In 1996 we began investigating the paleoseismic history of the Cascadia margin based on turbidite records. The possibility that a good earthquake record existed along the margin had been suggested in the late 1960's, and proposed formally in 1990 by John Adams. We conducted our first major cruise to test this hypothesis in 1999 aboard the R/V Melville. We have published several papers since then that expore in detail the realtionship between great earthquakes and turbidite stratigraphy in Cascadia. on that first cruise, we also collected several cores from Noyo Canyon, ~ 90 km south of Cascadia, on the abysal plain seaward of the offshore Northern San Andreas fault. We were unsure about whether these cores could act as controls for the Cascadia work, provide a record of teh NSAF, or perhaps he a confusing stratigraphy. Initially we found a record that was compatible with the existing onshore paleoseismology of the NSAF, and embarked on a project to investigate this record with a cruise on the R/V Revelle in 2002. This work has been supported by the National Science Foundation Earth Science and Ocean Science Divisions, and the USGS NEHRP Program since 1999.
A large number of co-investigators, students, technicians and ships crew have been involved in this project since 2002, including: Chris Goldfinger, C. Hans Nelson Joel E. Johnson, Ann E. Morey, Julia Gutiérrez-Pastor, Eugene Karabanov, Andrew T. Eriksson, Eulàlia Gràcia, Gita Dunhill, Jason Patton, Michaele Kashgarian, John Southon, Pete Kalk, Chris Moser, Bob Wilson, Jeff Beeson, Kelly Grijalva, Roland Bürgmann, and the officers and crews of the R/V Melville and R/V Revelle and the Shipboard Scientific Parties
Research supported by the National Science Foundation (awards EAR 0107120 and 0440427, OCE 0850931, and 1137986) and the U.S. Geological Survey (awards 04HQGR0056, 05HQGR0057, 05HQGR0066. 06HQGR0150, 06HQGR0107, and G09AP00016). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
The NSAF Turbidite Record
The San Andreas Fault (SAF) is probably the most intensively studied transform system in the world. Extending along the west coast of North America, from the Gulf of California to Cape Mendocino, the SAF is the largest component of a complex and wide plate boundary that extends eastward to encompass numerous other strike-slip fault strands and interactions with the Basin and Range extensional province. The Mendocino Triple Junction (MTJ) lies at the termination of the northern San Andreas, and has migrated northward since about 25-28 Ma. As the triple junction moves, the former subduction forearc transitions to right lateral transform motion and the SAF continues to lengthen. A controversial aspect of NSAF tectonics has been whether the fault is seismically segmented, or whether the long 1906 rupture was characteristic, or perhaps a mix of both.
On our 1999 and 2002 cruises, 74 piston, gravity and jumbo Kasten cores were collected from channel and canyon systems draining the northern California continental margin to investigate the record of periodic Holocene turbidites for possible connection to large magnitude earthquakes on the adjacent Northern San Andreas Fault. Poorly known channel systems were mapped with multibeam sonar to define pathways and channel confluences. Cores sampled all major and many minor channel systems extending from Cape Mendocino to just north of Monterey Bay. Sampling both along and across channels was done, and particular attention was paid to channel confluences, as these areas afford opportunities to test for synchronous triggering of turbidity currents.
Figure 1. Core locations from 1999 and 2002 cruises on R/V Melville and Roger Revelle. Bathymetric and topographic data compiled by OSU from NASA land data, archive and newly collected marine survey data during these cruises. Channel systems mapped from the new bathymetric grid and sidescan data. Core numbers are referred to in the text. Noyo Cores (including 49PC from Melville 1999 cruise) and 24GC, discussed in text are also shown in yellow. Offshore provenance from heavy mineral analyses are indicated in boxed text. At the confluence upstream of 24GC, Viscaino and Gualala mineralogies mix and result in stacked pulses, reflecting source mineralogies. Onshore paleoseismic sites Vedanta and Fort Ross are shown in red.
We use 14C ages, relative dating tests at channel confluences, and stratigraphic correlation using physical properties to determine whether turbidites deposited in separate channel systems are correlative, implying they were triggered by a common event. In Cascadia, both we and Adams(1990) used existing cores to test the possibility that the Cascadia cores contained a record of Holocene great earthquakes of the Cascadia margin. Adams examined core
logs for Cascadia Basin cores, and determined that nearly all of them had 13 turbidites overlying the Mazama ash which is dated at 7627 ± 150 cal yr BP (Zdanowicz et al., 1999). He found that cores along the length of Cascadia channel contain 13 turbidites and argued that these 13 turbidites correlate along the channel. Adams observed that cores from Juan de Fuca Canyon, and below the confluence of Willapa, Grays, and Quinault Canyons contain 13-16
turbidites above the Mazama ash. The correlative turbidites in Cascadia channel lie downstream of the confluence of these channels.If these events had been independently triggered events with more than a few hours separation in time, the channels below the confluence should contain the sum of the tributaries, from 26-31 turbidites, not 13 as observed (Figure 1). The importance of this simple observation is that it demonstrates synchronous triggering of turbidite events in tributaries, the headwaters of which are separated by 50-150 km. Similar inferences about regionally triggered synchronous turbidites in separate channels are reported in Pilkey (1988). This elegant relative dating technique and variants of it using isolated lower slope basins are used extensively in our Cascadia, Sumatra, and NSAF work.
The set of NSAF cores have yielded a turbidite record that is in good agreement with the shorter land record of Holocene NSAF earthquakes. Despite the intense scientific study of the NSAF stemming from the great seismic hazards to San Francisco, a reliable event history for this fault extending back through the Holocene has yet to be established. This is particularly true for the segments of the NSAF near and north of San Francisco. Establishment of an offshore record, reconciled with the land paleoseismic record, offers the opportunity to investigate long term earthquake behavior, stress interactions with other fault systems, and enables the use of these parameters in probabilistic hazard models.
Triggering mechanisms: Are they
Are these events all triggered by earthquakes? Common sense suggests that such a scenario is absurdly simplistic, yet our Cascadia work has led us to the unlikely conclusion that Adams(1990) was correct. We now discuss the methods used to test the hypothesis and why it seems to work. Adams(1990)suggested four plausible mechanisms for turbid flow triggering: 1) storm wave loading; 2) great earthquakes; 3) tsunamis; and 4) sediment loading. To these we add 5) crustal earthquakes, 6) slab earthquakes, 7) hyperpycnal flow, and 8) gas hydrate destabilization.
