The late Pliocene to Pleistocene Pico Formation is well represented between the Palos Verdes and Newport/Inglewood Faults [Wright, 1991]. The Pico Formation is subdivided into lower (late Pliocene), middle (early Pleistocene), and upper (late to early Pleistocene) parts primarily based on benthic foraminifers [Blake, 1991]. Unconformities are locally evident in the seismic data, but this interval has not been subdivided because any relationship to the better defined biostratigraphic subdivisions would be purely conjectural. Strata that occur in the same stratigraphic position as the Pico Formation, but are discontinuous with it, are (1) exposed on the lower slopes of the San Pedro Escarpment, (2) present in the subsurface beneath parts of Santa Monica Bay, and (3) buried deeply in San Pedro and Santa Monica Basins.
Figure 5 shows a typical example of the seismic stratigraphy of the Pico-equivalent strata that fill the basin between the Palos Verdes and Newport/Inglewood Faults (called Newport trough herein). Petroleum industry drilling data from the Beta oil field (Figure 4), summarized by Wright , indicate that the Pico Formation is a little over a kilometer thick in that area. Reflections are generally well developed and subparallel, although steeply crosscutting reflections irregularly occur. Continuity is highly variable; some reflections are coherent for several kilometers and others for less than 100 m and some parts of the section are almost unstratified. Amplitudes are highly variable and the vertical spacing between reflections is inconsistent.
3.4. Reflective Character of Late Pleistocene to Holocene Interval
 Sampling studies [e.g., Vedder, 1990] suggest that sediment of the late Pleistocene to Holocene interval is present almost everywhere in the coastal belt, but in most areas it is too thin to resolve seismically. There are several places where it is thick enough to resolve: (1) in San Pedro and Santa Monica Basins, (2) in Newport trough, (3) in the Santa Monica apron province and shelf, (4) along the edge of the San Pedro shelf break and on the slope below it, and (5) in other small, isolated basins. In the first two of these areas the interval is thick enough to exhibit distinct patterns of reflectivity.
 The seismic reflection data suggest that late Pleistocene to Holocene interval is several hundred meters thick in parts of San Pedro and Santa Monica Basins, but it is generally much thinner in Newport trough. It is stratified in most places with strong, subparallel reflections (e.g., left example of Figure 5). Coherency and amplitude are both highly variable along individual reflections and vertically through the section. Some reflections are coherent for several kilometers, but many others are continuous for less than 100 m. The entire section is reflective, but the vertical spacing of reflections is variable. In Newport trough the Holocene to late Pleistocene interval shows evidence of channeling. Dipping reflections are common and flat reflections occur in thin and semi-continuous zones (e.g., right example of Figure 5). Vertical spacing of reflections is highly irregular in these areas.
4. Structural and Stratigraphic Framework
 The following description of the coastal belt is organized into five areas, each with its own unique physiography and late Pliocene to Holocene structural history (Figure 6). These are the Newport trough area, the San Pedro shelf, the Palos Verdes Peninsula/San Pedro Escarpment area, Santa Monica Bay, and the deep offshore basins.
4.1. Newport Trough Area
Figure 7 is a simplified geologic map of the Newport-trough area on the shaded-relief base. Newport trough is one of the most prominent submarine physiographic features in the area and it is continuous with the midbathyal Gulf of Santa Catalina south of the coastal belt. Lasuen Knoll is a prominent submarine high and another large deep area lies to its west. The east edge of the Newport trough is a narrow shelf with a steep basin slope. San Gabriel and Newport Canyons form a complex network of channels and levees that are conspicuous on the shaded-relief image (Figures 4 and 7). The average water depth of this area is 450 to 600 m. On land the San Joaquin Hills dominate the landscape (Figures 1 and 7).
 Two regional north-northwest oriented fault systems and several smaller ones transect the area (Figure 7). These faults have local physiographic expression, but are subdued in many places. The Newport-Inglewood Fault Zone, has been mapped discontinuously along the top of the shelf break near the coast [e.g., Clarke et al., 1987] and it is seismically active (Figure 2). There are major structural and stratigraphic differences on either side of this fault, but it has very little or no local topographic expression probably because its surface trace follows the top of the shallow shelf where constant wave action bevels the surface. The Palos Verdes Fault Zone is continuous along the base of the steep west slope of Lasuen Knoll and has been mapped between there and the shelf break 1 km west of the head of San Gabriel Canyon [e.g., Vedder et al., 1986]. This fault has major physiographic expression and structural relief at the Lasuen-Knoll Escarpment, but it is hard to detect near San Gabriel submarine canyon where high rates of sedimentation and canyon formation are the dominant processes. One conspicuous short fault, the Avalon Knoll Fault of Marlow et al. , has a pronounced local physiographic signature. The fault trace is well defined in a prominent linear canyon near the western edge of area 1 (Figure 7). The Avalon Knoll Fault is not well defined north of the linear canyon where its displacement apparently diminishes in large northwest oriented anticlines (Figure 7).
 The crests of anticlines are conspicuous seafloor features in the area (Figure 7). Lasuen Knoll is the largest of these and its crest stands 350 m or more above the surrounding region. Several pronounced northwest oriented ridges on the slope south of the shelf are the result of folds in the pre-Holocene sedimentary rocks. Two ridges extend to the north-northwest from the north end of the Avalon Knoll Fault. Two others occur between the Avalon Knoll and Palos Verdes Faults. All of the folded ridges are elongated subparallel to the faults.
 Newport trough (Figure 1) is a major late Pliocene to Holocene depocenter between the Newport-Inglewood Fault Zone in the east and the Palos Verdes Fault and Lasuen Knoll in the west (Figure 8). In the central part of Newport trough the Pico and younger intervals exceed 1 km in thickness. Figure 9, is an interpretation of LARSE line 6 across this area. The Pico and Holocene intervals are characterized by discontinuous reflections that contrast with the continuous MPR reflections beneath them, which is even apparent in the low-frequency LARSE data. There also appears to be a thin, but continuous layer of stratified basement that overlies acoustic basement (possibly Catalina schist, since it is west of the Newport/Inglewood Fault) (Figure 9). Petroleum industry well data to the north beneath LA Basin suggests the stratified basement is probably Topanga Formation [Wright, 1991]. The buried floor of the basin is subhorizontal over a wide area and none of the fill is deformed except near the basin margins. The basin margins on each side of Newport trough are monoclines that are faulted by the large strike-slip faults that border the basin in most places (Figure 9). The southwest edge of the basin is the east flank of the Lasuen Knoll anticline (Figure 8).
