Professional Contributions

50 Years of Paleoseismology:

The Science and the Business

he modern era of Paleoseismology arguably began 50 years ago, shortly after the 1971 San Fernando, California earthquake (M6.6). This was the nation’s most damaging urban earthquake since the 1933 Long Beach earthquake (M6.4), and the first to occur after active fault studies for nuclear power plants had begun in the mid-1960s (e.g. Schlocker et al., 1963).

TJames P. McCalpin and Eldon M. Gath

Paleoseismology, the Science

Paleoseismic studies in the 1970s studied deformation along the traces of known faults, some of which had experienced historic surface ruptures. These “tectonic geomorphology” studies did not involve trenching the fault trace (which came later), but nevertheless yielded the first crude estimates of active fault parameters needed for seismic hazard analysis (e.g. Wallace, 1970). USGS continued this geomorphic emphasis up to 1977, when the National Earthquake Hazards Reduction Program (NEHRP) was created. In addition to the USGS Internal Program, the NEHRP External program for academics and consultants was more applied, relying more on trenching and studying fault hazards in hitherto unstudied (or under-studied) geographic areas. External studies discovered many previously unknown Holocene-active faults, which paved the way for the Quaternary Fault and Fold Database of the USA. http://bit.ly/earthquakehazards

Because Paleoseismology developed during the past 50 years, we can observe how the field underwent a stairstep evolution of rapid advances (triggered by new techniques), separated by plateaus in which the new techniques were applied to studies over large geographic areas (Figure 1).

The 1970s fault-centric nature of paleoseismic studies has continued to this day and has advantages and disadvantages. Such “primary” studies do yield the seismic source parameters for individual faults (surface rupture length, displacement per event, slip rate, recurrence interval, maximum/characteristic magnitude) needed for seismic hazard analysis (SHA), which is a forward model (from cause to effect). The weakness of the method is if active faults exist in the studied area that are not known, the seismic hazards are underestimated. This is particularly true in areas of blind faulting, or where the characteristic earthquake magnitude is at or below the threshold for surface rupture (M~6). The alternative approach is to study a site’s record of strong ground shaking directly, as preserved by evidence of prehistoric liquefaction and/or other ground failures (landslides, lateral spreads, toppled rocks, etc.). This “secondary” approach is not affected by the problem of unknown active faults. It does have three weaknesses, however. First, most sites are not particularly susceptible to liquefaction or ground failure. Second, the size of observed liquefaction/ground failure features, based on historic observations, does not scale linearly with the strength of ground shaking. Third, liquefaction and ground failure can be triggered by nonseismic means. As a result, current predictions of probabilistic fault displacement hazards (PFDHA) and probabilistic ground motions (PSHA) are built from fault-specific evidence, with usually no regard for the presence or absence of secondary evidence at (or near) the site, even used as a reality check on the forward model results.

Figure 1. New techniques (vertical lines) that have advanced Paleoseismology; black, dating methods; medium gray, remote sensing methods; light gray, institutional changes. Impact on the science is reflected by the height of vertical lines (in our subjective viewpoint). Curved arrows show how later advances in a single technique (luminescence dating) have superseded earlier methods. C-14 calib, calendar-correction of radiocarbon ages; LSAP, Low-Sun-Angle Photography; TL, thermoluminescence dating; AMS, Accelerator Mass Spectrometry radiocarbon dating; OSL, Optically-Stimulated Luminescence dating; CRN, Cosmogenic Radionuclide dating; Digital Elevation Models; SAR, single aliquot or single grain luminescence dating; ALS, Airborne Laser Scanning (lidar); TLS, terrestrial lidar; SfM, structure from motions DEMs; UAVs, unmanned aerial vehicles (drones).

What About Interpretive Paradigms?

A new interpretive paradigm (e.g., plate tectonics) can have a bigger scientific impact than any one new field or laboratory technique. New paradigms often arise from new types of data, which themselves became possible from a new technique (e.g., Figure 1). A recent example is surface rupture patterns and their prediction by PFDHA. At present, predictive relationships for surface rupture are based solely on empirical data (i.e., no underlying physical or kinematic model). Not surprisingly, predicted outputs carry high uncertainty. The new remote sensing

techniques mentioned in Figure 1 (ALS, TLS, SfM, UAVs) permit mapping every tiny rupture trace, resulting in hundreds to thousands of rupture traces defined, and point displacements measured. For example, the Norcia, Italy, rupture of 2016 yielded 5,400 point measurements of displacement, most on distributed faults. Such data densities were not possible previously but are statistically robust enough to support development and testing of physical/kinematic models of principal and distributed faulting. For example, fault rupture may be partly controlled by depth to bedrock or by the rheology of the surface materials, neither of which are used in present PFDHA models. Underpinning PFDHA with a physical model should, in theory, decrease the uncertainty in predicted fault trace length, location, and displacement, because data points that clearly contradict the physical model can be deleted. But will the final uncertainty become small enough that engineering geologists and their owner/clients will rely upon them for site specific design and hazard mitigation? This is where the academic aspirations run into the practical business realities of cost-benefit analysis.

Science Outlook for the Future

What are the most likely near-term scientific advances? Based on trends of the first 50 years, advances will probably occur in dating methods, remote sensing, and perhaps geophysics: 1) Rapid Dating (within a few days), so dates can be known before the paleoseismic trench must be backfilled. This would prevent the syndrome of “Oh, if I would have known that faulted bed was Holocene, I would have….” Commercial labs currently offer “rush” AMS radiocarbon dating within one week, but for luminescence dating even “rush” dating takes months. Few practitioners can get permission to leave a trench open for months, so a more rapid dating turnaround would help with trench interpretation. 2) “Virtual Trenching” via shallow geophysics. Twenty years ago, attempts were made to log stratigraphy and structure beneath normal fault scarps using P- and S-wave seismic tomography, which was touted as “seismic trenching” (e.g., Sheley et al., 2003). However, the tomograms could not distinguish thin or small deposits or displacements, which were easily visible on the log made by conventional trench logging. This research thread has stalled in the past decade, but should be taken up again, because as urban fault traces become totally developed, traditional trenching is no longer possible. One possibility to maintain high resolution with depth is to augment surface geophysical surveys with downhole surveys, which can use wave-guide principles to trace subsurface deposits between boreholes. This could potentially increase the depth of detailed fault-zone imaging to tens of meters, regardless of depth to water table. 3) New remote sensing techniques that image the shallow subsurface. Just as lidar penetrates surface vegetation to reveal the bare-earth topography, a useful advance would be an aerial sensor that penetrates the upper few meters of the subsurface and can measure its material properties (density, moisture, dielectric properties). Such a sensor could directly image young, low-density materials deposited in topographic traps (grabens, ramps) in active fault zones, even where the traps have no surface expression today because they are filled with sediment. These are good trenching targets. 4) Refined methods to identify and analyze the paleoseismic signature of M<6 earthquakes. Californians have been waiting 110 years for a repeat of an M8 earthquake on the San Andreas. But for every M8 earthquake in a seismic cycle, there will be 10 M7s and 100 M 6s (think of San Fernando M6.6; Whittier Narrows, M6.0; Coalinga M6.2; Loma Prieta, M6.9; Northridge, M6.7; South Napa, M6.0). The cumulative damage from 100 such M6s will be as large or larger than a single M8. But the evidence is much smaller and harder to find. 5) Increased usage of secondary paleoseismic evidence (liquefaction, ground failures) to provide reality checks in PSHA (e.g., Fan et al., 2019).