All of these mechanisms could trigger a
turbid event, but how often do they actually occur, and can
earthquake-triggered events be distinguished from other events?
Two basic techniques can be used to
distinguish seismic from non seismic events:
Sedimentological determination of individual event origin.
Regional correlations that require synchronous triggering.
Individual event determination can in some cases distinguish seismic turbidites from storm, tsunami, and other deposits using several methods. Nakajima and Kanai (2000) and Shiki et al. (2000) report that seismo-turbidites can in some cases be distinguished sedimentologically. They observe that historically known seismically derived turbidites in the Japan Sea and Lake Biwa are distinguished by wide areal extent, multiple coarse fraction pulses, variable provenance, and
greater depositional volume than storm-generated events. These investigators traced known seismo-turbidites to multiple slump events in many parts of a canyon system, generating multiple pulses in an amalgamated turbidity current, some of which sampled different lithologies that are separable in the turbidite deposit. These turbidites are complex, with reverse grading, cutouts, and multiple pulses. Gorsline et al. (2000) make similar observations regarding areal extent and volume of seismoturbidites. In general, these investigators observe that known storm triggered events are thinner, finer grained and have simple normally graded Bouma sequences, although complexity is also a function of proximity to the source, and some reports reach different conclusions (Mulder and Syvitski, 1996). While there may yet be applicable global, regional or local criteria to make such distinctions, these are at present poorly developed and somewhat contradictory.
Thus far in Cascadia and the San Andreas, we have for the most part not attempted to distinguish between triggering mechanisms directly because the physiography, numerous tephra layers, and long historical records favorable to this method in Japan are not present on the west coast of North America. Favorable factors that are present favor regional correlation and determination of synchronous triggering. Determination of synchronous triggering can eliminate non-earthquake triggers with the possible exception of storm wave loading or multiple hyperpycnal flows for very large storms, and perhaps triggering by a tele-tsunami. West Coast physiography favors filtering of non-seismic events from the record because a wide shelf separates river sources from canyon heads. Hyperpycnal flow, or direct turbid injection from rivers, can produce turbid flows, and can even mimic earthquakes in that they may affect several rivers over a span of days. We have found that while this certainly occurred during the Pleistocene when lowered sea-level resulted in direct river-canyon connections resulting in a major increase in turbidite frequency at the Holocene Pleistocene boundary. Once sea level-rise in the Holocene isolated most west coast canyons from the rivers, sediment was spread across the shelf and not directly injected in the canyons (e.g. Sternberg, 1986). Some exceptions are the Eel river record, which probably contains storm events, and the Viscaino channel along the northern San Andreas.>Both of these occur where the shelf is very narrow, and river injection is possible. Deep canyon heads also prevent triggering by storm wave loading and distant tsunami, the last two non-earthquake sources. For example, storm wave loading is an unlikely trigger in Cascadia, where, although deep water storm waves are large, the canyon heads where sediment accumulation occurs are at water depths of 150-400 m, too deep for sustained disturbance by maximal storm waves of ~20 meters. Tsunamis may also conceivably act as a regional trigger, however the tsunami from the 1964 Alaska Mw 9.0 event did not trigger a turbidite observed in any of the cores in Cascadia or the NSAF, although it did serious damage along the Pacific coast (Adams, 1990). Crustal or slab earthquakes could also trigger slumps and turbid flows, though not regionally. To test for this, we resampled the location of a 1986 box core in Mendocino channel, where the uppermost event is suspected to be the 1906 San Andreas event. Since 1986, the Mw7.1 Petrolia earthquake occurred in 1992, with an epicentral distance of only a few km from the canyon head. We found no turbidite in the 1999 box core, suggesting that triggering at that site may require earthquakes larger than Mw 7.2. Conversely, the Loma Prieta earthquake apparently did trigger a turbid flow event in Monterey Canyon at a greater epicentral distance (Garfield et al., 1994), though it is not known whether a discernable turbidite record exists from this event. Japanese investigators have suggested a minimum magnitude of ~Mw=7.2 for turbidite triggering, though we suspect that this minimum value is site and event specific.
Comparisons of hyperpycnal and earthquake generated turbidites
Taking advantage of favorable physiography, we have used spatial and temporal patterns of event correlations that are unlikely be the result of triggers other than earthquakes. We use multiple techniques to test for linkage between sites and thus test for synchroneity. Typically, paleoseismologic investigations use radiocarbon constraints to establish these linkages, but often are unable to determine synchronous event chronology due to the inherent limits in dating techniques. Relative dating techniques, if available, and if of sufficient resolution, are strongly preferred to test for synchroneity. The “confluence test” of Adams (1990) is powerful in that it requires synchronous triggering within a few hours. Comparisons of the number of events between time markers is a somewhat less powerful technique that can be applied in some cases. Recently we have begun to use direct physical property correlations, which are proving to be a powerful new method of testing for linkages (if present) between sites.
Using the “confluence test”, of which there are six along the NSAF margin. Subsequent to these results, we have found it possible to correlate the Holocene events directly using physical properties of the turbidites themselves in the cores. This method is discussed more fully below. Using these correlations we believe it is possible to establish synchronous triggering for much of the NSAF margin between San Francisco and Cape Mendocino. Along the NSAF, event synchroneity is tested not only with radiocarbon ages, but with correlation techniques within a radiocarbon constrained framework. Correlations using density and magnetic susceptibility patterns, which in this region are reasonable proxies for grain size, have been able to link sites along strike, a connection that radiocarbon evidence alone cannot establish with either onshore or offshore records.
