Technical Report
 

Paleoseismic Record of Peninsula Segment Earthquakes on the San Andreas Fault near San Francisco, CA and possible NSAF linkage to Cascadia

Public Deposited

Downloadable Content

Download PDF
Request Accessible Version
https://ir.library.oregonstate.edu/concern/technical_reports/ms35th18w

Descriptions

Attribute NameValues
Alternative Title
Creator
Abstract
  • Lake Merced results have established the lake as a viable paleoseismic site on the northern San Francisco peninsula. The lake records ~ 2300 years of event beds, 17 in total. Prior to that the lake was open to the ocean and contains marine shells consistent with an estuarine environment. The site was less tenable as a paleoseismic site at prior times, although that remains unexamined at present. Historic-era results from Lake Merced using bomb-carbon based age models show the site recorded three historic events, 1906, the 1957 Daly City earthquake, and Loma Prieta in 1989 with no intervening event beds. During the 20th century, these earthquakes resulted in the only three event beds in the lake during that time, providing a good test of the fidelity of the lake as an earthquake recorder. These results also show that this site records not only NSAF events, likely along the north coast and the peninsula, but also related earthquakes in the Santa Cruz Mountains, events in the probable pull apart basin at Daly City, and possibly others as well. The event bed corresponding to 1906 is ~ 1m thick, much greater than any other bed in the lake during the past 2300 years. Pb isotope data show this bed is contaminated with isotopes consistent mostly with lead based paint, and additionally tetraethyl Pb from gasoline. We interpret the thick, Pb laden deposit as likely anthropogenic material from the early development of San Francisco, and the great thickness of the bed may have been deposited over several decades as a hillslope effect following the earthquake. The Lake Merced record is largely compatible with onshore and offshore records for the north coast segment of the NSAF, and potentially with the lesser known peninsula segment. The lack of numerous “extra” events at Lake Merced relative to the north coast segment record may suggest that many, but not all north coast and peninsula segment events are one in the same, as was 1906, although numerous other interpretations are also possible. In an effort to examine the hypothesis of a stress relationship between the NSAF and Cascadia, we re-examined the evidence from both systems. Detailed analysis of cores near the northern limit of the NSAF at Noyo Canyon and in Southern Cascadia primarily at Trinidad Canyon, have yielded improved detailed records, age models, and evidence of historic earthquakes. In southern Cascadia, several significant event beds overlie the 1700 CE earthquake bed. High-resolution age models supported by bomb-carbon ages suggest that the significant 20th century beds likely include 1906 NSAF, 1980 Eureka, and 1992 Petrolia earthquake event beds. The presence of the 1906 beds in southern Cascadia suggests that others from the NSAF should be present as well, and also suggests that event beds from Cascadia earthquakes should be expected at similar ranges south of Cascadia in Noyo Canyon. In both Trinidad plunge pool and Noyo Canyon cores, a number of unusual doublet event beds are observed. The Noyo Canyon doublets are nearly all inverted, with the coarsest components at the top. In detail, these beds are composed of two stacked fining upward sequences, with the upper unit having greater density and coarser grain size. These unusual beds are unlikely to be the result of hydrodynamic transport effects or storm or climatic events. The doublet beds correspond to periods when both the NSAF and Cascadia had significant earthquakes at similar times. In all, of the 9 major earthquakes in the Cascadia record younger than 4100 years, 7 of them have a corresponding inverted doublet bed in Noyo Canyon with significant temporal overlap with or indistinguishable from Cascadia event timings. Five other Cascadia beds have temporal equivalents in Noyo Canyon, some of which do not correlate southward along the NSAF, suggesting Noyo recorded Cascadia events. For some of the doublet beds, time intervals above and below do not allow enough time for Cascadia beds to be present in Noyo Canyon, but mis-correlated on the basis of radiocarbon. We infer that the earthquake recurrence sequence that best explains the data is that for most Cascadia events < 4100 BP, a corresponding NSAF earthquake occurred shortly thereafter, producing the doublet stratigraphy in Noyo Canyon, and in Cascadia. The time separation between the Cascadia and NSAF ruptures is not known. We observe no interevent hemipelagic sedimentation between the two parts of the Noyo doublets, but erosion may have removed it. On the Cascadia side, the doublet for some events appears to be embedded in the tail of the first event, implying a very short time separation where the second event is settling into the still moving, waning turbidity current from the first event. The hypothesis of a stress-triggering interaction, and even partial synchronization of the NSAF with Cascadia is supported by the new stratigraphic and radiocarbon evidence of occurrence and timing. The data suggest that Cascadia ruptures first, and the NSAF thereafter. In Holocene times prior to ~ 4100 BP, the doublet stratigraphy is rare, and the weaker timing constraints available do not suggest such a relationship during that period. The 1906 event may have either ended a period of stress interactions, or is one of a small number of NSAF events unrelated to Cascadia in the past ~ 4100 years in a scenario of partial synchronization of the two faults.
License
Resource Type
Date Issued
Academic Affiliation
Subject
Rights Statement
Publisher
Peer Reviewed
Language
File Format
Accessibility Feature

Relationships

Parents:

This work has no parents.

Items

Thumbnail Title Date Uploaded Visibility Actions
file details LakeMerced2020FinalReport2021.pdf 2021-05-14 Public Download