Paleoseismology, the Business

Business needs for paleoseismic studies are separate from and independent of the science advances, which take place in academia and in government agencies (public policies based on new knowledge and experiences). The business need (or opportunity for paleoseismologists) is thereby generated by these public policies and regulations, while the engineering geologist’s ability to comply and solve their client’s problems is, to a large degree, facilitated by the scientific advances of the academic paleoseismology community. Unfortunately, these advances are often under the radar of the practicing engineering geology practitioner, or to quote Dr. Kerry Sieh in 2000, “the state of the knowledge is at least 10 years ahead of the actual use of that knowledge” (Yeats and Gath, 2005). This lag time led to the development of technical specialists in the applied paleoseismology discipline, perhaps starting as far back as Dr. Roy Shlemon and the fledgling California nuclear power industry in the mid-1970s.

Historically Dominant Market Sectors

Paleoseismic projects range over several market sectors: Water-related (Dams, Aqueducts, Tunnels); Energy-related (power plants, including nuclear; extraction sites, such as offshore drilling platforms; pipelines and terminals, such as LNG terminals); Waste disposal-related (high-level and low-level nuclear waste repositories; landfills); Transportation (highways, railroads); Land development-related (residential, commercial). In the 1960s–70s nuclear projects dominated the market at large scales, whereas in 1973 California’s Alquist-Priolo Fault Zoning Act required small-scale “paleoseismic” studies for residential and commercial land-use changes. The smaller budgets of the residential studies were offset by their sheer numbers (thousands), so cumulatively they were as important as the large-scale projects for critical facilities.

Paleoseismic Studies Driven by Regulations

Engineering geologists were aware of the fault rupture hazard before there was a scientific method to quantify that hazard

sufficient for risk reduction. Indeed, even the Alquist-Priolo Act was silent on the possibility of quantitative paleoseismology, requiring instead the strict avoidance (setback zone) from any Holocene-age faults. The Act was silent because the science did not yet exist from which to understand prior fault rupture timing, displacement magnitude, or even its spatial locations. The only mitigation permitted by law in California was, and unfortunately still is, strict structural avoidance. Since 1973 therefore, even though huge advances in paleoseismology have been made within the academic community, most engineering geologists in California have little need for them. Is this sediment layer or buried paleosol (soil) horizon Pleistocene or Holocene, and is it faulted or not? Period.

But large engineering projects are not necessarily subject to the limitations within the Alquist-Priolo Act, and structural mitigation for fault rupture displacement does look to paleoseismology to help answer design questions such as displacement magnitudes, kinematics, most recent event and knowledge of recurrence intervals. Of course, all of these parameters are unlikely to be obtained from any single study site, so it is also necessary for the practitioner to be able to define the uncertainties in a manner that can be used by the design engineer and understood by the project’s reviewers. Specific examples of these kinds of projects involve cutting the U.C. Berkeley football stadium in half to accommodate the Hayward fault’s current creep rate and future earthquake rupture, PG&E’s ongoing natural gas pipeline risk studies at fault crossings, and LA Metro’s fault rupture mitigation program for its subway tunnels.

Business Prospects for the Future

Predicting business trends in paleoseismology is even more uncertain than predicting its scientific advances. The current trending concerns are described below.

Market Sectors

Many countries have recently pledged to reduce or eliminate fossil fuels as a source of energy in favor of renewable energy, within the next decade or two. If this occurs (it will be expensive), it will reduce paleoseismic projects from the fossil fuel sector, such as oil and gas pipelines, offshore drilling rigs, and possibly nuclear power plants and waste repositories (after all, uranium is not a renewable energy source). The transition to renewable energy for stationary facilities (residential, commercial, industrial) and transportation (autos, Elon Musk’s Hyperloop, long-haul trucks, high-speed trains, airplanes) is basically a transition to electrical energy. Today electricity is generated by fossil-fuel powered, industrial-scale plants (=critical facilities), many of which require geologic hazard studies. In contrast, much future renewable electricity will be generated at widely dispersed points of use, which will not require geologic hazard studies. In the transportation sector, the increased speed of electrical vehicles such as high-speed trains may trigger a requirement for studying small ground movements, including tectonic ones. The amount of allowable track deflection for a 200–300-mph electric train is much smaller than for a 55-mph Amtrak coach.

The past 20 years has been a drought cycle in much of the western United States, and with the westward population shift from COVID-19, metro areas of the West are scrambling for new water supplies. At this time, Lakes Mead and Powell on the Colorado River contain only 37% and 34% of capacity, respectively. Lower levels will trigger a Lower Basin “water shortage condition,” resulting in decreased water allotments to Arizona, Nevada, and Mexico. This will spur new dam projects, pipelines, and aqueducts in earthquake country.

Regulatory Changes

California geologists hope to someday be able to use their paleoseismic tools and expertise under a modernized AlquistPriolo Act which opens up the mitigation alternatives to more than just avoidance. In the almost 50 years since its passage, its interpretation by regulators and state geologists has become increasingly prescriptive (Gath, 2015), while engineering mitigation of ground deformation has become increasingly performance and risk based, relying on huge increases in computer power for modeling, mechanical testing, materials science, and learning from earthquake studies to improve their professional practice (Committee, 2003). If engineering geologists were able to apply the current knowledge base in paleoseismology techniques to their projects the increase in knowledge for all of California’s faults would increase exponentially because now there would be hundreds if not thousands of projects per year wherein rupture recurrence, kinematics, and magnitudes would be built into the investigation plan and budget as it would finally be important for design and engineering mitigation. If the displacements exceed the capacity for mitigation, avoidance is still an option, but until engineers are allowed to try, paleoseismic-level geologic investigations cannot be defended as standard of care.

References

Committee to Develop a Long-Term Research Agenda for the Network for Earthquake Engineering Simulation (NEES), 2003, Preventing earthquake disasters: The grand challenge in earthquake engineering: A research agenda for the Network for Earthquake Engineering Simulation (NEES); National Academy of Science, National Research Council, Board on Infrastructure and the Constructed Environment, 138 p. Fan, X. and 16 others, 2019, Earthquake-induced chains of geologic hazards; patterns, mechanisms, and impacts: Rev. Geophys. 57: 421–503. Gath, E.M., 2015, Is the California Geological Survey’s position on active fault zoning and mitigation logical or ethical in light of 40+ years of earthquake geology research? (abs); Geol. Soc. Amer. Abstracts with Programs: 47(7):192.192. Schlocker, J., Bonilla, M.G. and Clebsch, A., Jr., 1963, Geologic and seismic investigations of a proposed nuclear power plant site on Bodega Head, Sonoma County, California; Part I, Geologic Investigations: U.S. Geol. Surv.

Trace Element Invest. Report TEI-837, p. 1–42. Sheley, D., Crosby, T., Zhou, M., Giacoma, J., Yu, J., He, R., and Schuster, G.T., 2002, 2D seismic trenching of colluvial wedges and faults: Tectonophys. 268(1): 51–69. Wallace, R.E., 1970, Earthquake recurrence intervals on the San Andreas fault: Geol. Soc. Amer. Bull. 81(10): 2875–2890. Yeats, R.S. and Gath, E.M., 2005, Paleoseismology of surface ruptures: research tool or standard of practice?: Association of Environmental & Engineering Geologists, AEG News, March, 2005, p. 21–23.

About the Authors

James McCalpin, BA 1972, University of Texas; MS 1975, University of Colorado; PhD 1981, Colorado School Mines, has worked in applied geomorphology, geohazards, and paleoseismology since 1976. He was County Geologist of Jefferson County (CO) in 1981–82, then taught geomorphology and engineering geology 1982–91 at Utah State University. Dr. McCalpin founded GEO-HAZ Consulting, Inc. in 1991 and has since performed nearly 200 consulting projects for clients worldwide. He has authored more than 160 papers in refereed journals and proceedings, 13 published geologic maps, and >130 consulting reports on geohazards for clients. His 1996 book Paleoseismology (Elsevier Publishing) won the AEG Holdredge (1999) and GSA Burwell (2000) awards. Paleosesismology 2nd Edition has been published in English (2009), Russian (2011), and Chinese (2020). Eldon Gath, Founder and President of Earth Consultants International of Santa Ana, California, is a Past-President and Life Member of AEG. He served on the AEG Board of Directors from 1990–98 and 2016–19, was the 2014–15 Jahns Lecturer, and granted the Floyd T. Johnston and E.B. Burwell awards. He has served on numerous technical committees and advisory boards for the US Geological Survey, National Research Council, Southern California Earthquake Center, State of California, IAEG, EERI, and others. Eldon has worked on projects throughout California and other western states, and in a dozen countries as diverse as Japan, Papua New Guinea, Panama, and Turkey. His projects have included gas storage fields, oil field redevelopment, city and county hazard management plans, pipelines, canals, dams, tunnels, and hundreds of “typical” engineering geology studies for development planning, design, and construction.