 The Lasuen-Knoll anticline is well defined by reflections from the MPR interval, which dips east beneath the basin (Figure 8). The Pico and late Pleistocene-to-Holocene intervals abruptly thin on the flank of the fold. Numerous unconformities can be identified in this part of the interval that record the uplift history of the anticline and the subsidence of the basin (Figure 8). The underlying MPR interval has few unconformities and is comparatively uniform in thickness, indicating that folding and basin formation mostly occurred in late Pliocene to Holocene time. Subsidence and sedimentation at midbathyal depths have occurred at nearly equal rates in the basin axis since the inception of Monterey Formation deposition and the modern basin is about at the same depth that it began. The shoaling of the anticline crest served to impound the influx of clastic material supplied by Newport submarine canyon and the smaller canyons that occur on the basin slope along the east edge of the basin. Several features that are interpreted as relict channels are perched atop Lasuen Knoll (Figure 8), a further indication of the youth of the uplift. The top 200 to 400 ms of basin fill show abundant reflector-truncations and other evidence of channeling, indicating that the canyon complex has been active throughout the Holocene and possibly much of the Pleistocene.
 The monocline at the east edge of Newport trough is poorly documented in our data set because it and the Newport-Inglewood Fault are within the 3 miles of shore, limiting our access. The MPR interval lies at depths in excess of 1.5 km below sea level to the west of the fault (Figure 9) and the Monterey Formation is exposed in hills above sea level along the coast to its east [e.g., Morton, 1999]. A wedge-shaped body of poorly reflective rocks is evident in many of the seismic lines on the eastern edge of the gulf (Figures 8 and 9) low in the Pico interval. This body of rock is continuous well south of the coastal belt (H. Ryan, written communication, 2003). It might be extensive debris flow deposit or it could be prebasin rocks that are related to the Peninsular Ranges in the upper plate of the San Onofre Detachment Fault [Bohannon and Geist, 1998].
 The Palos Verdes Fault cuts the folded west edge of Newport trough and it has a different appearance in almost every profile that crosses it. The fault at Lasuen Knoll is well west of the monocline (Figure 7). The Palos Verdes Fault dips steeply west at the north end of Lasuen Knoll and has the geometry of a normal fault in 2-D profile (Figure 8). The MPR interval defines a broad open anticline in its footwall and there is a deep basin filled with Pico interval and younger strata to its west (Figure 8). Two kilometers to the north, the fault has a reverse geometry with a tight anticline in the MPR interval in its hanging wall and the deep basin is in its footwall (Figure 11, line 68). The fault zone is very complex at the southwest end of the knoll and its geometry is ambiguous on the seismic profiles. The sharp relief associated with the fault at Lasuen Knoll suggests its youth, but this stretch of the fault is not known for seismic activity (Figure 2).
 The Palos Verdes Fault cuts the folded basin edge between Lasuen Knoll and the Huntington/Newport shelf break. Here the fault zone widens into several branches, of varied geometry in 2-D profiles. Most of the branches are buried by deposits of the Pico interval (Figure 10, lines 66 and 67). The branches that can be traced to the surface have small displacements and little or no surface expression, but this section of the fault is associated with a small amount of recorded seismicity (Figure 2). The LARSE data and drilling in the Beta field indicate that the Palos Verdes Fault dips gently east near the surface and dips steeply west at depth [this report, Figure 10; Wright, 1991, Figure 13]. The Pico and younger intervals are thick to the east of the Palos Verdes Fault, but thin or absent to its west. The MPR and older intervals show little or no change in thickness across the zone. This suggests that the bulk of the fault displacement has occurred in the last few million years.
 A faulted and folded ridge occurs 3 to 4 km west of the Palos Verdes Fault (Figures 7 and 10). The ridge is an anticline in the lower Monterey Formation [Vedder et al., 1986] and the unnamed fault to its east offsets Holocene strata. The fault also locally defines the western basin margin of the Pico and younger intervals (Figure 10). To the north the unnamed fault bends northwest, away from the Palos Verdes Fault, where it looses displacement in an anticline beneath Pliocene strata (Figure 7). To the south the unnamed fault cannot be recognized at any depth in the profiles south of line 68. It is only present in between Lasuen Knoll and the Huntington/Newport shelf break.
Marlow et al.  named the Avalon Knoll Fault (Figures 7 and 11) and they stated that “it might extend onshore and contribute to the missing balance of Quaternary uplift determined for the Palos Verdes Hills and not accounted for by vertical uplift along the onshore Palos Verdes Fault.” There is no evidence that it does that, however. The Avalon Knoll Fault has pronounced physiographic expression, but it is not a continuous structure (Figure 7). The fault ends to the north at a northwest oriented anticline that is cored by Monterey Formation on the basin slope (Figure 7). The Avalon Knoll Fault has a straight surface trace in the area where it is best expressed, is up on its east side (about 500 m), and it dips steeply west near the surface (Figure 11, line 68). The fault is buried north of there and where it possibly has a small normal component of displacement (Figure 11, line 67). Its pronounced surface expression and the apparent control it has exerted on Holocene sedimentation patterns (Figure 11) suggest that it has been active during the Holocene. Substantial thickness differences in all of the seismic intervals identified on both sides of the fault strongly suggest a large component of horizontal displacement. The northwest orientation of the anticline at its northern termination and the thrust-faulted anticline to its west (Figure 11, line 67) are consistent with a constraining-bend-geometry in a right-slip system.
4.2. San Pedro Shelf and Basin Slope
Figure 12 is a simplified geologic map of an area that includes the San Pedro Shelf. The base map also shows 300-kHz acoustic backscatter draped over shaded relief to aid in identifying outcrop areas. The shelf is the most prominent submarine physiographic feature and it is deeply incised by San Pedro Sea Valley on its west edge. The basin slope is broad, relatively steep, and is cut by small straight canyons, except for a large area near the base of the slope with a smooth, gentle, undulatory topography, that we call the lower-slope structure (“lss” in Figure 12). The terrain on land to the north of the Long Beach coastal plain is smooth with the exception of low folded hills marking the inland trace of the Newport/Inglewood Fault Zone (Figure 1). The Palos Verdes Fault Zone, which bisects the shelf in a northwest direction, has subtle expression in the bathymetry and follows the northeast edge of the shelf outcrops (light shades in the backscatter, Figure 12). The average water depth on the shelf is less than 50 m.