We have tested the turbidite record along the NSAF for synchronous triggering of turbidity currents as a method for determining the origin of these deposits, whether from earthquake, or other sources. Using 14C ages, relative dating tests at channel confluences, and direct correlation of physical properties to determine whether turbidites deposited in separate channel systems are correlative, that is, they were triggered by a common event. We concluded that along-strike physical property correlations, supported by application of the “confluence test” and good correspondence with land paleoseismic dates both in individual matching, and total number of events, supports the use of offshore turbidites as paleoseismic indicators for the NSAF. We suggest several other lines of evidence that support this conclusion. First, the recurrence interval overall is consistent with other paleoseismic evidence from the North Coast segment of the fault. Second, the mean recurrence interval changes abruptly (more than doubling) at the MTJ, from a value consistent with NSAF earthquakes south of the MTJ, to a value consistent with Cascadia earthquakes north of the MTJ. We can think of no other reason for such a rate change, and indeed if such a change were due to external factors such as storm frequency or sediment loads, the frequency should increase northward, not southward as observed. Third, we observe that the thickness of the turbidites decreases southward as the shelf widens and the distance from the canyon heads to the fault increases. This occurs despite closer proximity to the main sediment sources, the Russian River and San Francisco Bay, which are to the south. This relationship suggests that epicentral distance, and not sediment supply, is the controlling factor for turbidite size, at least during Holocene high sea-levels. This observation also supports our primary inference of earthquake origin, and is consistent with observations in Cascadia where robust turbidites are found in channels fed by small sediment sources, such as the Mendocino Channel fed by the Mattole River.
Figure 2. “Confluence Test” as applied to the central NSAF margin channel systems for the period ~4000-10,000 ybp. Inset map shows eight confluences (southern two dashed, not yet tested). The correlation figure uses magnetic susceptibility records from six cores at four sites above confluence “C” (representative lithologic log for 13PC shown, 13PC gamma density shown in light blue). Magnetic susceptibility for three cores below confluence “C”, and one core below confluences “B”, “C”, “D”, and “E” are also shown. Green bars separate these three core sets, and are linked to the map. Core 25GC (red) also shown alongside 49PC with hemipelagic intervals (circled H) shown between turbidites. Magnetic trace for 49PC shows disturbance by sampling, density trace is shown for comparison. Mismatching hemipelagic intervals shown green, all others in red. Green wavy lines are known erosive bases. Turbidite numbers shown on 49PC in green. Peak radiocarbon ages are shown without ranges to reduce clutter. Color bands show correlation ties for key turbidites. While some details of each of the correlated turbidites vary from site to site as would be expected, the stratigraphy represented by these turbidites remains largely unaffected by passage through the confluences with other channels. The total number of turbidites is nearly constant, indicating that turbidite arrival at confluences from separate channel systems was most likely synchronous. See figure 6 for mag. sus. and density scales.
The NSAF late Holocene turbidite record examined thus far has passed these tests, and can be correlated with multiple proxies along multiple canyon systems from the MTJ to offshore San Francisco. The inference of earthquake origin is further supported by an abrupt change in turbidite recurrence interval at the Mendocino Triple Junction, from ~200 years, a value consistent with onshore NSAF earthquakes to the south, to a value of ~520 years consistent with the Cascadia subduction zone to the north.
Figure 3. In progress Holocene correlation diagram for NSAF system cores. Light blue traces are gamma density, dark blue traces are magnetic susceptibility. (Kasten cores “KC” do not have associated density data as they are too large for the scanner). This correlation is approximately oriented in a 14C age framework, with some vertical stretching of cores based on differing sedimentation rates. Peak probability calendar ages corrected for sample thickness are shown in black (gray if questionable) below the dated turbidite, 2s range omitted for clarity. Upper ~2500 years of data shown in next figure using 3 mm high-resolution magnetic susceptibility data. Significant correlation horizons, are shown, but not all individual events are depicted to reduce clutter. Color scheme matches events in Figures 5 and 7 for events common to both figures.
Figure 4. Correlation diagram for the uppermost ~ 3000 years in NSAF system cores. Channel locations shown in Figure 1. Light blue traces are Gamma density, dark blue are magnetic susceptibility. 1 cm resolution magnetic susceptibility data are shown for Noyo cores, and 3 mm high-resolution data are shown for all other cores. The 3mm magnetic susceptibility data is stretched horizontally (variable colored traces) to show the low amplitude signal resolved with this resolution. Event numbers in blue at core centers. NSAF cores all have a number of relatively thick events that serve as excellent datums, and are the basis for our regional correlations in Figure 6. The uppermost of these marker events, a very distinctive multipulse event dated at ~ 2100 ybp, is shown labeled “11” in red. Gualala, Albion, Bodega and Cordell and 54KC core tops have been vertically expanded for viewing to offset low sedimentation rates in the upper 2100 years for all cores. OxCal age peaks and ranges shown on cores 49PC and 54KC at left; these ages do not match figure 6 ages, for which data are insufficient for OxCal analysis. Gualala ages have been corrected for sample thickness only. Correlation for weak events T7a and T9a uncertain and not shown beyond Noyo Channel. T7a is datable, and included in our time series; T9a appears to be limited to two adjacent proximal core sites, and is not included in the recurrence statistics. 49PC high-resolution magnetic trace is punctuated by sample voids, indicated by yellow squares. Color scheme matches events in Figures 5 and 6 for events common to both figures.
Preliminary comparisons of our event ages with existing and in progress work at onshore coastal sites show good correspondence, further circumstantial evidence that the offshore record is primarily earthquake generated. During the last ~2100 years, we observe 11 most likely correlative turbidites, including one likely generated by the 1906 earthquake, that can be traced between Noyo Canyon, near the MTJ, and Cordell Channel near Point Reyes. Using combined constraints from physical property correlation, radiocarbon ages, and inter-event sedimentation, we conclude that it is likely that at least 8 of 11 events recorded both onshore and offshore in the past 2100 years have rupture lengths of at least 250 km, and extend from the MTJ region to near the latitude of San Francisco.
Our data, and those from Vedanta and Fort Ross, suggest an age near AD 1700-1720 for the NSAF penultimate event. While Vedanta and Fort Ross are sufficiently distant from Cascadia to preclude confusion, the Noyo Canyon offshore site is much closer at 90 km. The 14C ages cannot distinguish the penultimate NSAF from the Cascadia AD 1700 event. However this event, as with most of the other 10 events discussed here, can be correlated well to the south, making it unlikely that the Cascadia record is confused in the NSAF record.
Figure 5. OxCal age model for the youngest 15 events in the NSAF offshore system, and comparison to onshore ages. Inter-event times based on hemipelagic sediment thickness (represented by gray bars) were used to constrain original 14C calendar age distributions (gray traces) using the SEQUENCE option in OxCal. Inter-event times were estimated by converting hemipelagic sediment thickness between each pair of events to time using the sedimentation rate. Events dated more than once were combined in OxCal prior to calibration if results were in agreement; if not in agreement, the younger radiocarbon age was used in the final model. Five ages are calculated from sedimentation rates where not enough forams were present for 14C dating. The resulting probability distributions (filled black, grey for undated events) are mostly in good agreement with land ages from Fort Ross except for T3-4 and T7a (green lines) and Vedanta (red lines). See inset for geographic locations. See Electronic Supplement for OxCal input data and sedimentation rate curves.