Annual Meeting Sponsor Opportunities

Sponsoring at our 2021 Annual Meeting in San Antonio, Texas is effective advertising that builds lasting recognition and good will for your company or organization.

Network with Industry Professionals:

■ Public and Private Owners of Utilities, Cities, Municipal

Water Districts, Water and Sewer Departments, Waste

Treatment Plant and Facilities ■ Transportation Departments, Developers, Financiers and more ■ Integrated Project Team Leaders ■ Senior Level Executives representing Environmental

Engineers, Contractors and Specialists ■ All the major suppliers and specialty contractors that support major Infrastructure projects

Plus These Benefits:

■ Increased exposure to potential clients & teaming partners ■ Reconnect with past colleagues and clients ■ Expand your business to a global market ■ Build lasting recognition for your company

Who Should Become a Sponsor? There are opportunites for large corporations:

Premium Event Package – $2995 ($4500 value)

which includes a corner exhibit booth, two full meeting registrations, sponsorship of the USB Drive Giveaway (your logo on USB drive), mobile app banner ad, and your logo on the attendee registration bag.

Smaller companies have numerous options:

$100-$500 Levels:

■ Speaker/Moderator Breakfasts – $250 each ■ Field Courses – $200 ■ Guest Tours – $150 ■ Student Mini-Grant Program - $250 ( Provides the

Student/Professional Networking Session giveaway to offset student travel expenses. ■ Technical Sessions and Symposia – $500 each. Your logo will be placed on the session opening slide. ■ Technical Session Breaks (Wednesday Morning, Wednesday

Afternoon, and Friday Afternoon left) – $250 each ■ Young at Heart Student and Young Professional Event – $250 ■ Mobile App—Sponsored Push Message Notifications (offered Wednesday through Friday only!) – $100 per message. Send real-time messages to app users! Notifications appear on user’s phone even if they’re not currently using the app! ■ Sponsored Maps – $250 Exhibit Hall Map, OR $250 Conference Center floor plan.

WE ARE FLEXIBLE! Please contact AEG Meeting Manager Heather Clark at 303-518-0618 or heather@aegweb.org.

The Role of Paleoseismic Investigations in the Engineering of Pipeline Fault Crossings

James Hengesh, Douglas Nyman, and David Waring

Introduction

ipelines are geographically distributed systems that cross a variety of geological environments and are exposed to a diverse range of ground conditions and geological hazards. These hazards must be characterized P for the route selection, basic design, detailed design, and construction stages of a project. Experience has proven that failing to recognize and mitigate geological hazards for pipeline projects can result in a loss of pipeline integrity leading to direct costs from pipeline damage, environmental impacts, repair and monitoring costs, and indirect losses due to business interruption (Lee et al., 2009). For example, several segments of the

Trans-Ecuador oil pipeline in the Matagua River Valley, Ecuador, were severed and dislocated tens of meters by flow failures during the 1987 Mw 7.1 Ecuador earthquake (Nieto et al. 1991). Costs associated with repairs and lost revenue from the 1987 earthquake reached roughly U.S. $1 billion (1987 dollars;

O’Rourke and Liu, 1999). Fault crossings are significant because pipelines that cross a surface fault rupture must deform longitudinally and in flexure to accommodate the associated permanent ground deformation (Nyman and Bouckovalas, 2019; Honegger and Nyman, 2017). If a pipeline crosses a potentially active fault, it is necessary to characterize its activity, location, orientation, amount and type of potential displacement, slip characteristics, and zone of disturbance. Paleoseismic investigations can play a key role in determining the activity of a fault and parameters needed for pipeline engineering and design.

Data Requirements for Characterization of Fault Crossings

Once it is confirmed that an existing pipeline or proposed route crosses an active fault, it is necessary to evaluate the fault’s characteristics to support the engineering and design of the fault crossing mitigation measures. At a minimum, the following information is required to characterize active fault crossings: 1. Fault activity and slip rate; 2. Location of the main fault trace, secondary splays, and width of permanent ground deformation zone; 3. Fault geometry – strike direction and dip angle; 4. Style of faulting – strike-slip, normal, reverse, or oblique; 5. Slip partitioning among fault splays; 6. Fault length, width, area, and displacement distribution per surface rupture event; 7. Maximum earthquake magnitude distributions and recurrence intervals; 8. Design fault displacement (defined as orthogonal components of displacement in three-dimensional space); 9. Detailed topographic profile; and, 10. Soil types and geotechnical parameters in the locality of the fault crossing.

Compiling the above information requires that geological investigations be carried out to provide the engineer with the information necessary for design of the fault crossing.

Investigative Approaches

Geological and Geomorphological Mapping Approach

Up through the 1990s, the geological investigations carried out in support of pipeline engineering projects tended to be limited to geological and geomorphic mapping. This approach can be sufficient if there are Quaternary-age deposits or geomorphic surfaces that cross a fault zone. Figure 1 shows an example of a Quaternary age depositional surface that is offset by faulting. In other cases, the continuity of deposits or landforms may provide evidence of the absence of tectonic activity and thereby confirm that an engineered pipeline fault crossing design is unnecessary.

Surficial geological mapping has successfully provided information to support engineering analyses and design of mitigation measures for major pipeline systems. For example, the geological investigations carried out for the Trans-Alaska Pipeline System (TAPS) in 1973 successfully identified the locations of the Denali fault (Cluff et al., 2003). The information

Figure 1. Given suitable deposits and exposures a faults location and sense of deformation can be estimated based on field observations. At this location in Owens Valley, California, the 767 Ka Bishop Tuff and late Quaternary fluvial terrace deposits (black arrows) are offset (red arrows); note the left-lateral oblique offset of Owens River Gorge. Oblique aerial view to southeast.

developed during these investigations was used to support the engineering of the TAPS pipeline at the fault crossing. In 2002, the fault crossing design for the Denali fault was tested. The Mw 7.9 Denali earthquake produced 5.5 m of ground displacement at the TAPS crossing location consistent with the estimated 6.1 m design displacement. The mitigation measures put in place at the Denali fault crossing enabled the pipeline to successfully accommodate the permanent ground deformation produced by the event (Sorensen and Meyer, 2003; Hall et al., 2003).

Characterizing a fault using only a surficial mapping approach can be challenging in a number of situations, such as: along low activity faults; in areas with poor exposure; at locations that lack suitable Quaternary deposits or morphological features; and where the style of deformation may be complex. The accurate determination of 3D fault slip vectors is essential for establishing an optimal pipeline crossing alignment or analyzing an existing crossing. Thus, in most cases, reliance on surface mapping alone is insufficient for definition of displacement parameters, and there is a clear role for paleoseismic trenching investigations.

