RESULTS OF SCEC-SPONSORED RESEARCH DURING 1997

Kristin D. Weaver and James F. Dolan
Department of Earth Sciences
University of Southern California

-Los Angeles, CA 90089-0740

Raymond fault Project USC graduate student Kristin Weaver and Dolan had a highly productive summer field season on the Raymond fault. We excavated two trenches totaling 42 m in length across what we believe is the second-most-promising trench site on the Raymond fault (more on our efforts to trench the most promising Raymond fault site below) at the Los Angeles County Arboretum in Arcadia, just east of Pasadena (Figure 1). This research formed the first half of Kris Weaver's MS thesis research on Raymond fault. We plan to excavate at least one other site on the Raymond fault during the summer of l 998.

We excavated our 1997 trench along the southern edge of 500-m-long pressure ridge at the southwest end of the arboretum (Figure 2). We chose this site because we thought there might be late Holocene strata onlapping the base of the scarp that might be used to constrain the ages of the most recent few surface ruptures on the fault. The trench exposed well-bedded, predominantly sandy and silty strata that have been deformed by four strands of the fault (Figure 3). Two of these strands were exposed within the trench, whereas the two others are inferred on the basis of stratigraphic and structural relationships in the trench.

The northern strand is inferred to occur just north of the trench on the basis of the monotonic northward thickening of the argillic horizon of the active soil (Figure 3). We interpret this northward thickening as a result of a north-side-up displacement along the inferred northern strand.

The next strand to the south comprises a narrow, well-defined zone of predominantly south-dipping fault splays (Figure 3). This fault zone separates gently south-dipping strata to the south from horizontal strata to the north. We obtained two ~35,000 yr BP AMS radiocarbon dates from detrital charcoal collected near the base of the trench just south of the fault. These ages are consistent with our field estimates of the degree of soil development at the north end of the trench based on the thickness and color of the argillic horizon, the number and degree of development of clay films, and the presence of numerous 'laminar B' horizons within the well-bedded sands in the lower half of the trench. Although the fault dips moderately steeply southward, and exhibits what initially appeared to be south-side-down normal displacements in the upper few meters, the lack of any vertical separation of distinctive stratigraphic units below 2 m depth (Figures 3 and 4) indicates that this is a pure strike-slip fault that has locally juxtaposed dissimilar stratigraphic sequences. In the upper two m the fault separates well-bedded sands capped by a 1.5 to 2 m-thick argillic horizon to the north from several massive, wedge-shaped deposits to the south (Figure 4).

The origin of these wedge-shaped deposits remains unclear. If the stratigraphic relationships exposed lower in the trench did not preclude major south-side-down normal displacements, we would interpret these as colluvial wedges shed off a south-dipping normal fault scarp. But since this strand exhibits pure strike-slip displacement, the wedges must represent either: (l) colluvial wedges shed off a south-facing scarp caused by strike-slip juxtaposition of irregular topography; or (2) channel deposits. We exposed the fault and associated wedge-shaped deposits in three faces: the two original trench walls, and a later exposure we cut 3 m east of the original east wall of the trench. Geometrically, the wedges appear to extend along the strike of the fault, which is, of course, along the slope of the scarp. This relationship suggests that if these are channel deposits, then the channel must have been flowing along the fault. While the presently
active channel at the site does flow along the base of the scarp, the base of the wedgeshaped deposits within the scarp is more than one meter higher than the highest possible correlative units within the recent channel deposits at the south end of the site (discussed below). This relationship, could, of course, be due to uplift of the wedge-shaped deposits during an earthquake along a deeper strand to the south. We suspect, however, that the wedge-shaped deposits are not channel deposits because of the exact coincidence of the northern edge of the wedge-shaped deposits with the upward projection of the fault, and the very abrupt, 60o angle formed by the base of the wedge-shaped deposits and the steep northern edge of the deposit in the west wall of the trench. We believe that the coincidence of the north edge of the deposit with the upward projection of the fault argues against this being a purely sedimentary contact, and that the contact is, in fact, a fault that extends all the way to the ground surface. Nevertheless, we are approaching this interpretation with caution because the three-dimensional geometry of the deposit is not well established (i. e., we do not know if the paleo-surface topography is consistent with the existence of a south-facing paleo-scarp which could have generated colluvial wedges). One additional point that suggests that the wedges may be true colluvial wedges and not channel deposits is that the wedges are completely massive and composed of what appears to be, at least in part, old A horizon material. In contrast, all of the channel deposits that we observed elsewhere in the trench, including those at depths similar to the subsurface depth of the wedges, were pervasively and finely bedded on the mm to cm scale. These relationships, and the preservation of distinct contacts between the wedge-shaped deposits, argues against their massive nature being due to bioturbation. Thus, we think that the wedges probably are indeed colluvial wedges, but we consider this interpretation to be preliminary, and the origin of these features remains somewhat puzzling.