 The Palos Verdes Fault on the shelf is known from seismic data and petroleum industry well data to be a steep zone in the axis of a large anticline or monocline that is elevated on its southwest side. More than 1.5 km of vertical displacement has been documented across the entire folded zone in the shelf area [e.g., Fischer et al., 1987a; Wright, 1991]. A right-lateral component of slip is also commonly assumed for the fault because (1) it has a northwest orientation, (2) nearby fold axes are consistent with right slip, and (3) some differences in basement and stratigraphy on either side of the zone are easier to explain with strike-slip models [Fischer et al., 1987a]. Although no lateral displacement has actually been documented on the shelf, about 300 m of right slip (averaging 2.5 to 3.8 mm/yr) was established on land east of the Palos Verdes Hills where the fault offsets the ancestral Los Angeles River channel [Stephenson et al., 1995].
 The Palos Verdes Fault is a young structure on the San Pedro Shelf. The fault offsets near-surface reflections in Holocene strata on high-resolution seismic profiles and there is subtle seafloor relief associated with it in many places [Fischer et al., 1987b]. However Holocene displacement is slight and most of the fault activity took place prior to deposition of the upper part of the Pico interval. Thus the fault is chiefly pre-early Pleistocene in age. This is documented on our seismic reflection profiles, three of which are shown in Figure 13. The vertical component of offset is greater than 1 km in middle Pico Formation and the underlying strata, but the base of the upper Pico Formation laps across the fault zone with little or no observable offset. Reflections in the upper Pico Formation above the onlap are flat lying with only slight disruption evident by faults. Strata of the upper Pico Formation are not folded, although older strata are. This strongly suggests that vertical displacement rates have slowed considerably in the last 100 kyr.
 The character of the San Pedro shelf is different on either side of the Palos Verdes Fault Zone. On the northeast side the shelf is underlain by unconsolidated to poorly consolidated Holocene and Pleistocene sediments that are easily eroded. The position of the southern shelf break in the thick sediment section (Figure 12) is a function of the rate of sedimentation versus the rate of erosion, which chiefly occurs by slumping at the shelf edge, and it may have varied with time. On the southwest side of the fault there are outcrops of folded Miocene rocks that resist erosion. The position of the shelf edge southwest of the fault zone has essentially been fixed in space at the edge of the uplifted competent rocks. The flat surface and uniform depth of the shelf on both sides of the fault results from a balance of sediment trapping northeast of the uplift and wave erosion of the more competent uplifted rocks to its southwest. Most of the sediments introduced by the rivers that cross LA Basin are transported southward across the shelf to the heads of San Gabriel and Newport Canyons. Some sediment has apparently spilled to the southwest across the uplift and is preserved in a lens-shaped Holocene (?) and Pleistocene deposit at the shelf edge and on the basin slope [Fischer and Lee, 1992]. The San Pedro Sea Valley incises the western shelf edge where poorly indurated Pico interval strata occur in the axis of a large syncline with an east-west orientation (Figure 12).
 Seismic profiles that cross the base of the basin slope to the southeast of the lower-slope structure (“lss” in Figure 12) show numerous fold and fault structures. The structures are primarily delineated by the MPR interval that is apparent in all the sections as well-defined, continuous, and parallel reflections. The interval generally dips southwest parallel to the slope, but it is also deformed into numerous broad, open folds. Some of the folds result in seafloor relief that is evident in the slope-surface morphology depicted on the shaded-relief map (e.g., Figure 12). Most of the folds are cut by faults that exhibit reverse separations. In many places the younger intervals, including the Quaternary, can be seen filling swales developed on the upslope flanks of topographically expressed anticlines. The Miocene and Pliocene intervals (MPR interval through upper part of Pico interval) dip to the southwest beneath the Quaternary fill of San Pedro Basin (Figure 14). Some open folds near the edge of the basin are evident in the basin fill, but dips on the fold flanks are typically 10° or less even in the older rocks beneath the fill. The fill is undeformed to the southwest of the sections. The deformation documents post-late Miocene upper crustal shortening that we estimate to be about 10% in a northeast-southwest direction. The adjacent San Pedro Basin escaped this strain except near its edge.
 The uplift of the basement, relative to the basin, is primarily the result of regional folding. Thrust and reverse faults might be locally important to the uplift, but most faults are secondary to the smaller folds. The contact between basin fill and dipping Miocene strata is well defined in the bathymetry, but is not demonstrably faulted anywhere. The monoclinal folding that has resulted in the uplift of the strata is young and abrupt and the basin fill simply laps onto the fold flank.
 The lower-slope structure (“lss” in Figure 12) is an area of gentle topography on the lower basin slope off the west side of San Pedro shelf. The structure forms an anomalous westward broadening of the basin slope that has the morphology of a large slump (Figure 12). Seismic reflection profiles across the structure (Figure 15) show that acoustic basement is draped by thin sediment of the Pico interval and that the sediments and basement are cut by several normal faults that break the seafloor near its west edge. These observations suggest the structure is late Pliocene to Quaternary in age and that the faulting in the distal region is young and perhaps still active. The faults dip west and are down to the west and the top-of-basement surface defines an east tilted block in most profiles. Small buried horsts and grabens are also evident. Basement rocks might be exposed along the western edge of the structure in the uplifted crest of the main tilted block (line 200, Figure 15). The thick fill in San Pedro Basin is in fault contact with, or onlaps, the steeply inclined west and southwest edges of the basement block. The older fill in some places is slightly inclined to the west and more highly deformed in others. Numerous straight channels dissect the surface of the upper slope immediately above the lower-slope structure. Most of the channels do not continue downslope beyond the upper boundary of the lower-slope structure. The lack of continuity of the upper slope channels is an enigma. The apparent headwall of the structure appears to be overlapped by Pico Formation without any offset or disruption. The faults document a local extensional strain estimated to be about 10° in a west-northwest direction. The unusual surface morphology of the lower-slope structure might be due to slumping caused by slight westward stretching of rocks in the lower-basin slope during late Pliocene and Quaternary time.
 Several examples of landslide scars, escarpments, and debris-flow deposits can be identified on the lower parts of the basin slope and in the proximal fill of San Pedro Basin. The deposits show up as locally unstratified lenses within the basin fill adjacent to the basement/fill contact. One deposit is identified in Figure 15. Scars and unusual topography are evident on the shaded relief image (Figure 12).
4.3. Palos Verdes Shelf and San Pedro Escarpment
 The most pronounced topographic and structural relief in the coastal belt occurs along the San Pedro Escarpment offshore of the Palos Verdes Peninsula (Figure 12). The crest of the Palos Verdes Hills is 400 to 450 m above sea level and the top of Catalina Schist basement is exposed at elevations of about 330 m. Ten kilometers to the southwest, the floor of San Pedro Basin lies 800 to 850 m below sea level and the top of schist basement is buried under an additional kilometer or more of young sediment. The escarpment has slopes as steep as 14° [Bohannon and Gardner, 2004, Figure 2].