Paleoseismic Trenching Approach

Paleoseismic trenching is one of the most effective techniques for evaluating surface fault rupture hazards that may impact pipelines. A trench exposes the stratigraphic and structural features in the shallow subsurface along the fault zone (Figure 2a) and enables the direct observation, description, measurement, and dating of critical features. The purpose of a trenching investigation is to determine the precise location where a pipeline crosses a fault, and to document the fault’s activity, orientation, style of faulting, and amount of displacement expected during an earthquake. Where conditions are favorable, trenches also can yield information on dates of past surface rupturing events, earthquake recurrence intervals, fault slip rates, and slip vectors (Figure 2b). One of the principal benefits of fault trenching for pipeline projects is the ability to more reliably characterize key input parameters to support engineering activities. We recently carried out paleoseismic trenching investigations along several existing pipelines to validate fault locations and design parameters for faults that previously were assessed using only surface mapping techniques. Although the surface mapping was carried out by experienced geologists, the limitations of the approach included residual uncertainties in fault location, activity and style of deformation.

The fault validation trenching program resulted in significant revisions to the interpretations of fault parameters that were developed based on surface mapping. The results of the trenching revealed the following: ■ Some faults previously assessed as active did not meet the criteria of Holocene activity; ■ Some previously mapped fault traces were mislocated; and, ■ The previously interpreted fault geometry, style of deformation and slip vectors were incorrect.

Data from the paleoseismic trenching program reduced the uncertainties associated with the input parameters needed for fault crossing engineering. This information has led to design of fault crossing mitigation measures where the pipelines have significantly greater displacement capacity and a much lower residual risk of losing pressure integrity.

a b

Figure 2. (a) Example of paleoseismic trench completed to investigate a potential pipeline fault crossing; (b) fault zone (white arrows) exposed in trench showing horizontal slip vectors on fault plane (black arrow).

Supporting Risk Assessments

Owners of large pipeline systems are now completing systemwide risk analyses to assess their corporate exposure to a range of potential impacts. These risk analyses typically include a broad range of operational and environmental issues, including geohazard events. Because paleoseismic trenching provides the opportunity to document a fault’s movement history and to sample and date materials related to past earthquakes, fault slip rates and earthquake recurrence intervals often can be assessed. These data can yield an annualized frequency, or rate of earthquake occurrence that may be incorporated in the risk assessments. By combining the displacement capacity of the pipeline at a fault crossing with the rate of occurrence of the design earthquake, the residual risk of pipeline failure (defined as loss of pressure integrity) can be assessed (Figure 3—next page). This residual risk of failure can be incorporated into a company’s overall risk assessment. In areas where fault crossing mitigation measures have not been put in place, this type of risk assessment also can be used to communicate within a corporate risk framework that areas of high risk exist, and to identify and prioritize locations for future investigation and mitigation.

Conclusions

Paleoseismic trenching investigations have been completed along several existing pipeline routes where fault crossing designs were based on fault characterizations sourced from surface geological and geomorphological mapping. The results have shown that the surficial mapping approach can have large uncertainties that may lead to: (i) inaccurately identifying the location of an active fault trace; (ii) inferring that a nontectonic feature is fault related; (iii) incorrectly assessing that an older fault is Holocene active (or vice-versa); (iv) mischaracterizing the fault orientation, style of deformation, and 3D slip vectors

Figure 3. Example of probability of failure analysis for a mitigated pipeline fault crossing. Figure courtesy of GR8 GEO.

for a fault crossing; and (v) inaccurately estimating fault slip rates and recurrence intervals.

All of these uncertainties can affect project engineering and the quantification of risk. Incorrectly characterizing the fault parameters can result in misplacement of pipeline crossing mitigation measures, unfavorable alignment of the pipeline across the zone of deformation, and inadequate pipeline displacement capacity to accommodate ground surface rupture. The longterm effect is that the pipeline system may have considerably greater residual risk of failure than if resources were invested at the start of the project to develop reliable inputs to support subsequent engineering activities. Although each fault location is unique and needs to be assessed on a case-by-case basis, the results of our field investigations have clearly demonstrated the value of paleoseismic trenching. The data developed during trenching programs support of the implementation of robust engineering mitigation measures that reduce the risk of failure at pipeline fault crossings.

References

Cluff, L.S., Page, R.A., Slemmons, D.B., and Crouse, C.B., 2003, Seismic hazard exposure for the Trans-Alaska Pipe-line, in Beavers, J.E., ed., Advancing mitigation technologies and disaster response for lifeline systems: ASCE, Sixth U.S. Conference and Workshop on Lifeline Earthquake Engineering, ASCE Technical Council on Lifeline Engineering, August 10–13, Long Beach, CA, p. 535-546. Hall, W.J., D.J. Nyman, E.R. Johnson, and Norton, J.D., 2003, Performance of the Trans-Alaska pipeline in the November 3, 2002 Denali fault earthquake, in Beavers, J.E., ed., Advancing mitigation technologies and disaster response for lifeline systems: ASCE, Sixth U.S. Conference and Workshop on Lifeline Engineering, August 10-13, Long Beach, CA, p. 522-534. Honegger, D.G., and Nyman, D.J., 2017, Pipeline design and assessment guideline (2017 Revision): Pipeline Research Council International, Inc., Catalog No. L51927-R01. Lee, E.M., Audibert, J.M.E., Hengesh, J.V. and Nyman, D.J., 2009, Landsliderelated ruptures of the Camisea pipeline system, Peru: Quarterly Journal of Engineering Geology and Hydrogeology; v. 42, p. 251-259, doi:10.1144/1470-9236/08-061. Nieto, A.S., and Schuster, R.L., Plaza-Nieto, G., 1991, Mass wasting and flooding: in Chung, R.M., ed., The March 5, 1987, Ecuador earthquakes, mass wasting and socioeconomic effects: National Research Council, Natural Disaster Studies, v. 5, p. 51-82, National Academy Press, Washington, D.C. Nyman, D.J., and Bouckovalas, G.A., 2019, Assessment and mitigation of seismic geohazards for pipelines, in Rizkalla, M., and Read, R., eds., Pipeline geohazards: planning, design, construction and operation, ASME, Chapter 11. O’Rourke, M.J. and Liu X., 1999, Response of buried pipelines subject to earthquake effects: MCEER Monograph No. 3, 247p. Sorensen, S.P., and Meyer, K.J., 2003, Effect of the Denali fault rupture on the Trans-Alaska Pipeline,” in Beavers, J.E., ed., Advancing mitigation technologies and disaster response for lifeline systems: ASCE, Sixth U.S. Conference and Workshop on Lifeline Earthquake Engineering, August 10-13, Long Beach, CA, p. 576-586.

About the Authors

James Hengesh, Principal Geologist at Interface Geohazard Consulting, has 34 years of experience conducting investigations to characterize geological and seismic hazards for major infrastructure projects worldwide. Dr. Hengesh has served as technical leader and expert reviewer for assessments of both onshore and offshore pipelines. He has characterized surface fault rupture hazards to support engineering analyses and design of mitigation measures for pipelines in Alaska, Albania, Australia, Azerbaijan, Georgia, Greece, India, Papua New Guinea, Peru, Russia, Tanzania, Trinidad, and Turkey. Douglas Nyman, Principal Engineer of D.J. Nyman & Associates, has 45 years of experience in the mitigation of earthquake and geological hazards for pipeline systems. He provides engineering services to support design of onshore and offshore pipelines to withstand large ground displacements and extreme or unusual load conditions. David Waring is an Engineering Geologist, who graduated from Imperial College London. David is currently employed by BP and provides onshore geotechnical and geohazard technical assurance for projects and operational sites across the integrated energy company.

iPad Lidar Scanning for 3D Trenching:

A New Methodology for Paleoseismologists Demonstrated on the Dog Valley Fault, Truckee, California

Ian K.D. Pierce and Rich D. Koehler

Introduction

he advent of Structure-from-Motion (SfM) photogrammetry has revolutionized the acquisition of cheap, rapid, and accurate 3D full color data for a wide variety of tectonic and geomorphic studies (e.g., Westoby T et al., 2012; Reitman et al., 2015; Angster et al., 2016; Pierce et al., 2021; Delano et al., 2021). SfM is now the preferred methodology for producing detailed and accurate trench photomosaics. One of the problems facing this application is a lack of a local reference or scale. Various methods have been developed to counter this, including the use of highly accurate, but labor intensive and costly survey equipment (e.g.: Reitman et al., 2015) or scale bars placed in the scene (e.g.: Delano et al., 2021). Here we demonstrate a new methodology using the laser scanner built into a 2020 iPad Pro to provide reference and accurately scale SfM models. For this study, we focus on the Dog Valley fault north of Lake Tahoe, California, and present reconnaissance fault trace mapping (lidar- and fieldbased) and preliminary paleoseismic trenching results from an exposure along the eastern end of the Dog Valley fault. This ongoing work will contribute towards a better understanding of regional seismic hazards and the role of the Dog Valley fault in accommodating strain in the northern Walker Lane.