Stress Linkage to Cascadia?
We have also compared the Cascadia and NSAF paleoseismic records, interested in whther any relationships could be observed between the two records. Onshore and offshore paleoseismic records from the Cascadia subduction zone suggest that margin wide and segmented southern Cascadia earthquakes precede NSAF events by ~ 0-80 years, averaging 25-45 years for 11 of 15 probable earthquakes or 73% of NSAF events in the past 3000 years, with two additional Cascadia events too poorly constrained for comparison. On the other hand, NSAF events precede Cascadia earthquakes on average by ~ 150-200 years, most likely too long for a stress change effect. Modeling of the static coseismic and both viscoelastic and afterslip-induced postseimic stress changes suggests that co-seismic stress changes from Cascadia earthquakes are more than sufficient to trigger NSAF events, if they nucleate along the northernmost section of the NSAF near Point Delgada.
Figure 6. OxCal age model for the youngest 15 events in the NSAF offshore system, and comparison to onshore NSAF ages. Cascadia OxCal PDF’s are shown in blue, with lighter blue used where only Hemipelagic ages are available. Land ages from OxCal combines are shown in red. Cascadia mean event ages are also shown with blue arrows for well dated turbidite events, Purple arrows for hemipelagic age estimates, and light red arrows for onshore paleoseismic events. See text for discussion and tables for data used and criteria, and discussion of temporal relationships. Inter-event times based on hemipelagic sediment thickness (represented by gray segments of NSAF PDF’s) were used to constrain original 14C calendar age distributions (gray traces) using the SEQUENCE option in OxCal. Inter-event times were estimated by converting hemipelagic sediment thickness between each pair of events to time using the sedimentation rate. Events dated more than once were combined in OxCal prior to calibration if results were in agreement; if not in agreement, the younger radiocarbon age was used in the final model. Five ages are calculated from sedimentation rates where not enough forams were present for 14C dating. The resulting probability distributions (filled black, grey for undated events) are mostly in good agreement with land ages from Fort Ross except for T3-4 and T7a (green lines; Kelson et al., 2006) Vedanta (red lines; Zhang et al., 2006) Bolinas Lagoon and Bodega Bay (Purple lines, Knudsen et al., 2002) and Point Arena (light blue lines, Prentice et al., 2000). Additional Vedanta event is also shown (Fumal pers. comm. 2007) See inset for geographic locations. See Electronic Supplement for OxCal input data and sedimentation rate curves.
Figure 7. a) Comparison of CFS changes (bars) on the NSAF from A) coseismic deformation B) afterslip C) viscoelastic relaxation D) combined coseismic, afterslip, and viscoelastic relaxation E) coseismic deformation and viscoelastic relaxation from Mendocino Transform Fault (MTF) and Little Salmon Fault (LSF). Fault segments in black are source faults, green segments signify receiver faults. F) CFS change on the Cascadia receiver faults from a NSAF earthquake.
Alternative Explanations? Hyperpycnal flow
Hyperpycnal flow is the density driven underflow from storm discharge of rivers into marine or lacustrine systems, and has been proposed as a link to turbidity currents in a variety of settings. Documentation of hyperpycnal flows into lakes and shelf basins is abundant, however evidence of such flows entering canyons systems and moving into deep water is sparse. Most, if not all examples involve very short distances between the river mouth and canyon head, either during Pleistocene low-stand conditions, or in systems that have very narrow shelves during high stand conditions. Hyperpycnal flows extend further from river mouths with high discharge (Alexander and Mulder, 2002), but documentation is sparse. Wright et al. (2001) infer that hyperpycnal flow is strongly affected by ambient currents, and generally deliver sediment to the slope only upon relaxation of longshore currents. Most investigators cite Pleistocene examples when referring to flows reaching the abyssal plain or lower fan reaches (e.g. Mulder et al., 2003; Normark et al., 1998; Normark and Reid, 2002; Piper et al., 1999). Under low stand conditions, rivers and canyons are directly connected, and such flows are expected. These authors relate the deep water deposition of hyperpycnites to sea level control, or alternatively to climate shifts. An example of high stand hyperpycnal flow has been reported for the Var River, in which the canyon and river mouth are less than 1 km apart (Klaucke et al., 2000; Mulder et al., 1998a). Many large river systems deposit most of their load in river mouth bars, with lesser quantities making it past such bars in to a delta front slope (eg. Yellow River, Li et al., 1998). Whether hyperpycnal flows can reach the deep sea via canyon systems incised during the Pleistocene appears to be a function of shelf width, river peak storm discharge, Holocene aggradation of Pleistocene canyons, and the wave and current climate during peak storm discharge. However, the requirements for and evidence of hyperpycnal flows to the deep ocean under high-stand conditions (excepting very narrow shelves) remain unknown at best (Mulder et al., 2001).
Comparisons of hyperpycnal and earthquake generated turbidites
In some proximal settings such as large lakes, shelf basins, and fjords, records of both earthquakes and storm deposits are found. In one of the best comparisons, St.-Onge et al (2004) show that details of both seismic and hyperpycnal deposition in the Saguenay Fjord in eastern Canada are diagnostic, and argue that hyperpycnal deposits are distinguished by reverse grading at the base, followed by normal grading. The diagnostic reverse-then-normal grading for hyperpycnal deposits has been widely reported, and is the result of waxing, then waning flow associated with the storm. In the Saguenay Fjord, six events have normal grading, and are inferred to be earthquake generated. Four others have similar basal units, but are topped by a reverse graded unit, and then a normally graded unit, with no evidence of hemipelagic sediment between these units. These events are interpreted as an earthquake, followed by a hyperpycnite that resulted from the breaching of a landslide dam caused by the original earthquake. The dam breaching is a variant of the more common hyperpycnal scenario involving waxing and waning depletive flow (Kneller, 1995), but would likely produce a similar flow hydrograph (St-Onge et al., 2004).