 Some of the relief is the direct result of folding. Dibblee  indicated that the Palos Verdes Hills are a large anticlinorium with a northwest axis and schist in its core. Monterey Formation and interbedded basalt flows dip subparallel to the topographic slope on the southwest flank of the anticlinorium. Seaward dipping submarine outcrops of Monterey Formation, basalt flows and possibly a small amount of schist occur on the narrow shelf southwest of Palos Verdes [Dibblee, 1999]. On the basis of their study of marine terraces, Ward and Valensise  estimated that the anticlinorium has grown to its present elevation from an original flat seafloor about 850 m below sea level in the last 2.8 Myr. The hills have been emergent since the last million years.
 A large part of the relief is also the result of faulting. Ward and Valensise  modeled the anticlinal growth as a response to vertical displacement on the Palos Verdes Fault Zone. Northeast of the anticlinorium, the Palos Verdes Fault is a reverse fault with a displacement on the basement surface of 1200 m [Wright, 1991]. Southwest of the fold the San Pedro Escarpment Fault has been mapped at the base of the escarpment [Nardin and Henyey, 1978; Wright, 1991]. The linear trace of the abrupt break at the base of the escarpment strongly suggests fault control, but little is known about the fault. On many regional maps [e.g., Vedder et al., 1986] the contact at the base of the escarpment is simply mapped as unfaulted basin fill that laps onto the uplifted fold flank. Figure 16 shows three examples of seismic profiles that cross the base of the escarpment. Shallow reflections in the fill are subhorizontal and deeper reflections have a gentle apparent dip to the southwest. The southwest dipping contact between the fill and the uplifted fold flank is steeper than the dip of strata on either side as well as steeper than the slope of the escarpment (Figure 16). Thus we interpret the contact to be a fault and not simply an onlap of young sediment on a dipping fold flank. The fault dips moderately southwest, displaces very young strata, and has a normal component of displacement on the order of 1 to 2 km, which is the approximate depth to the base of the basin fill. The Palos Verdes and San Pedro Escarpment Faults have similar magnitudes of vertical offset and they share the same northwest orientation (Figure 12), although one is contractional and the other is extensional.
 The San Pedro Escarpment is the site of numerous slope failures [Bohannon and Gardner, 2004; Hampton et al., 2002]. The largest of these originated in the large failure scar that is imaged on the escarpment (Figure 12) and resulted in a massive deposit on the Basin floor called the Palos Verdes debris avalanche deposit [Bohannon and Gardner, 2004; Locat et al., 2004]. The Palos Verdes debris avalanche deposit is about 7500 years old [Normark et al., 2004], has a volume of about 0.34 km3, and is the product of a single failure event that might have generated a local tsunami [Bohannon and Gardner, 2004]. The failures are symptomatic of youthful uplift, which results in unstable slope conditions, on the southwest side of the Palos Verdes anticlinorium. The escarpment should be considered a likely candidate for future slope failures.
4.4. Santa Monica Bay
 The main physiographic elements of Santa Monica Bay are (1) a 4 to 10 km-wide, shallow shelf that stretches from Point Dume to Palos Verdes, (2) a rocky marginal plateau between Redondo and Santa Monica Canyons, and (3) a deep, seaward sloping apron between Santa Monica Canyon and the Point Dume shelf (Figure 17). The deep-water Santa Monica Basin is separated from the apron by a low-relief ridge that extends northwest to southeast between Dume Canyon and the lower reaches of Santa Monica Canyon (Figure 18). Three major faults, the aforementioned San Pedro Escarpment Fault, the Redondo Canyon Fault, and the San Pedro Basin Fault (Figure 17) are important in this area.
Nardin and Henyey  described a large reverse fault, called the Redondo Canyon Fault, that they thought branched from the Palos Verdes Fault Zone to follow the axis of Redondo Canyon to the base of the slope. The multibeam bathymetric data clearly show that the floor of the canyon is sinuous and not fault controlled (Figure 17). Our seismic data indicate that the Redondo Canyon Fault is a zone of reverse faults that is oriented west-northwest and crosses the lower reaches of the canyon (Figure 18). The main branch of this zone passes just north of the area of high backscatter (outcrops of MPR interval) on the shelf just southeast of the canyon (Figure 17). As Nardin and Henyey  suggested, this fault probably connects with the Palos Verdes Fault Zone (Figure 18), absorbing some of its displacement. The MPR interval is exposed on the Palos Verdes shelf south of the fault (Figure 17) and is buried by up to a kilometer of Quaternary strata beneath the Santa Monica shelf on its north side. Thus, vertical separation on the Redondo Canyon Fault Zone accounts for a large part of the difference in structural elevation on the basement surface between the Palos Verdes Hills and the marginal plateau of the Santa Monica shelf. Figure 18 shows that the Redondo Canyon Fault Zone splits into several branches in an area of faulted anticlines on the lower slope west of Redondo Canyon. The Redondo Canyon Fault Zone might be active, particularly where it crosses Redondo Canyon. Here the fault appears to extend to the surface on Line 103 (not shown in a figure). Holocene deposits cover many branches of this fault zone elsewhere. The Holocene deposits are slightly deformed above some of the branches of the Redondo Canyon Fault Zone, suggesting that it might primarily be a blind-reverse fault in places.
 The San Pedro Basin Fault (Figures 17 and 18) is a linear, northwest-oriented zone of faults, some of which displace Holocene strata, extending from the southwestern part of San Pedro Basin to Dume Canyon [Vedder et al., 1986]. Near the south edge of our study area this fault zone occurs at the southwest margin of San Pedro Basin where it juxtaposes 500 to 700 m of Quaternary basin fill against a basement high that extends north of Redondo Knoll (line 46, Figure 18). The faults bounding the basement high all have normal-separation. The seafloor and the underlying sediments near the basement high are deformed in a broad box anticline in the San Pedro Basin Fault Zone. The fault zone is wide near the surface in this area, but normal-separation and reverse-separation faults merge at depth creating a flower structure (line 46, Figure 18). North of the Redondo-Knoll basement high the San Pedro Basin Fault creates another high that separates the Quaternary fill of San Pedro and Santa Monica Basins (lines 44 and 45, Figure 18) in an area where the two basins might be otherwise joined. The fault might not break the surface, but Holocene fill is displaced and the sea floor is deformed in a narrow arch at the fault zone (lines 44 and 45, Figure 18). The near vertical nature of the zone, the presence of flower structures, and the fact that normal- and reverse- separations are present in the same narrow zone all suggest that the displacement on this fault is primarily strike-slip. The folds in the basin fill define an abrupt zone of contraction across the fault zone where it crosses the deep basins. Normal-separation faults marking the margin of San Pedro Basin at Redondo Knoll probably originated in a prior phase of extensional faulting.