The Dog Valley Fault

The Dog Valley fault in northeast California is a northeaststriking, left-lateral strike-slip fault that extends for ~25 km from north of Truckee, California, to the north flank of Peavine Mountain near Reno, Nevada (Figures 1 and 2). At this latitude, about 10 to 15 percent (5–7 mm/yr) of Pacific/North American plate relative motion (dextral shear) is distributed across the northern Walker Lane with about 2–3 mm/yr concentrated along its western margin (Hammond et al., 2011; Bormann, 2013, 2016; Pierce et al., 2021). This deformation is accommodated by active transtensional faulting along the Dog Valley fault as well as a set of conjugate faults including the northwest-striking, right-lateral Polaris (e.g., Hunter et al., 2011) and Truckee faults. The system has generated several strong historical earthquakes (1966 M6.6 and two ~M6 earthquakes in 1914 and 1948), numerous smaller earthquakes (e.g.: 2021 M4.7), and poses a significant surface fault rupture and strong ground motion hazard for the communities of Truckee and Reno, as well as several water storage dams in the region.

Previous assessments of the Dog Valley fault have been limited to seismic hazards evaluations related to dam safety (Olig et al., 2005; Hawkins et al., 1986). Olig et al. (2005) described tectonic geomorphic features along the fault including side-hill benches, ridge-crest saddles, aligned linear drainages, and reversals in scarp directions. Additionally, Olig et al. (2005) inferred a cumulative left-lateral displacement of 3.6-4.0 km since ~3 Ma. Several trenching efforts by Hawkins et al. (1986) did not expose the fault, however their trenches may have been located across scarps with adverse groundwater conditions and/or secondary cracks related to the 1966 earthquake.

Despite this documented activity, the Dog Valley fault is not included in the US Geological Survey National Seismic Hazard model. Slip rate estimates for the Dog Valley fault are uncertain due to the lack of absolute age control for faulted surfaces, and paleoseismic parameters are non-existent. Thus, we initiated a study to better characterize this fault for seismic hazards applications.

Figure 1. Regional fault map. The Dog Valley fault is part of the Northern Walker Lane, an area of distributed dextral shear that accommodates ~5-7 mm/yr of northwest directed dextral shear, subparallel to the San Andreas fault. Faults (red) modified after USGS Quaternary Fault & Fold Database.

Figure 2. Map of tectonic geomorphic features and fault traces along the Dog Valley and Polaris faults. Yellow stars indicate sites of previous studies.

Figure 3. Photo of faulted Miocene andesitic breccia and tuff in a roadcut adjacent to the east side of Stampede Dam. Holocene fault scarps are visible in lidar data in the hillside above this roadcut. Figure 4. UPPER: Photo of excavator at the toe of the fault scarp. LOWER: East wall of the trench showing sharp juxtaposition of fluvial units across an apparently dipping fault.

Mapping

Tectonic geomorphic features and fault traces were mapped along the length of the Dog Valley fault based on interpretation of lidar hillshade maps and field reconnaissance. Mapping along the Dog Valley fault indicates that it is characterized by rightstepping, en échelon fault strands expressed by subdued geomorphic features including oppositely facing scarps, closed depressions, aligned linear ridges, and sidehill benches (Figure 2). The lidar data reveal that the trace of the Dog Valley fault goes through the Stampede Dam, maintained by the US Bureau of Reclamation. Field observations indicate that strands of the Dog Valley fault offset Miocene andesite flow breccias and tuff breccias exposed in a roadcut immediately northeast of the dam (Figure 3). Thus, future fault rupture may pose potentially severe, and underappreciated seismic hazard for the downstream communities including Reno. A paleoseismic trench (DV1) was excavated approximately 3 km northeast of Stampede Reservoir where the fault forms a 2-m high northwest-facing scarp across the mouth of a small alluvial valley (Figure 4). An apparent left deflection of an ephemeral stream channel adjacent to the trench occurs along the strike of the fault.

A New Trenching Methodology

We were fortunate to have access to a small excavator, which remained on site for the duration of the project. After initially logging the first cut of the trench, we used the excavator to progressively peel back trench walls, ~20 cm at a time, creating both fault-parallel and fault-orthogonal slices. Each trench slice was scanned using a 2020 iPad Pro lidar scanner with the SiteScape app and photographed using overlapping photos from a Google Pixel2 smartphone camera. The photos were processed using Agisoft Metashape software to construct SfM models. The dense colored point clouds from the SfM models were then exported into CloudCompare software, which was used first to reference the lidar scans using flags in areas of the trench wall that did not change during progressive excavation as tie points (Figure 5, middle). Then, the SfM point clouds

Figure 5. Top: Photo of the 3D trench mid-excavation. Middle: Similarly oriented aligned lidar scans and higher resolution SfM point cloud (bottom) fitted to the lidar scans. The dotted white lines in the lower scene shows the channel margins that are left-laterally displaced ~80 cm by the fault (dashed red). Figure 6. Plan view of CloudCompare scene showing the referenced point clouds of 10 of the 14 slices. Fault approximate location dashed in red.

Figure 7. Orthophotomosaic and stratigraphic log of the west wall of trench DV1 along the Dog Valley fault.

were referenced to each of the corresponding lidar scans, again using flags or small stones in each of the lidar and SfM clouds as tie points (Figure 5, lower). The final result is a single high-resolution, spatially accurate 3D model of all 14 trench slices (Figure 6).

The Trench

The trench exposed a sequence of low-energy saturated interbedded fluvial overbank sands with cobble lenses and peaty buried meadow soils (Units 3 and 4). These deposits are clearly juxtaposed against relatively higher-energy dry indurated fluvial coarse sandy silt and gravel channel deposits (Units 5 and 6) (Figure 7). A massive-to-weakly stratified brown-gray silt (Unit 2) and the modern soil (Unit 1) overlies these deposits. The stratigraphic relationships indicate the occurrence of at least one earthquake that broke the entire stratigraphic package. Radiocarbon analyses on five charcoal samples constrain the age of the earthquake to after ~8,100 cal yr. BP (samples and ages shown on trench log, Figure 7). Several upward fault terminations were observed lower in the stratigraphy however, the similarity of deposits and the flooded conditions at the base of the trench precluded confidently attributing these features to additional paleoseismic events. A fault-parallel slice of the trench (P2 in Figure 6) exposed a sand dike that nearly reached the surface, indicating limited post-faulting deposition and precluding dating any post-event deposit that could bracket the age of the event.

The observed facies and thickness changes of stratigraphic units across the fault are consistent with strike-slip displacement. Thus, to assess the amount of lateral displacement in the most recent event, we expanded trench DV1 by cutting fourteen parallel and perpendicular slices of the wall to track a prominent channel margin. The 3D trench scene was used to reconstruct the offset channel margin. The model reconstruction confirmed field measurements indicating that the channel intersects the fault at a high angle and was left-laterally displaced during the most recent earthquake by ~80 cm (Figure 5).