In another well studied example, a major river flood in the1969 El Nino input ~25 million tons of sediment (20X the yearly Columbia sediment load) to the Santa Ana River in Southern California over a 24 hour period, in close proximity to nearby canyon heads. Sediment from this extreme flood did not continue down canyons as hyperpycnal flow, but deposited as a distinct yellow unit on the shelf and upper slope. Over the next 10 years it moved downslope as turbid layer transport caused by storm wave resuspension, similar to that observed by Puig et al. (2004) for the Eel system, and deposited as yellow layers between varves of the Santa Barbara Basin (Drake et al., 1972).
Other examples of reverse-then-normal grading for hyperpycnites have been reported and compared to normally graded failure deposits in the Var system (Mulder et al., 2001), Lake Biwa (Shiki et al, 2000a), and modeled by Mulder et al. (1998b), Felix (2002) and others. The dynamics of longitudinal and temporal variability and their effects have been discussed in detail by Kneller and McCaffrey, 2003, and Mulder et al., 2003. Hyperpycnites are also commonly very organic rich as compared to seismic turbidites, having their sources in floods rather than in resuspension of older canyon wall material as in earthquake triggering (Mulder et al., 2001; Shiki et al., 2000; Nakajima et al, 2000). It has been suggested that this distinction may be used as a basis for distinguishing earthquake and storm deposits using OSL dating.
On the northern California margin, all of the canyon systems have Holocene shelf widths of 10-50 km, with the shelf widening southward. One canyon, Viscaino Canyon traverses the shelf and comes within ~ 1 km of the shoreline, similar to the configuration of the Monterey Canyon. Viscaino Canyon and channel were found to contain heavy loads of what are likely to be littoral sand captured during longshore transport during the winter season. On the abyssal plain, Viscaino channel did not have this heavy sand load, and was usable for paleoseismic work. In the northernmost California and thePacific Northwest, large storms both increase discharge, and produce strong southerly winds that disperse sediment northward on the shelf via the Davidson current (Sternberg, 1986; Wheatcroft et al., 1997; Sommerfield and Nittrouer, 1999; Wolf et al., 1999) as well as hundreds of kilometers seaward as evidenced by satellite imagery (Wheatcroft et al., 1997). The Eel River system, with its narrow shelf is the best candidate in Cascadia for delivering hyperpycnal flows to the canyon head. However, Puig et al. (2004) saw no evidence of hyperpycnal flow into the Eel Canyon head in observations made during storms in the Canyon head at 120 m water depth. They did observe some cross-shelf sediment transport due to wave loading at this depth. Other observations made for depths of 280 m, and 900 m were also linked to storms, with wave loading the presumed mechanism, but no link to hyperpycnal discharge or to turbidity currents delivering material to the base of the slope and beyond was found (Puig et al., 2003, 2004). In all NSAF canyon systems, the late Pleistocene transgression and subsequent aggradation have erased any topographic canyon or channel expression across the shelf.
NSAF turbidite characteristics
NSAF turbidites show rare to no evidence of the characteristic reverse grading and high organic content observed for hyperpycnal deposits. We observe normally graded sequences with multiple amalgamated pulses, sharp erosional bases, and normal grading of individual pulses when present. The number and form of these multiple pulses persists over wide areas, and are the basis for individual event correlation. In further testing for event origin in our cores, we find no evidence of the1964 and 1995 extreme El Nino river flood events as hyperpycnal turbidite deposits in any core.
Storm events that may trigger hyperpycnal flow occur relatively frequently in settings where they have been observed. Such events occur every few years in Lake Baikal and Crater Lake yet there are only 4-6 Holocene turbidites in Lake Baikal and a few per thousand years in Crater Lake, far less than the hundreds Holocene hyperpycnal flow events. Similarly, the number of storm events along the northern NSAF and southern Cascadia that exceed 15m significant wave height and 100 knots windspeet is 2-3 per year, or 600-900 such storms since the AD 1700 Cascadia earthquake, and 200-300 since the 1906 earthquake. We see no evidence of any turbidites above the 1906 NSAF or 1700 Cascadia events except in the southermost part of Cascadia, where there are both proximal river sources, and more seismic sources. The three turbidites that overly the likely 1700 AD event there are fewer thatn the number of storms, but are as yet unexplained. We conclude that hyperpycnal flow is an important mechanism for delivery of sediments to the shelf or upper slope, but other triggering mechanisms (sediment overloading failures, storm wave loading, or earthquakes) must be responsible for the observed frequency of robust turbidites in lacustrine or marine turbidite systems.
Goldfinger, C., Nelson, C.H., and Johnson, J.E., 2003, Deep-Water Turbidites as Holocene Earthquake Proxies: The Cascadia Subduction Zone and Northern San Andreas Fault Systems: Annali Geofisica, v. 46, p. 1169-1194.
—, 2003, Holocene Earthquake Records From the Cascadia Subduction Zone and Northern San Andreas Fault Based on Precise Dating of Offshore Turbidites: Annual Reviews of Earth and Planetary Sciences, v. 31, p. 555-577.
Goldfinger, C., Morey, A.E., Nelson, C.H., Gutiérrez-Pastor, J., Johnson, J.E., Karabanov, E., Chaytor, J., Ericsson, A., and shipboard scientific party, 2007, Rupture lengths and temporal history of significant earthquakes on the Offshore and Northcoast segments of the Northern San Andreas Fault based on turbidite stratigraphy, Earth and Planetary Science Letters, v. 254, p. 9-27.
Goldfinger, C., Grijalva, K., Burgmann, R., Morey, A.E., Johnson, J.E., Nelson, C.H., Gutierrez-Pastor, J., Karabanov, E., Chaytor, J.D., Patton, J., and Gracia, E., 2008, Late Holocene Rupture of the Northern San Andreas Fault and Possible Stress Linkage to the Cascadia Subduction Zone, Bulletin of the Seismological Society of America, v. 98, p. 861-889.
Goldfinger, C., Patton, J.R., Morey, A.M., 2009, Reply to comment on "Late Holocene Rupture of the Northern San Andreas Fault and Possible Stress Linkage to the Cascadia Subduction Zone, Goldfinger, C., Grijalva, K., Burgmann, R., Morey, A., Johnson, J.E., Nelson, C.H., Ericsson, A., Gutiérrez-Pastor, J., Patton, J., Karabanov, E., Gracia, E.", Bulletin of the Seismological Society of America, v. 98, p. 861-889, 8 pp.