Figure 18 indicates that the structural character of the northeast margin of San Pedro Basin is more complex to the west of Redondo Canyon than it is along the San Pedro Escarpment. Most of this complexity corresponds to the intersection of the three large fault zones (Figure 18). The basin is a deep and narrow graben bounded by faults with large normal separations where it is crossed by line 46 (Figure 18). The normal-separation faults on the north side of the basin are related to the San Pedro Escarpment Fault. The northern margin of the graben is deformed by blind-reverse faults where it is crossed by line 44 (Figure 18) and these faults are probably continuous with the Redondo Canyon Fault Zone. The northern margin of the basin is not highly faulted where it is crossed by line 42 (Figure 18), because the Redondo Canyon and San Pedro Escarpment Faults merge with the San Pedro Basin Fault east of the profile. The basin fill is depositional on the steep, but unbroken, southwestern flank of the Santa Monica shelf uplift in the latter area.
 Normal-separation faults are conspicuous in the San Pedro Basin Fault Zone where it crosses Santa Monica Canyon (lines 41 and 48, Figure 19). Here a narrow, fault-bounded basement high characterizes the fault zone. The high is flanked on either side by thick basin deposits. Although there is a thin Quaternary cover above the basement high, the Santa Monica and San Pedro Basins are distinctly separated by the fault zone. The basement high within the fault zone widens slightly with depth because the larger faults dip away from it. The fill of San Pedro Basin thickens toward the fault zone, forming a wedge-shaped deposit that is bounded on its southwest side by a normal-separation fault (lines 41 and 48, Figure 19). The entire wedge of basin fill also dips southwest and is perched high on the folded flank of the marginal plateau. Although this wedge is continuous with the fill in other parts of San Pedro Basin it is elevated and deformed, so we call this the elevated portion of San Pedro Basin. It is also higher than Santa Monica Basin on the other side of the fault zone. The fill of Santa Monica Basin is tilted to the southwest, away from the fault zone. Holocene basin fill is confined to Santa Monica Basin where it is thick due to a local influx from Santa Monica Canyon.
 A low ridge separates the shelf apron northwest of Santa Monica Canyon from the deep basin to the southwest. Seismic line 52 crosses this ridge in an area of high backscatter where the older basin fill (probably early Pleistocene or late Pliocene in age) is exposed in the core of an anticline (Figure 17). The southwest side of the ridge slopes gently down to the modern floor of Santa Monica Basin 7 km to the south and the older fill dips subparallel to the slope. The north flank of the anticline is cut by a steep reverse-separation fault that is part of the San Pedro Basin Fault Zone (line 52, Figure 19). Thick Quaternary basin fill, mostly the upper part of the Pico interval, in the elevated part of San Pedro Basin forms the underpinning of the entire shelf apron to the northeast of the San Pedro Basin Fault Zone. Holocene deposits overlap the fault and are continuous down slope from the apron to the deep basin in most places between Santa Monica and Dume Canyons. However, the Holocene beds do not overlap the highest part of the folded ridge (line 52, Figure 19) so the anticline there might be a Holocene structure.
 The LARSE 02 seismic profile shows that the Quaternary fill in the elevated portion of San Pedro Basin beneath the Santa Monica shelf apron forms a wedge-shaped deposit that is more than 1.5 s thick near the shelf break [Fisher et al., 2003]. The deposits thin gradually toward the marginal plateau where they lap on to the erosion surface at the top of folded MPR interval rocks. The slope on the erosion surface is gentle and there is no evidence of large-scale faulting between the shelf apron and the marginal plateau, consequently the Palos Verdes Fault does not extend into the western parts of Santa Monica Bay [Fisher et al., 2003]. Late Pliocene and Pleistocene strata (Pico Formation) of LA Basin are about 1-km thick beneath the nearby Santa Monica shoreline [Wright, 1991; Figure 15]. Thus there also is no reason to suspect that the Palos Verdes Fault extends to the northwest of the peninsula beneath the shoreline. Fisher et al.  concluded that the Palos Verdes Fault looses its displacement offshore of Manhattan Beach on the north flank of the marginal-plateau uplift. The late Pliocene and Pleistocene deposits of western LA Basin beneath Santa Monica are continuous with those in the elevated portion of San Pedro Basin beneath the shelf apron.
 There was continuous subsidence and deposition connecting the Santa Monica area, the shelf apron, and San Pedro Basin proper during the entire late Pliocene and Quaternary time period. The fill (Pico Formation and equivalent strata) in these parts of LA and San Pedro Basins forms a wedge that wraps around the western nose of the marginal plateau and the northwestern termination of the Palos Verdes Fault Zone. The wedge is thin where it laps on to the plateau and it is thick adjacent to the Dume and San Pedro Basin Faults.
4.5. Area 5: Deep-Water Santa Monica and San Pedro Basins
 Santa Monica and San Pedro Basins are 700 to 900 m deep and relatively flat-floored (Figure 1). Redondo Knoll, a basement high that is about 300 m deep, separates them. Avalon Knoll borders San Pedro Basin on the southeast. Santa Monica is the larger of the two basins, and it extends well beyond our study area to the northwest (Figure 1). Figure 20 is a map of the two basins based on regional Sea Beam bathymetry merged with MBES data where available. A multichannel seismic reflection profile (L490SC-125) longitudinally crosses these two basins (Figure 4). These data and an interpretation are shown in Figure 21.
 The fill of Santa Monica Basin forms a thick wedge of sediment that onlaps the west tilted basement block of Redondo Knoll (Figure 20). The fill in the northwestern part of the basin exceeds 2.5 km in thickness. Two distinct sequences can be delineated; a shallow one that is flat lying and a deeper one that gently dips to the northwest away from Redondo Knoll. Each sequence is a kilometer or more in thickness in the basin depocenter. The dip in the lower sequence is obvious with high vertical exaggeration, but is only about 1° in a natural cross section (Figure 20). The internal geometry of the fill sequences suggests that they are local deposits rather than fragmented parts of once-more-widespread deposits such as the MPR interval strata. The slight west-tilt of the deeper sequence and the Redondo-Knoll block are probably due to a normal component of displacement on the San Pedro Basin Fault Zone that bounds the east side of the knoll. There is very little or no evidence of faulting on the margins of Santa Monica Basin that are within our study area (Figure 20).