Discussion/Conclusions

Observations from trench DV1 along the Dog Valley fault indicate the occurrence of at least one earthquake that occurred after ~8,100 cal yr B.P., which was associated with ~0.8 m of left-lateral displacement. Comparing this result to previously reported earthquake timing data from the conjugate Polaris fault indicates that the most recent earthquake along both faults post-dates ~7-8 ka (Melody et al., 2012). Although broadly constrained, the earthquake history allows the possibility that ruptures along each fault occurred closely spaced in time or possibly contemporaneously. Conjugate ruptures have occurred in several historical earthquakes both globally and in the Walker Lane (e.g., 2019 Ridgecrest; 2015 Nine-Mile Ranch; 1986 Chalfant Valley) and suggests that this style of strain release may be an underappreciated seismic hazard in the Walker Lane.

Acknowledgements

We are grateful for discussions in the field with T. Dawson, G. Seitz, J. Zachariasen, S. Wesnousky, J. Bormann, J. McNeil, K. Adams, and K. Knudsen, and assistance from R. Arrowsmith in reconstructing the trench model. Research supported by US Geological Survey (G20AP00055). Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the United States Government.

References

Angster, S., Wesnousky, S.G., Huang, W., Kent, G., Nakata, T., and Goto, H., 2016, Application of UAV photography to refining the slip rate on the Pyramid Lake fault zone, Nevada: Bulletin of the Seismological Society of America, 106, no. 2, 785–798. Bormann, J., 2013, New insights into strain accumulation and release in the central and northern Walker Lane, Pacific–North American plate boundary, California and Nevada, USA: PhD Dissertation, Univ. of Nevada, Reno, Nevada. Bormann, J., Hammond, W.C., Kreemer, C., and Blewitt, G., 2016, Accommodation of missing shear strain in the central Walker Lane, western North America: Constraints from dense GPS measurements: Earth and Planetary Science Letters, 440, 169–177. Delano, J.E., Briggs, R.W., DuRoss, C.B., and Gold, R.D., 2021, Quick and dirty (and accurate) 3D paleoseismic trench models using coded scale bars:

Seismological Research Letters, XX, 1–12. Hammond, W. C., Blewitt, G., and Kreemer C., 2011, Block modeling of crustal deformation of the northern Walker Lane and Basin and Range from GPS velocities: Journal of Geophysical Research, 116. Hawkins, F.F., LaForge, R., and Hanson, R.A., 1986, Seismotectonic study of the Truckee/Lake Tahoe area northeastern Sierra Nevada California for Stampede, Prosser Creek, Boca, and Lake Tahoe dams: U.S. Bureau of Reclamation Report no. 85-4. Hunter, L.E., Howle, J.J., Rose, R.S., and Bawden, G.W., 2011, Lidar-assisted identification of an active fault near Truckee, California: Bulletin of the Seismological Society of America, 101, 1162–1181. Melody, A.D., Whitney, B.B., and Slack, C.G., 2012, Late Pleistocene and Holocene faulting in the western Truckee Basin north of Truckee, California:

Bulletin of the Seismological Society of America, 102, 2219 –2224. Olig, S., Sawyer, T., Anderson, L., Wright, D., Wong, I., and Terra, F., 2005, Insights into Quaternary strain patterns in the northern Walker Lane from mapping and source characterization of faults near Truckee, California:

Seismolological Research Letters, v. 75, p. 251. Pierce, I.K.D., Wesnousky, S.G., Owen, L.A., Bormann, J.M., Li, X., and Caffee, M., 2021, Accommodation of plate motion in an incipient strike-slip system: The central Walker Lane: Tectonics, 40, e2019TC005612. Reitman, N.G., Bennett, S.E.K., Gold, R.D., Briggs, R.W., and DuRoss, C.B., 2015, High-resolution trench photomosaics from image-based modeling: Workflow and error analysis: Bulletin of the Seismological Society of America, 105, no. 5, 2354–2366. Westoby, M.J., Brasington, J., Glasser, N.F., Hambrey, M.J., Reynolds, J.M., 2012, ‘Structure-from-Motion’ photogrammetry: A low-cost, effective tool for geoscience applications: Geomorphology 179, 300–314. https://doi.org/10.1016/j.geomorph.2012.08.021.

About the Authors:

Dr. Ian Pierce is a Postdoctoral Research

Fellow at the Centre for the

Observation and Modeling of Earthquakes Volcanoes and Tectonics at the University of Oxford, UK. Ian Pierce is supported by the Leverhulme Trust. He completed his PhD, entitled Active Faulting in the Central Walker Lane, in 2019.

Dr. Rich Koehler is an Associate Professor of Geology at the Nevada

Bureau of Mines and Geology and University of Nevada, Reno. He has worked closely with Dr. Pierce on several recent projects.

CALL FOR PAPERS

AEG’s journal Environmental & Engineering Geoscience (E&EG) is currently seeking articles for review and possible publication. E&EG is dedicated to publishing peerreviewed, high quality, original research and case studies on environmental geology, engineering geology, geotechnical engineering, geomorphology, low-temperature geochemistry, applied hydrogeology, and near-surface processes. Visit https://www.editorialmanager.com/eeg, the Peer Track on-line submission platform, for Instructions for Authors, the Style Guide, and to login. Published quarterly and hosted at GeoScienceWorld (http://eeg .geoscienceworld.org/), E&EG publishes articles ahead-of-issue, as soon as they are ready. Current and past issues are available to GSA members who select the journal as part of their annual dues and subscriptions options and to AEG members as a benefit of membership. For more information or to discuss possible special issue topics, please contact a co-editor: Abdul Shakoor (ashakoor@kent.edu) or Eric Peterson (ewpeter@ilstu.edu).

AEG/USSD Workshop AEG 2023 Annual Meeting …for These Future AEG EventsSave the Dates

Westin Fort Lauderdale Beach Resort

December 6–8, 2021

This workshop will focus on Best Practices in Specialty Geotechnical Construction Techniques for Dam Exploration and Remediation. Information will be posted soon on www.aegmeetings.org.

Portland Marriott Downtown Waterfront – Portland, Oregon

September 19–24

AEG 2022 Annual Meeting

Planet Hollywood Resort – Las Vegas, Nevada

September 13–17

Your fun begins on The Strip, Las Vegas’ main artery and home to bustling casinos, worldclass restaurants, ornately designed resorts and over-the-top shows. Overlooking the Las Vegas Strip, our Annual Meeting hotel—Planet Hollywood Resort—is in the middle of everything you will want to experience. This Annual Meeting will be exciting, educational, and a great opportunity to reconnect with your colleagues and meet new ones. The In-Person AEG 2020 Annual Meeting has been rescheduled for 2023. Portland, Oregon’s largest city, sits on the Columbia and Willamette rivers, in the shadow of snow-capped Mount Hood. It’s known for its parks, bridges, and bicycle paths, as well as for its ecofriendliness and its microbreweries and coffeehouses. The city hosts thriving art, theater, and music scenes. Surrounded by Portland hot spots, the Marriott Portland Downtown Waterfront hotel reflects the vibrancy of the city and is located directly on the waterfront in downtown, surrounded by more than 60 breweries. Wind down with friends by sampling some of the 150 types of whiskey at their on-site restaurant, Proof Reader. The hotel even helps you go green with complimentary car-charging stations, biodegradable straws and easy access to light rail from the nearby Morrison Subway Station. At night, retire to elevated guest rooms with 24-hour room service, waterfall showers, free Wi-Fi, premium channels and views of Portland. Don’t miss all of the exciting Field Courses, Guest Tours and the Special Event at the Oswago Hills Vineyard planned for this incredible Annual Meeting.