Goldfinger, C., 2009, Subaqueous Paleoseismology, in Mcalpin, J., ed., Paleoseismology, 2nd edition, Elsevier, p. 119-169.
Goldfinger, C., 2011, Submarine Paleoseismology Based on Turbidite Records, Annual Reviews of Marine Science, v. 3, p. 35-66.
Goldfinger, C., Nelson, C.H., Morey, A., Johnson, J.E., Gutierrez-Pastor, J., Eriksson, A.T., Karabanov, E., Patton, J., Gracia, E., Enkin, R., Dallimore, A., Dunhill, G., and Vallier, T., 2011, Turbidite Event History: Methods and Implications for Holocene Paleoseismicity of the Cascadia Subduction Zone, USGS Professional Paper 1661-F, Reston, VA, U.S. Geological Survey, p. 362 p, 64 Figures. In press. Unformatted preprint avaialble at http://activetectonics.coas.oregonstate.edu/Goldfinger_1661.zip
Gracia, E., Vizcaino, A., Escutia, C., Asioli, A., Rodes, A., Palla, R., Garcia-Orellana, J., Lebreiro, S., Goldfinger, C., 2010, Holocene Earthquake Record Offshore Portugal (SW Iberia): Testing turbidite paleoseismology in a slow-convergence margin, Quaternary Science Reviews, v. 29, p. 1156-1172.
Gutierrez-Pastor, J., Nelson, C. H., Goldfinger, C., Johnson, J.E., Escutia, C., Eriksson, A., Morey, A. and The Shipboard Scientific Party, 2009, In Press, Earthquake Control of Holocene Turbidite Frequency Confirmed by Hemipelagic Sedimentation Chronology on the Cascadia and Northern California Active Continental Margins GS/SEPM, Special Publication “External Controls on Deep-Water Depositional Systems”, Editors: Ben Kneller, Bill McCaffrey and Ole J. Martinsen, 30p.
Nelson, C.H., Goldfinger, C., Johnson, J.E., and Dunhill, G., 2000, Variation of modern turbidite systems along the subduction zone margin of Cascadia Basin and implications for turbidite reservoir beds, in Weimer, P.W., and al., e., eds., Deep-water Reservoirs of the World: 20th Annual Research Conference, Gulf Coast Section Society of Economic Paleontologists and Mineralogists, p. 31 pp.
Nelson, C. H. Escutia, C., Goldfinger, C., Karabanov, E., Gutiérrez-Pastor, J., 2009, In Press, Tectonic, volcanic, sedimentary, climatic, sea level, oceanographic, and anthropogenic controls on turbidite systems SEPM Special Paper on turbidites, in prep.
Priest, G. R., Goldfinger, C., Wang, W., Witter, R. C. Zhang, Y., 2009, Confidence levels for tsunami-inundation limits in northern Oregon inferred from a 10,000-year history of great earthquakes at the Cascadia subduction zone, Natural Hazards v. 54, p. 27-73. DOI 10.1007/s11069-009-9453-5.
Priest G.R, Goldfinger C, Wang K, Witter RC, Zhang Y, Baptista A.M., 2009, Tsunami hazard assessment of the Northern Oregon coast: a multi-deterministic approach tested at Cannon Beach, Clatsop County, Oregon. Oregon Department of Geology Mineral Industries Special Paper 41, 89p. plus Appendix and GIS files.
Witter, R.C., Zhang, Y., Wang, K., Priest, G.R., Goldfinger, C., Stimely, L., English, J.T., and Ferro, P.A., 2011, Simulating tsunami inundation at Bandon, Coos County, Oregon, using hypothetical Cascadia and Alaska earthquake scenarios, Oregon Department of Geology and Mineral Industries Special Paper 43, Salem, Oregon, 63 p. and GIS files
Alexander, J., and Mulder, T., 2002, Experimental quasi-steady density currents: Marine Geology, v. 186, no. 3-4, p. 195-210.
Adams, J., 1990, Paleoseismicity of the Cascadia subduction zone: Evidence from turbidites off the Oregon-Washington Margin: Tectonics, v. 9, p. 569-584.
Beattie, P.D., and Dade, W.B., 1996, Is scaling in turbidite deposition consistent with forcing by earthquakes?: Journal of Sedimentary Research, v. 66, p. 909-915.
Beck, C., Mercier de Lépinay, B., Schneider, J.-L., Cremer, M., Çagatay, N., Wendenbaum, E., Boutareaud, S., Ménot, G., Schmidt, S., Weber, O., Eris, K., Armijo, R., Meyer, B., Pondard, N., Gutscher, M.-A., Turon, J.L., Labeyrie, L., Cortijo, E., Gallet, Y., Bouquerel, H., Gorur, N., Gervais, A., Castera, M.H., Londeix, L., de Rességuier, A., and Jaouen, A., 2007, Late Quaternary co-seismic sedimentation in the Sea of Marmara's deep basins: Sedimentary Geology, v. 199, p. 65-89.
Bertrand, S., Charlet, F., Chapron, E., Fagel, N., and De Batist, M., 2008, Reconstruction of the Holocene seismotectonic activity of the Southern Andes from seismites recorded in Lago Icalma, Chile, 39°S: Palaeogeography, Palaeoclimatology, Palaeoecology, v. 259, p. 301-322.
Blumberg, S., Lamy, F., Arz, H.W., Echtler, H.P., Wiedicke, M., Haug, G.H., and Oncken, O., 2008, Turbiditic trench deposits at the South-Chilean active margin: A Pleistocene–Holocene record of climate and tectonics: Earth and Planetary Science Letters, v. 268, p. 52-539.
Dallimore, A., Thomson, R.E., and Bertram, M.A., 2005, Modern to late Holocene deposition in an anoxic fjord on the west coast of Canada: Implications for regional oceanography, climate and paleoseismic history: Marine Geology, v. 219, p. 47-60.
Doig, R., 1990, 2300 yr history of seismicity from silting events in Lake Tadoussac, Charlevoix, Quebec: Geology, v. 18, p. 820-823.
Goldfinger, C., Patton, J., Morey, A.E., and Nelson, C.H., 2009, Reply to “Comment on Late Holocene Rupture of the Northern San Andreas Fault and Possible Stress Linkage to the Cascadia Subduction Zone" Bulletin of the Seismological Society of America, v. 99, p. 2599–2606.