 San Pedro Basin is smaller than Santa Monica Basin and the basin margins are faulted in most places (Figure 20). Santa Monica Basin is also not as deep and the fill totals slightly more than 1 km in thickness. In the seismic data San Pedro Basin has the appearance of a half graben that developed in the hanging wall of the San Pedro Basin Fault Zone (Figure 21) and this fault probably bounds the entire south and west margins of the basin (Figure 20). The surface trace of the San Pedro Basin Fault Zone is mapped in Figure 20 based on geomorphology and it is not straight as one might expect of a strike-slip fault. It is oriented north-south on the east side of Redondo Knoll (Figure 20), where it has the 2-D geometry of a normal fault (Figure 21). Northeast of Catalina Island it is oriented northwest and has the type of irregular surface trace that is typical of a normal fault at the base of a mountain front (Figure 20). Our best estimate of the type of displacement bordering San Pedro Basin is shown in Figure 21. There is little or no evidence of faulting on the easternmost edge of the basin (Figure 20). The gross geometry of the basin is that of a large pull-apart in a northwest oriented right-slip system and it may be just that. However, the magnitude of right slip, if any exists on any of the faults, is unknown.
 Santa Monica and San Pedro Basins were major Quaternary depocenters and may have accumulated Pliocene sediment as well. Teng and Gorsline [1989, 1991] showed roughly 2 s two-way travel time (TWTT) of sediment filling nearby Santa Monica Basin and they thought the deepest fill might be as old as early Pliocene (unit P1). If this is true, then there might be time-equivalent strata to the Repetto Formation deep in the basins, but their assessment was based entirely on seismic interpretations. Much higher Holocene sedimentation rates than were thought to exist previously have been reported from the Deep Sea Drilling Project hole in Santa Monica Basin [Fisher et al., 2001, 2003; Lyle et al., 1997; Normark et al., 1998]. Piper and Normark  and Fisher et al.  suggested that the upper 1.5 km of fill in Santa Monica Basin is no older than Quaternary and might be entirely younger than 600 ka. Even if pre-Holocene sedimentation rates were not as high during the early phases of basin filling, unit P1 of Teng and Gorsline [1989, 1991] is probably no older than late Pliocene, so strata equivalent in age to the Repetto Formation are not likely to be present.
 Faults can be interpreted to cut the deeper fill in the central parts of each basin, but few of them have much offset. Both sequences are mostly undeformed over large areas. The demonstrable high rates of sedimentation [Fisher et al., 2001], the lack of internal deformation, and the local nature of the fill sequences all suggest that these basins are late Pliocene and younger features that primarily developed in response to rapid regional subsidence. A slight amount of west-northwest extension, probably less than 10%, is evident across San Pedro Basin, but there is little or no evidence for extension in Santa Monica Basin. The amount and rate of subsidence greatly exceed what would be predicted given the magnitude of extension in either basin.
5.1. Early Miocene to Early Pliocene Tectonism
 Approximately 90° of clockwise rotational translation of the Santa Monica and San Gabriel Mountains of the western Transverse Ranges (WTR) occurred relative to the Santa Ana Mountains and the rest of the Peninsular Ranges mostly during Miocene time [e.g., Hornafius et al., 1986; Luyendyk, 1991]. The LA depression widened throughout that time period in the wake of WTR rotation. Crouch and Suppe  speculated that a large area of uplifted and unroofed schist-basement was produced in the floor of the depression as the angular distance increased between the WTR and Santa Ana belts.
Wright's  data and interpretations indicate that the unroofing of Catalina Schist in the greater LA depression might not have been a simple process. Figure 22 summarizes some of the most important points in Wright's  cross sections. Catalina Schist was entirely unroofed prior to deposition of Topanga Group (Miocene) strata everywhere southwest of the Newport/Inglewood Fault Zone. That area may have been initially mountainous and provided a source of schist debris for the San Onofre Breccia and Topanga conglomerates [i.e., Stuart, 1979; Wright, 1991]. Basement has not been drilled over a large area of central LA Basin northeast of the Newport/Inglewood Fault Zone where the Topanga is very thick, but data near the Newport/Inglewood Fault Zone suggest that Peninsular-Ranges basement is present there. Paleogene sedimentary rocks or Peninsular Range basement are known beneath most of the northeastern half of LA Basin to the southwest of the Whittier Fault and in some places adjacent to the northeast side of the Newport/Inglewood Fault Zone. Northeast of the Whittier Fault the basement is probably all of Peninsular Range affinity. Thus the basement occurs in fault-bounded belts with northwest orientations. The bounding faults are mostly steep with suspected right-slip displacements.
 The kinematic evolution of WTR-rotation in the LA area is commonly explained by one of two types of models, “deck of cards slip” and “core complex extension.” Deformation is accommodated by strike-slip between large northwest oriented blocks of crust in the deck of cards models [e.g., Hornafius et al., 1986; Wright, 1991], but the steep block-bounding faults and horizontal displacements do little to explain how Catalina Schist became exposed. Core complex extension was evoked to solve that problem [e.g., Bohannon and Geist, 1998; Crouch and Suppe, 1993] by a process of tectonic denudation in which schist from the middle crust became exposed after the overlying Peninsular Ranges/WTR basement was removed by detachment faulting. Wright's  subsurface analysis shows strike-slip-bounded basement blocks with northwest orientations, which are consistent with the deck of cards models, so he thought that schist had already been exposed by early Miocene time and that it simply translated northward to its present position beneath the coastal belt as the LA depression widened in response to WTR rotation. The early Miocene exposure of the schist is also consistent with the sedimentary record provided by the San Onofre Breccia and Topanga Group, both of which contain schist debris. Core complex extension is probably the best way to explain the early Miocene uplift and exposure of schist. Bohannon and Geist  depicted an early Miocene rift with a north-south orientation that developed about 20 Myr ago between the then adjacent WTR and Peninsular Ranges blocks. They postulated that most of the upper plate denudation occurred on the Oceanside Detachment Fault, a low-angle normal fault with an east dip that was originally described by Crouch and Suppe . A slight modification of one of Wright's  cross sections shows how the subsurface of the LA depression might look if the latter model is correct (Figure 23a). The exposed schist formed an eroding high area during this time. The LA depression, with its thick post-Topanga Group sediments, began subsiding during later phases of the rifting after the Oceanside Detachment Fault became inactive. These later phases (after 15 Ma) involved more rotation of the WTR block, rotation that was accommodated by deck of cards slip in the LA depression.