Paleoseismology in Wine Country:

Paleoseismic Investigations along the Rodgers Creek Fault in Sonoma County, California

Robert W. Givler, John Baldwin, and Christopher Bloszies

Introduction

he most recent Uniform California Earthquake Rupture Forecast (UCERF 3.0) (Field et al., 2015) assigns a probability of 32% that the HaywardRodgers Creek fault (HRCF) will produce an earthquake T of M≥6.7 in the next 30 years, the highest probability for any

San Francisco Bay Region fault other than the San Andreas fault (Figure 1). As compared to previous models, the UCERF 3.0 model also increased the number of multi-fault ruptures and doubled the probability of M>7.2 events (Field et al., 2015). The age of the most recent event (MRE) and earthquake recurrence are important parameters (especially for time-dependent models like UCERF 3.0) for developing rupture scenarios, as well as computing probabilities of future major earthquakes along the HRCF. Paleoseismic data currently considered in the UCERF 3.0 model for the Rodgers Creek fault (RCF) are sparse and based primarily on two sites along the central part of the RCF (Figure 1; Budding et al., 1991; Schwartz et al., 1992;

Hecker et al., 2005). Our recent research conducted at two sites located along the southern RCF in Sonoma County,

California—Tolay Marsh (Rams Gate Winery) and Cline

Vineyard Sites—focused on obtaining paleoseismic data to improve existing forecasting models.

Rodgers Creek Fault

The RCF represents the central part of a 275-km-long fault system that includes the Hayward, Rodgers Creek-Healdsburg, and Maacama faults (Figure 1; McLaughlin et al., 2012). The 60-km-long RCF has a late Holocene dextral slip rate of 6.4 to 10.4 mm/yr (Budding et al., 1991; Schwartz et al., 1992; Hecker et al., 2005; Blisniuk and Walker, 2018). The central and southern sections of the RCF are divided by a possible 1km-wide step over, and the southern section is separated from the Hayward fault across a 4-km-wide right bend through San Pablo Bay (Figure 1; Watt et al., 2016). The RCF has had several historic moderate-sized earthquakes including the 1969 M5.6 and M5.7 earthquakes near Santa Rosa (Wong, 1991), and the poorly located 1898 M6.2 to M6.7 Mare Island events (Figure 2; Toppozada et al., 1992).

Figure 1. Southern Rodgers Creek fault (primary strands shown in red from Hecker et al., 2018) and paleoseismic sites along the fault (Buntz et al. 1991; Schwartz et al., 1992; Randolph-Loar, 2002; 2003; Hecker et al., 2005; Givler et al., 2010; 2016; 2018). Blue dashed line presents the historical margin of tidal marsh. Black lines are other Quaternary faults.

Figure 2. Tolay Marsh site geologic map (Witter et al., 2006) illustrating the location of the USGS seismic line collected as part of this study and gouge core transects #1-3. Holocene and Late Quaternary faults shown in red and orange. Suspected Holocene faults shown as thin green lines. Blue line represents the historic limit of tidal marshes. Geologic units include artificial fill (af), Quaternary Holocene fan (Qhf), Quaternary Holocene channel (Qhc), Quaternary Holocene terrace (Qht), Quaternary Holocene Bay mud (Qhbm), and bedrock (br).

Active aseismic creep has been well documented both south and north of the RCF, along the Hayward and the Maacama faults, respectively (Figure 1). Recent studies also document creep along the RCF, at an approximate rate of 1.5±0.3 mm/yr (Lienkaemper et al., 2014; McFarland et al., 2016). Data from geodetic measurements and from PS-InSAR datasets indicate a faster creep rate up to or >6 mm/yr, near Santa Rosa (Floyd, et al., 2009; Funning et al., 2007). The lowest measured rate of creep of 0.3±0.5 mm/yr is along the southern RCF, located 3.6 km to the south of the Cline Vineyard site (site RCWD in McFarland et al., 2016). Dynamic rupture modeling of the HRCF suggests that earthquakes may nucleate on locked parts of the fault (e.g., the southern RCF) and rupture through creeping strands (e.g., southern Hayward fault) leading to M7+ earthquakes (Harris et al., 2021).

Existing RCF Paleoseismic Data

Four previous paleoseismic investigations on the RCF provide the basis for the late Holocene event chronology for the central and southern sections of the RCF (Figure 2; Budding et. al, 1991; Schwartz et. al, 1992; Randolph-Loar, 2002; RandolphLoar et al., 2004; Hecker et al., 2005). Collectively, these studies yield an average earthquake recurrence interval of 131–370 years (Budding et. al, 1991; Schwartz et. al, 1992; Hecker et al., 2005). The findings of these studies are summarized below: Triangle G Ranch and Beebe Ranch Sites on the central RCF (Figure 1)—Schwartz et al. (1992) interpret two events at the Triangle G Ranch: (1) an oldest event that occurred shortly before 993 to 1193 AD (962–762 yr BP), and (2) a younger event that occurred after 993-1193 AD (962–762 yr BP) but before 1275–1413 AD (680 to 542 yr BP). Using the occurrence of non-native pollen and review of historical records, Hecker et al. (2005) constrain the timing of the MRE at this site to after A.D. 1715 and probably before A.D. 1776. Hecker et al. (2005) estimate the MRE right-laterally offset a buried paleochannel a minimum of 2.2 m (+1.2, -0.8). They also compare the timing of the RCF MRE to the timing of the last prehistoric ruptures on the Hayward fault (A.D. 1640–1776; 315–234 yr BP), to support an interpretation of a possible linked rupture of both faults across San Pablo Bay. Martinelli Ranch Site on the southern RCF (Figure 1)—RandolphLoar et al. (2004) performed a paleoseismic study at Martinelli Ranch that constrained a single surface rupturing between 1.0–0.79 ka and 10.7–9.6 ka. Donnell Ranch Site on the southern RCF (Figure 1)—Located on the central portion of the southern RCF, Randolph-Loar (2002) identified one event prior to 6.8 ka, two events between 6.8 and 3.7 ka, and three post 3.7 ka.

Tolay Marsh Site (Rams Gate Winery)

The Tolay Marsh site is located at the intersection of the southern RCF and the northern edge of San Pablo Bay (Figures 1 and 2). This site is situated along the San Pablo Bay in an area relatively undisturbed, and currently is isolated from the bay by narrow levees bordering slack-water sloughs that meander across the estuary and flow south to the bay. At the Tolay Marsh site, we:

Figure 3. Tolay Marsh seismic reflection profile (uninterpreted and interpreted on the left and right, respectively) (Givler et al., 2010). Topographic profile at the top of the lines illustrates the width of the fault zone with respect to topography. No vertical exaggeration in seismic line. Red lines are interpreted faults. Colors on interpreted line are increasing P-wave velocity. See Figure 2 for seismic line location.

(1) collected and interpreted seismic reflection data across the fault in collaboration with the US Geological Survey (USGS) (Figure 3); and (2) collected cores of shallow estuarine stratigraphy. We then used radiocarbon dating and diatom analyses to confirm the timing and evidence of rapid co-seismic submergence (Knudsen et al., 2002). Diatom analyses were completed by Eileen Hemphill-Halley (Humbolt State University).

At the interface with the marsh, the site is bounded by a northwest-southeast trending Tertiary volcanic ridge cut by multiple Holocene and Quaternary active fault strands (Figure 2; Wagner et al., 2011). The alignment of the fault strands with southwest-and northeast-facing fault scarps, linear channels, closed depressions, springs, and bedrock topographic highs suggest structural complexity and long-term deformation (Hart, 1992; Randolph-Loar et a1., 2004). Previous trenches at the site provide further documentation on differentiating Holocene strands from older Quaternary strands. Based on our seismic reflection profile across the fault, we interpret that the fault forms a broad, approximately 500- to 600-m-wide bedrock pop-up structure intersected by multiple shallow faults (Figure 3). Some of the faults imaged in the seismic line coincide with Quaternary and Holocene fault traces mapped at the ground surface (Givler et al., 2010). Unfortunately, the presence of the flood control levees to the south along strike prevented gouge core transects where the fault projects into the marsh (Figure 2).