Goldfinger, C., Patton, J.R., Morey, A., and Witter, R., 2010, Segmentation and Probabilities for Cascadia Great Earthquakes based on Onshore and Offshore Paleoseismic Data, Seismological Society of America Annual Meeting: Portland, OR, Seismological Research Letters.
Goldfinger, C., Witter, R., Priest, G.R., Wang, K., and Zhang, Y., 2010, Cascadia Supercycles: Energy Management of the long Cascadia Earthquake Series: Seismological Society of America Annual Meeting, v. 81, p. 290.
Gorsline, D.S., De Diego, T., and Nava-Sanchez, E.H., 2000, Seismically triggered turbidites in small margin basins: Alfonso Basin, Western Gulf of California and Santa Monica Basin, California Borderland: Sedimentary Geology, v. 135, p. 21-35.
Gorsline, D.S., Shiki, T., Chough, S.K., and Einsele, G., 1996, Depositional events in Santa Monica Basin, California borderland, over the past five centuries, Elsevier, Amsterdam, Netherlands, p. 73-88.
Guyard, H., St-Onge, G., Chapron, E., Anselmetti, F., and Francus, P., 2007, The Ad 1881 Earthquake-Triggered Slump And Late Holocene Flood-Induced Turbidites From Proglacial Lake Bramant, Western French Alps, Submarine Mass Movements and Their Consequences, p. 279-286.
Huh, C.A., Su, C.C., Liang, W.T., and Ling, C.Y., 2004, Linkages between turbidites in the southern Okinawa Trough and submarine earthquakes: Geophysical Research Letters, v. 31.
Huh, C.-A., Su, C.-C., Wang, C.-H., Lee, S.-Y., and Lin, I.-T., 2006, Sedimentation in the Southern Okinawa Trough -- Rates, turbidites and a sediment budget: Marine Geology, v. 231, p. 129-139.
Ikehara, K., 2000, Deep-sea turbidite record of great earthquake along the eastern Nankai Trough and eastern margin of Japan Sea, Active Fault Research for the New Millenium. Proceedings of the Hokudan International Symposium and School on Active Faulting: Hokudan-cho, Awaji Island, Hyogo, Japan, p. p. 119-120.
—, 2000, Paleoseismicity analysis using earthquake-induced sediments: Bull Geol. Survey Japan, v. 51, p. 89-102.
—, 2004, Estimation of recurrence intervals of large earthquakes using deep-sea turbidites around Japan, in Satake, K., Goldfinger, C., ed., Workshop on Turbidites as Earthquake Recorders: Tsukuba, Japan, AIST, Geological Survey of Japan.
Ikehara, K., and Ashi, J., 2004, Deep-Sea Turbidite Evidences on the Recurrence of Large Earthquakes Along the Eastern Nankai Trough: Eos Trans. AGU, v. 86, p. Fall Meeting Supplement, Abstract T11A-0354.
Inouchi, Y., Kinugasa, Y., Kumon, F., Nakano, S., Yasumatsu, S., and Shiki, T., 1996, Turbidites as records of intense palaeoearthquakes in Lake Biwa, Japan: Sed. Geol., v. v. 104, p. p. 117-125.
Johnson, J.E., Paull, C.K., Normark, W., and Ussler, W., 2005, Late Holocene Turbidity Currents in Monterey Canyon and Fan Channel: Implications for Interpreting Active Margin Turbidite Records: Eos Trans. AGU, v. 86, p. abstract OS21A-1521.
Karlin, R.E., and Abella, S.E.B., 1994, A history of past earthquakes in the Puget Sound area recorded in Holocene sediments from Lake Washington: Proceedings of the workshop on Paleoseismology Open-File Report - U. S. Geological, v. Survey, p. 90.
—, 1996, A history of Pacific Northwest earthquakes recorded in Holocene sediments from Lake Washington: Journal of Geophysical Research, B, Solid Earth and Planets, v. 101, p. 6137-6150.
Karlin, R.E., Holmes, M., Abella, S.E.B., and Sylwester, R., 2004, Holocene landslides and a 3500-year record of Pacific Northwest earthquakes from sediments in Lake Washington: Geological Society of America Bulletin, v. 116, p. 94-108.
Karlin, R.E., Verosub, K., and Harris, A., 1996, A paleoearthquake record in varved Holocene sediments from ODP Leg 169S, Saanich Inlet, British Columbia: Geological Society of America, Cordilleran Section and associated societies, 96th annual meeting; abstracts Abstracts with Programs - Geological Society of America, v. 32, p. 22.
—, 2000, A paleoearthquake record in varved Holocene sediments from ODP Leg 169S, Saanich Inlet, British Columbia: Geological Society of America, Cordilleran Section and associated societies, 96th annual meeting; abstracts Abstracts with Programs - Geological Society of America, v. 32, p. 22.
Kastens, K.A., 1984, Earthquakes as a triggering mechanism for debris flows and turbidites on the Calabrian Ridge: Mar. Geol., v. v. 55, p. p. 13-33.
—, 1984, Earthquakes as a triggering mechanism for debris flows and turbidites on the Calabrian Ridge: Marine Geology, v. vol. 55, p. pp. 13-33.
Lozefski, G., McHugh, C.M., Cormier, M., Seeber, L., Cagatay, N., and Okay, N., 2004, PROVENANCE OF TURBIDITE SANDS IN THE MARMARA SEA, TURKEY: A TOOL FOR SUBMARINE PALEOSEISMOLOGY: Geological Society of America Abstracts with Programs, v. 36, p. 120.
Nakajima, T., 2000, Initiation processes of turbidity currents; implications for assessments of recurrence intervals of offshore earthquakes using turbidites: Bulletin of the Geological Survey of Japan, v. 51, p. 79-87.
—, 2000, Initiation processes of turbidity currents; implications for assessments of recurrence intervals of offshore earthquakes using turbidites: Chishitsu Chosajo Geppo = Bulletin of the Geological Survey of Japan, v. 51, p. 79-87.
—, 2001, Translated: Initiation processes of turbidity currents; implications for assessments of recurrence intervals of offshore earthquakes using turbidites: Bull Geol. Survey Japan, v. v. 51, p. p. 79-87.
—, 2003, Earthquake potential inferred from turbidites on the Sado Ridge, in the eastern margin of the Japan Sea,: Earth Planest and Space.