 The entire area around Los Angeles subsided during most of the middle and late Miocene and as much as 3 km of Relizian-age and younger sediment accumulated in the central trough of the basin [Wright, 1991]. The subsidence must have been very rapid. Some, if not all, of the schist in the coastal belt was above sea level in the early Miocene in order to provide a source of clasts in the San Onofre Breccia, yet the middle and late Miocene sediment was deposited in a marine environment with benthic fauna commonly interpreted to live at midbathyal or greater depths [Blake, 1991]. Wright's  interpretations show that the marine sedimentary units are much thinner beneath the coastal belt where they rest directly on schist basement, than they are in the central trough of the LA depression. Only the oldest Monterey (Relizian and some Mohnian) is present in the uplifted parts of the coastal belt, younger beds were not deposited or are eroded. These observations indicate that the initial subsidence, probably more than 1 km, took place during the Sausesian (about 16 to 20 Ma) shortly after the schist was uplifted and exposed during the initial rifting episode. The central trough of LA Basin has subsided as much as 6 km since the middle Miocene (Figure 23). The interpretations shown in Figure 24a suggest that the oldest beds of the MPR interval and the underlying Topanga Group were deposited on a surface of schist that might have been tectonically exhumed from beneath the Oceanside Detachment Fault and its upper plate of Peninsular Ranges basement during the last phases of normal displacement on that fault. The depositional relations at the base of the MPR interval can be seen in seismic data at the northernmost end of the LARSE 1R seismic reflection line (Figure 23b). A buttress unconformity, with onlap to the south-southwest, is evident in the reflections at the base of the interval.
 All of our high-resolution multichannel seismic lines that cross the base of the slope show similar onlap relations of the MPR interval on the basement. These relations are documented in Figures 11, 14, 15, and 18. The MPR interval, which is easily identified by its uniform stratification and continuous, even reflections that exhibit little variability in amplitude, exceeds 0.5 TWTT in thickness and is well represented everywhere northeast of the basin slope and beneath LA Basin. The interval is wedge shaped beneath the slope and the strata all terminate abruptly at the basement contact. We interpret this relation to be a buttress unconformity that is overlapped by late Pliocene and younger strata, which rest directly on basement everywhere in San Pedro and Santa Monica Basins.
 The thinning of the sedimentary strata of the MPR interval is also conspicuous across the Santa Monica shelf between the northern parts of the LA and Santa Monica Basins. Figure 24a is an interpretive cross section that illustrates these relations. Monterey Formation and younger strata are about 6 km thick in the central trough of LA depression. The lower part of the Monterey Formation on the Santa Monica shelf is present in a thin, gently folded sheet. In the deep water at the southwest end of the cross section a thin veneer of Quaternary sediment directly overlies Catalina Schist basement. The schist-cored basement high at the southwest end of the section is part of Redondo Knoll, a high that separates Santa Monica and San Pedro Basins.
 The lack of sedimentary cover on Redondo Knoll indicates that it was an area of erosion or nondeposition during late Miocene and Pliocene time. The absence of late Miocene and Pliocene sediment is typical of the interior parts of Santa Monica and San Pedro Basins. Vedder et al.  mapped widespread Miocene sedimentary rocks on the submarine ridge flanking the southwest margin of Santa Monica Basin. LARSE line 03 crosses Santa Monica Basin and shows the relationship between Miocene strata, basement, and basin fill on this part of the basin margin (Figure 24b). Line 03 shows that the Miocene strata mapped on the ridge northwest of Catalina Island by Vedder et al.  onlap basement and pinch out beneath the southwest edge of Santa Monica Basin. Between that pinch out and the one beneath the shelf edge, the younger basin fill directly rests on basement composed of schist and possibly igneous rocks of Miocene age. Thus two opposing middle and late Miocene Basin margins are preserved beneath each side of the modern Santa Monica Basin. Because middle-Miocene to late Pliocene sediments were deposited everywhere there was a deep basin or adequate subsidence, we conclude that the Santa Monica-San Pedro Basin area was emergent to shallow and not subsiding during that time interval. This area, which includes Catalina Island, probably formed a sill between the open ocean to the west and the deep basins in the LA region. This shelf or emergent area was a relict of the earlier schist highland that was the main source for the San Onofre Breccia.
 In summary the LA region looked very different in early Pliocene time than it does today. Most of what is land now was under water then and the local southwest margin of the Monterey/Repetto midbathyal basin had roughly the same orientation and location as that of the present-day basin slope. However, the basin slope had the opposite sense of physiographic and structural relief. The deep basin was in the LA area. The modern offshore area was emergent or a shallow shelf. The Palos Verdes area was a small submarine knoll on the northeast slope between the shelf and basin. The Santa Monica Mountains, which were submerged during most of the Miocene, had started to rise above sea level in the early Pliocene.
5.2. Late Pliocene to Holocene Tectonism
 The period from the late Pliocene to Holocene was a time of major changes in the regional physiography, depositional patterns, and structural relief of the coastal belt. During this time the Santa Monica and San Pedro Basins rapidly subsided to their present midbathyal depths. The shelves and Palos Verdes Hills were uplifted from deep submarine knolls. The Los Angeles Basin area filled to sea level and above with sediment that was impounded behind the rising sill created by the uplifts. Subsidence accompanied filling in many parts of the basin. Newport trough continually subsided to accumulate more than a kilometer of sediment in a midbathyal environment. The Santa Monica Mountains rapidly rose to their present elevations. The uplifts in particular were the sites of large faults, folding, and mass movements. The growth of these structures has resulted in the patterns of modern seismicity and the structures played a dominant role in generating many of the present physiographic features.
 The complete reversal of physiographic relief between the region presently offshore and the Los Angeles area is one of the most profound signatures of the local Quaternary tectonism. This change was previously unrecognized, so we are not aware of any published tectonic models that deal with it. Simple fold and thrust models [Namson and Davis, 1991; Shaw and Suppe, 1996] can explain the Pliocene-to-Holocene uplift of the shelves and Palos Verdes Hills, but they fail to explain the simultaneous subsidence of the adjacent offshore basins. Flexural thrust loading is not viable in this case because the Palos Verdes anticlinorium is bounded on the south by a normal fault and constitutes a small load relative to the size and depth of the offshore basins. In addition, the flexural rigidity of the local crust is probably low due to high thermal gradients [e.g., Yeats and Beall, 1991] and the deepest parts of the basins are in their centers, far from any potential load (Figure 25).