In the southern part of the trench we exposed a third fault.zone. This zone, which we refer to as the 'vertical strand', is 7 m wide and consists of several different splays that cut upward through the south-dipping strata (Figure 5). We obtained a ~9,000 yr BP AMS radiocarbon age on a sample of detrital charcoal collected from the deformed strata near the base of the trench. These faulted, tilted strata are onlapped by horizontal, wellbedded, friable sands, from which we obtained three AMS radiocarbon ages. These three ages were in correct stratigraphic succession, suggesting that they may not be reworked (Figure 5). We obtained a ~ 1,000 yr BP age from one of the deepest exposed horizontal beds. This bed is overlain by a thinly bedded sand layer from which we obtained a ~900 yr BP age. The shallowest, youngest sample is not shown on figure 5, as it came from the opposite trench wall. This ~500 yr BP sample came from a layer that projects across the trench to the top of the layer containing the 900-year BP date.

The tilted strata at the far southern end of the trench suggest the presence of yet another strand located south of the fault. We cannot excavate any farther south at this point as the property on the other side of the boundary fence is occupied by the backyards of several houses, most of which have built-in pools.

The relationships near the southern end of the trench constrain the age of the most recent surface rupture on the Raymond fault to have occurred before ~ 1,000 years ago (the age of the oldest undeformed strata) and after ~9,000 years ago (the maximum age of the deformed strata). We can use our minimum age of the most recent event together with published data from an earlier study of the fault by Crook and others (1987) to place tighter constraints on the most recent event. Crook and others' (1987) trenches near Sunny Slope reservoir 2.5 km west of our trench site revealed a series of narrow fissures filled with black, peaty marsh deposits. These organic-rich, clayey mudstones yielded a bulk radiocarbon age of ~2100 yr BP, providing a maximum age for the most recent surface rupture on the Raymond fault. Thus, the most recent event on the Raymond fault probably occurred between ~ 1,000 and ~2,000 years ago.

This result is markedly different from results obtained from the Hollywood fault, along strike of the Raymond fault 25 km to the west (Figure 1). A series of adjacent bucket-auger holes just west of downtown Hollywood revealed evidence that the most recent surface rupture on the Hollywood fault occurred between ~5,000 and ~8,00010,000 years ago (Dolan and others, manuscript in prep.). These data indicate that the Raymond fault did not rupture together with the Hollywood fault during its most recent surface rupture. This does not, of course, preclude the possibility that earlier ruptures did involve both of these faults. More paleoseismologic data are needed from other faults within the area (e. g.,~Sierra Madre and Verdugo faults) as well as the western part of the Raymond fault west of its intersection with the Verdugo-Eagle Rock fault system, in order to understand the past interactions of ruptures along these faults. Nevertheless, we are encouraged by these early results. Such studies will be critical in understanding the long-distance and long-term interaction of the stress field within southern California.

A final note on the Raymond fault project. In addition to the Arboretum excavations, we attempted to get permission from the City of San Marino to excavate a trench across the Raymond fault in the median strip of Sierra Madre Boulevard. This is the only site on the Raymond fault where the active trace of the fault is buried by a young alluvial fan. We would expect this to be the site of relatively continuous Holocene sedimentary record and we therefore consider this to be the most promising paleoseismologic trench site on the fault. Trenches that were excavated across the Raymond fault at other sites during the early 1980's (see Crook and others, 1987) generally lacked Holocene stratigraphy. As noted above, our own trench site at the Arboretum exhibited a similar problem, with no strata between ~9,000 and 1,000 years old exposed in the trench [The problem at the Arboretum site is logistical. If we had been able to extend the trench farther south beneath the swimming pool of the adjacent property, we probably could have exposed mid Holocene strata]. We are continuing our efforts to gain permission to trench the Sierra Madre Boulevard site. In this regard, Caltech geology professors Brian Wernicke and Joann Stock, both of whom live in San Marino, have agreed to speak on our behalf at an upcoming City Council meeting. I am optimistic that the city can be convinced to allow us permission to trench the site.