—, 2004, Earthquake potential north of Sado Ridge, Japan Sea, inferred from turbidites, in Satake, K., Goldfinger, C., ed., Workshop on Turbidites as Earthquake Recorders: Tsukuba, Japan, AIST, Geological Survey of Japan.
Nakajima, T., and Kanai, Y., 2000, Sedimentary features of seismoturbidites triggered by the 1983 and older historical earthquakes in the eastern margin of the Japan Sea: Sedimentary Geology, v. 135, p. 1-19.
Noda, A., 2004, Turbidites along Kushiro Canyon, in Satake, K., Goldfinger, C., ed., Workshop on Turbidites as Earthquake Recorders: Tsukuba, Japan, AIST, Geological Survey of Japan.
Noda, A., Tsujino, T., Furukawa, R., and Yoshimoto, N., 2004, Character, Provenance, and Recurrence Intervals of Holocene Turbidites in the Kushiro Submarine Canyon, Eastern Hokkaido Forearc, Japan: Eos, Trans. AGU, Fall Meeting Abstracts, v. 86, p. T13C1369.
Noda, A., TuZino, T., Kanai, Y., Furukawa, R., and Uchida, J.-i., 2008, Paleoseismicity along the southern Kuril Trench deduced from submarine-fan turbidites: Marine Geology, v. 254, p. 73-90.
Okamura, Y., 2004, Paleoseismology of deep-sea faults based on marine surveys in Japan Sea, in Satake, K., Goldfinger, C., ed., Workshop on Turbidites as Earthquake Recorders: Tsukuba, Japan, AIST, Geological Survey of Japan.
Patton, J.R., Goldfinger, C., Morey, A., Erhardt, M., Black, B., Garrett, A.M., Djadjadihardja, Y.S., and Hanifa, U., 2009, 7.5 ka earthquake recurrence history in the region of the 2004 Sumatra-Andaman earthquake: Geological Society of America Abstracts with Programs, v. 41, p. 408.
—, 2010, Temporal Clustering and Recurrence of Holocene Paleoearthquakes in the Region of the 2004 Sumatra-Andaman Earthquake, Seismological Society of America Annual Meeting, Volume 41: Portland, OR, p. 408.
Prins, M.A., Postma, G., and Weltje, G.J., 2000, Controls on terrigenous sediment supply to the Arabian Sea during the late Quaternary; the Makran continental slope: Marine Geology, v. 169, p. 351-371.
Prunier, C.F., Karlin, R.E., and Pratt, T.L., 1996, Seismic landslides in Puget Sound (SLIPS); IV, Holocene neotectonics and mass-wasting in Lake Sammamish related to the Seattle Fault: Enriched: AGU 1996 fall meeting Eos, Transactions, American Geophysical Union, v. 77, p. 500.
Sarı, E., and Çağatay, M., 2006, Turbidites and their association with past earthquakes in the deep Çınarcık Basin of the Marmara Sea: Geo-Marine Letters, v. 26, p. 69-76.
Schnellmann, M., Anselmetti, F.S., Giardini, D., McKenzie, J.A., and Ward, S.N., 2002, Prehistoric earthquake history revealed by lacustrine slump deposits: Geology (Boulder), v. 30, p. 1131-1134.
Shiki, T., Cita, M. B., and Gorsline, D. S., 2000a, Sedimentary features of seismites, seismo-turbidites and tsunamiites—an introduction: Sedimentary Geology, vol. 135, p. vii-ix.
Shiki, T., Kumon, F., Inouchi, Y., Kontani, Y., Sakamoto, T., Tateishi, M., Matsubara, H., and Fukuyama , K., 2000b, Sedimentary features of the seismo-turbidites, Lake Biwa , Japan : Sedimentary Geology, v. 135, p. 37-50.
Skinner, M.R., and Bornhold, B.D., 2003, Slope Failures and Paleoseismicity, Effingham Inlet, Southern Vancouver Island, British Columbia, Canada, in Locat, J., and Mienert, J., eds., Submarine Mass Movements and Their Consequences: 1st International Symposium: Nice, France, Kluwer Academic Publishers, p. 375-382.
Soh, W., 2004, Coseismic seafloor displacement along Enshu fault detected from NSS coring, in Satake, K., Goldfinger, C., ed., Workshop on Turbidites as Earthquake Recorders: Tsukuba, Japan, AIST, Geological Survey of Japan.
Sternberg, R. W., 1986, Transport and accumulation of river-derived sediment on the Washington continental shelf: J. Geol. Soc. London, v. 143, p. 945-956.
St-Onge, G., Mulder, T., Piper, D.J.W., Hillaire-Marcel, C., and Stoner, J.S., 2004, Earthquake and flood-induced turbidites in the Saguenay Fjord (Québec): a Holocene paleoseismicity record: Quaternary Science Reviews, v. 23, p. 283-294.
Syvitski, J.P.M., and Schafer, C.T., 1996, Evidence for an earthquake-triggered basin collapse in Saguenay Fjord, Canada: Sedimentary Geology, v. 104, p. 127-153.
Thomson, J., and Weaver, P.P.E., 1994, An AMS radiocarbon method to determine the emplacement time of recent deep-sea turbidites: Sed. Geol., v. 89, p. 1-7.
Ujiie, H., Hyun, S., Oyakaa, T., Nakamura, T., Miyamoto, Y., and Park, j.-O., 1999, Holocene turbidite cores from the southern Ryukyu Trench slope: Suggestions of periodic earthquakes: Geol. Soc. Japan, v. v. 103NO. 6.
Ujiie, H., Nakamura, T., Miyamoto, Y., Park, J.-O., Hyun, S., and Oyakawa, T., 1997, Holocene turbidite cores from the southern Ryukyu Trench slope; suggestions of periodic earthquakes: Chishitsugaku Zasshi = Journal of the Geological Society of Japan, v. 103, p. 590-603.
Völker, D., Reichel, T., Wiedicke, M., and Heubeck, C., 2008, Turbidites deposited on Southern Central Chilean seamounts: Evidence for energetic turbidity currents: Marine Geology, v. 251, p. 15-31.
Witter, R., Zhang, Y., Goldfinger, C., Priest, G.R., and Wang, K., 2010, Validating Numerical Tsunami Simulations in Southern Oregon Using Late Holocene Records of Great Cascadia Earthquakes and Tsunamis, Seismological Society of America Annual Meeting: Portland, OR, Seismological Research Letters.