 All of the large basins of the coastal belt are essentially deep sags in the crust. Many basin margins are not faulted and monoclinal flexures are the most conspicuous structures that separate basin from uplift, even where large faults are also present at basin edges. Deep flexural basins such as these are typically the result of local crustal thinning and that is also the best way to explain the observed high rates of subsidence.
 The importance of subsidence cannot be overlooked in tectonic discussions of this part of southern California. All of the basins in the coastal belt, including large parts of Los Angeles Basin, have rapidly subsided during all or at least part of the Quaternary. Fold and thrust scenarios invariably result in crustal thickening and uplift and if this area is truly part of a regionally extensive compressional belt it should be generally on the rise. Although LA Basin went from submarine to subaerial during that time, much of the vertical change was due to rapid sedimentation rather than tectonism. Vertical tectonism in the other large basins was all strongly negative resulting in a greater total surface area that was negative than positive throughout region (Figure 25).
 Some extension is evident in the eastern part of the offshore basins, but in all cases the amount of subsidence greatly outweighs the observed magnitude of extension. Uniform-lithospheric-stretching models [e.g., McKenzie, 1978] have been used to describe the subsidence history of LA Basin [Mayer, 1991]. However, that type of model directly relates the amount of subsidence with the magnitude of extension and a corresponding upward perturbation in the Moho due to uniform thinning of the entire mechanical lithosphere. In this case, there is only a small amount of extension and no evidence for a large Moho perturbation has ever been demonstrated. Thus we conclude that the entire lithosphere was not uniformly thinned, only the middle and deep crustal columns beneath the basins. Although speculative, this concept allows us to explain how the basins sank passively into the vacated space independent of the small amount of upper crustal extensional strain. Conversely, the deep crust under local areas of uplift might be getting thicker in a similar manner. Dynamic details of this process are unresolved, but possibly a plate-capture argument [e.g., Bohannon and Geist, 1998; Bohannon and Parsons, 1995; Nicholson et al., 1994] could be invoked to explain how strong remnants of the subducted Farrallon slab with Pacific plate motion might be differentially inducing horizontal flow in the deep and middle crustal rocks from below.
 Coastal-belt uplifts seem to be more readily explained with fold and thrust concepts or as the result of confining bends in large lateral-slip fault systems, since they obviously formed by upper crustal shortening and large strike-slip faults border them in some cases. Legg  has successfully used the latter concept to explain similar large features in the inner continental borderland to the south. However, the structural variability of the coastal belt uplifts raises questions as to how universally any one concept can be applied. The Santa Monica shelf anticlinoria is not the result of a confining bend in a larger strike-slip fault system because it is not fault bounded and is surrounded on three sides by a thick wedge-shaped body of fill in the Santa Monica Basin and shelf areas (Figure 25). Two of the largest local earthquakes indicate that there is a blind thrust deep beneath this wedge of fill [Hauksson and Saldivar, 1989], but there is no evidence that the basin formed as the result of thrusting. The north side of the Palos Verdes Hills uplift might be a confining bend [Fischer et al., 1987a], but to its south, the normal displacement of the same magnitude on the San Pedro Escarpment Fault probably did not form in either that setting or as part of a thrust belt. Even Lasuen Knoll, which might be one of the best examples of a large confining bend, is in many places bounded to the west by a normal-separation fault.
 Another important consideration that bears on Quaternary tectonism in the coastal belt is that it is characterized by simultaneously active faults with different apparent displacements. One potentially flawed fundamental premise that recurs in most tectonic explanations of this region is that there was a complete shift from Miocene transtension to Quaternary transpression. Some explanations implicitly assume three different episodes of extension, strike slip, and compression and although the postulated temporal order of each might vary they are typically considered distinct from the other. Our observations indicate that all three types of deformation have been simultaneously active during the Quaternary over most of the region and that all three have played a significant role in causing the observed regional strain. No single consistent change or sequence of changes can be documented and universally applied over the whole area. There is evidence for normal and reverse faulting in many of our seismic profiles. Young folds appear to be cut by normal faults. We recognize that such disparity can be more apparent than real in 2-D sections that cross strike-slip regimes and that complex strike-slip zones can also include both types of faults as secondary structures. However, in many places the varied types of faults suggest important regional strain variability that is not considered by existing models. We think the latter is especially true near the base of the San Pedro Escarpment where large normal faults that bound the northeast edge of San Pedro Basin interact with reverse faults and folds in the uplift of the Palos Verdes Hills. This is a transitional area between the rapidly subsiding extensional basins to the southwest and the folded uplifts to the northeast.
 The coastal belt is part of a broad right-shear margin accommodating roughly 280 km of Pacific plate displacement at N37°W since 5.1 Ma relative to a fixed North American plate [Atwater and Stock, 1998] and that strain signature can be found throughout the region (Figure 25). However the wide range of structures and deformation styles that occur locally indicate a level of regional complexity that goes well beyond any simple right-shear explanation. Constraining- and releasing-bend geometries can be documented, but not every uplift or blind thrust can be explained that way (Figure 25). The basins in particular seem to have developed as deep crustal sags that are not consistent with simple releasing-bend or fold and thrust explanations. Fundamental, but puzzling questions remain regarding how right shear is partitioned between the Pacific plate offshore and the interior of the North American continent through this zone. If stress is transmitted horizontally through the upper crust, as a right-shear couple might imply, how is it that subsiding basins with extensional strain signatures or no internal strain can flank narrow compressed uplifts? Why would deep crustal sags develop at all in such a system? We conclude instead that the stress has been primarily imparted from below. The kinematic history that we think we see seems to be more consistent with one that is associated with rafted blocks of upper crust differentially linked to some type of strong basal layer in the deep crust or upper mantle. As such, our view is somewhat similar to the flake tectonic models presented by Yeats  and Yeats and Ehlig  in which local crustal areas or blocks are thought to move as independent rafts developing their own internal strain.
 A simple conclusion can be drawn from our analysis regarding seismicity. Deep earthquakes on blind thrusts should only be expected beneath the cores of large anticlinal uplifts. That is where they have been observed to occur, but one might expect them almost anywhere beneath the LA area if published fold and thrust models are correct. In those models the uplifts are linked to major flat faults and thrust ramps many of which lie beneath the adjacent subsiding basins. Because most parts of the basins are not under compression at any depth in our view, we do not expect deep earthquakes beneath them except where they are cut by strike-slip faults. One exception is the northern end of the Newport/Inglewood zone where LA Basin is under compression.
 This paper was reviewed by Holly Ryan, Brian Edward, and Brian Wernicke and has been improved by their comments.