Eastern Sierra Madre Fault Project This was a particularly frustrating part of my (Dolan's) efforts this year. Last year I had identified a promising site on the easternmost part of the Sierra Madre fault, just west of San Antonio Canyon, where the frontal reverse-fault system exhibits a major lateral ramp and a northeastward step from the eastern Sierra Madre fault to the Cucamonga fault (Figure 1). I was particularly enthusiastic about this site because I thought it might afford the opportunity to compare the ages of past earthquakes on the eastern Sierra Madre fault with those on the Cucamonga fault. The landowner was initially quite enthusiastic about the project. As the field season approached, however, I was unable to contact him, despite repeated attempts both by mail and telephone. I finally drove over to his house in Claremont and met him as he was leaving his garage. He then informed me that he had decided to sell the property and that his lawyer had advised him to keep me away from the site at all costs. As I said, quite frustrating.

I then began looking for an alternative site along the eastern part of the Sierra Madre fault, with the same research rationale. I have identified what I believe to be a very promising site in San Dimas, and I have contacted the lease-holder (the property is cityowned, fortunately). He was also quite receptive to the idea, as the site lies at the edge of an open field and the proposed trench would have minimal impact on any existing activity. I am currently in the process of contacting the city of San Dimas in order to gain official permission to excavate the site during Summer 1998, using funds already in hand.

Figure 1. Regional neotectonic map for metropolitan southern California showing major active faults. Raymond fault shown in red. Fault locations are from Ziony and Jones (1989), Vedder and others (1986) and Dolan and Sieh (1992). Santa Rosa Island fault is off figure to west. Closed teeth denote reverse-fault surface traces; open teeth show upper edges of blind thrust fault ramps. Strike-slip fault surface traces identified by double arrows. ELATB=East Los Angeles thrust belt; Hol Flt=Hollywood fault; RMF=Red Mountain fault; SCIF=Santa Cruz Island fault; SJF=San Jose fault; SSF=Santa Susana fault; VERF=Verdugo-Eagle Rock fault; LA=Los Angeles; LB=Long Beach; M=Malibu; NB=Newport Beach; Ox=Oxnard; P=Pasadena; PH=Port Hueneme; PM=Point Mugu; PP=Pacific Palisades; SJcF=San Jacinto fault; V=Ventura; WN=Whittier Narrows. Dark shading shows Santa Monica Mountains.

Figure 2. Detail of map of Raymond fault from Crook and others (1987), showing location of our 1997 trench site along the southern edge of a pressure ridge in the LA County Arboretum.

Figure 3. Trench logs of our 1997 LA County Arboretum trench across the Raymond fault. Blue layer at top is the A horizon of the active soil. Plum color beneath it in northern 13 m of trench is a moderately developed argillic horizon. Note onlapping relationship in southern part of trench between tilted and faulted south-dipping strata and undeformed latest Holocene strata.

Figure 4. Detail of log of west wall of trench at the surface projection of the south-dipping fault zone exposed in north-central part of trench. Lack of vertical displacement of the pink layer at 2.5 m depth and underlying green layer precludes sign)ficant dip slip on this strand and suggests that it is a pure strike-slip fault. Origin of the wedge-shaped deposits (purple, green, beige, and pale blue) that overlie a sequence of well-stratified sands (orange and yellow) remains enigmatic. they are either colluvial wedges shed off a south-facing scarp or a series of channel deposits that have been uplifted above the level of present fluvial deposition by slip on a fault somewhere to the south. See text for discussion of possibilities.

Figure 5. Detail of log of east wall of south end of trench. Note onlapping relationship between tilted and faulted early Holocene deposits and flat-lying late Holocene strata. Ages are 14c AMS radiocarbon dates.