Shallow gouge and vibratory cores data, performed in transects parallel to the USGS seismic line, were used to provide stratigraphic context and age control on the deformation across the primary fault strand(s) (Figure 4). Stratigraphic and paleontologic analyses of gouge and vibracore samples document a sequence of late Pleistocene to Holocene alluvial and late Holocene marsh sediments, such as marsh soils (peaty muds), quiet water deposits (mud), and algal mat-rich horizons overlying Pleistocene(?) coarse-grained overbank alluvium (Figure 4). On the northeastern side of the RCF and along gouge transects #1 and #2, the marsh stratigraphy shallows toward the ground surface and defines a northeast-facing subsurface escarpment coincident with present-day topography and pop-up structure (Figures 2 and 4). Distinct elevation differences between closely spaced gouge cores also allow for the interpretation of possible southwest-side-up displacement of late Holocene strata (Unit 2d in Figure 4). This interpretation is supported by present-day tectonic geomorphology (Figure 2), geophysically imaged subsurface offsets (Figure 3), and trench data to the northwest (see below).

Based on the correlations between paleoecological (diatom) stratigraphic data, at least one coseismically-related submergence event occurred at the site in the vicinity of the gouge core transects. Diatom and radiometric analyses of a buried peaty mud horizon (Unit 2d) indicate a rapid co-seismic submergence event that abruptly changed the diatom paleoecological

Figure 4. Gouge and vibracore transect #2 illustrating the shallow marsh stratigraphy along the northeast margin of Tolay marsh. Stratigraphic relationships and paleontological diatom analyses of fossils preserved in these buried deposits from viabracore VC2 indicate possible rapid coseismic submergence between Units 2d and 2e, which we tentatively correlate with a possible offset of unit 2d between GC29 and GC20 (Givler et al., 2010).

conditions after 2150–2350 cal yr BP and before 1900–2300 cal yr BP (Figure 4). Utilizing Bayesian models, we constrain the timing of this event between 2100 and 2300 yr BP (B.C. 346 and 152) at the 95.4% confidence limit. Following criteria defined in Knudsen et al. (2002), we interpret this as evidence of a large earthquake on the RCF.

Cline Vineyards

The Cline Vineyards site is located 1.3 km south of Highway 116 (Figure 1) on a Holocene fluvial terrace where the southern RCF juxtaposes two members of the Tertiary Sonoma Volcanics (Figure 5; Wagner et. al., 2002a; 2002b; Wagner et al., 2011). Near the Cline Vineyards site, fault traces associated with the RCF are defined by prominent tectonic-related geomorphology: northwest-trending topographic scarps, shutter ridges, deflected drainages, sag ponds, and aligned springs (Figure 3) (Hart, 1992; Randolph et al., 2004; Givler et al., 2019; Hecker et al., 2018). The results of our site surficial geologic mapping and subsequent paleoseismic trenching suggest the Cline Vineyards site is traversed by two RCF strands, an eastern and a western fault strand that displace late Holocene age fluvial terrace deposits (Qt1 and Qt2). Trench exposures (Figures 6 and 7) and 14C data confirm the Qt1 terrace surface is composed of early to late Holocene fluvial deposits separated by as many as five different buried soils, with the latest deposition on the Qt1 surface occurring after 680–760 cal yr B.P. (Unit 600). The multiple trenches excavated at the site provide direct evidence for late Holocene deformation that includes a combination of co-seismic folding and faulting distributed across the eastern and western strands (Figure 5).

Event timing at the Cline Vineyard site includes the following: (1) Event E1 occurred after A.D. 1220-1280 (670–740 cal yr BP) based on findings of trenches T-3 and T-2; (2) Events E2 and E3 occurred before A.D. 1040–1160 (800–910 cal yr BP) from Trench T-3; (3) Event E4 faulted gravels in trenches T-5, T-7, and T-8 sometime after approximately B.C. 1050 (3000 cal yr BP)—this event may correlate with events E1, E2, or E3 based on the limited deposition on the Qt1 surface in the late Holocene; and (4) several earlier events older than B.C. 2050 (4000 cal yr BP) are inferred based on the available data in trenches T-5, T-7 and T-8, but these events remain poorly constrained.

Rogers Creek Fault Event Correlation

To potentially correlate paleoseismic events interpreted at our two sites with past studies, we developed earthquake-timing information at the 95.4% confidence limit using Bayesian modeling software (OxCal - Bronk Ramsey, 2009; Figure 8).

Figure 5. Map of the Cline Vineyard Site illustrating locations of paleoseismic trenches and key geologic units at the site. Abbreviations: Qt1 and Qt2- Quaternary terrace 1 and 2, Qfh – Quaternary Holocene fan, Qc – Quaternary colluvium, Tsv-Tertiary Sonoma Volcanics (shown in grey). The western strand of the RCF is mapped along exposures within the drainage trending northwest and crossed by trenches T-2, T-3, T-4, T-5a, and T-6 before stepping to the right to the eastern strand intersected by trenches T-1, T-5b, T-2 and T-8.

Figure 6. Photomosaic (above) and interpreted log (below) of the north wall of trench T-3. Key units include Sonoma Volcanics bedrock (Unit 10), Holocene fluvial deposits (Units 110 to 500), Holocene colluvium (Unit 150 and 350), late Holocene fluvial gravel (Unit 550) and Units 600/700 represent the most recent soils and overbank deposits. See Givler et al. (2019) for detailed unit descriptions.

Figure 7. Interpreted log of the north wall of trench T-2. Key units include Sonoma Volcanics bedrock (Unit 1), Highly weathered bedrock (Unit 2), colluvium (Units 4, 5, 6, 11, and 12), Holocene fluvial deposits (Units 7, 8, 9, 10), Holocene colluvium (Unit 150 and 350), late Holocene fluvial gravel. See Givler et al. (2016) for detailed unit descriptions. Blue line represents the interpreted event horizon for the MRE (E1).

Figure 8. Summary of earthquake-timing data for the Cline and Tolay Marsh Sites. Earthquake PDFs generated using OxCal Bayesian modeling software (Bronk Ramsey, 2009). Alternative model for the Cline Site E1 is shown as a dotted line. Triangle G earthquake-timing data (Schwartz et al., 1992; Hecker et al. 2005) shown at the top of the figure for events E1, E2, and E3.

Although the constraints on event timing are broad at the Cline Vineyard and Tolay Marsh sites, the events correlate with other paleoseismic events identified on the RCF to the north at the Triangle G Site (Figure 1). For example, the timing of Cline Vineyard E1 (after A.D. 1220–1280) is broadly consistent with the Triangle G E1 (A.D. 1715–1776) identified by Hecker et al. (2005). Alternatively, it is possible the youngest faulting predates the surface soil in Cline Vineyard trench T-3 (Unit 600 in Figure 6;) which places the MRE between A.D. 1080 and 1260 (95.4% confidence). In the alternative interpretation, the MRE (E1) and Event E2 at the Cline Vineyard Site occurred before A.D. 1040–1160, which overlaps with the penultimate event

(E2) (A.D. 990–1410) at Triangle G Ranch (Hecker et al., 2005). Tolay Marsh Event E1 timing (B.C. 346–152, 95.4% confidence interval) also overlaps broadly with the timing constraints for the Cline Vineyard E2 and Triangle G E3 (before A.D. 900). Thus, our results generally support co-seismic surface rupture of the southern RCF with the RCF to the north (although these events can also be interpreted as separate events). This event chronology also overlaps with the northern Hayward fault to the south (Hecker et al., 2005) and supports the interpretation of co-seismic ruptures across San Pablo Bay between the two faults (Figure 1; Watt et al., 2016). These joint rupture scenarios are supported by dynamic rupture modeling (Harris et al., 2021), which favors nucleation on the “locked” southern RCF and propagation onto creeping sections of the Hayward fault to the south in larger events (M7+).

Summary

Herein we summarize paleoseismic research at two separate sites along the southern RCF (Figure 1). The available paleoseismic data from previous studies and this research indicate that only two to three events have occurred along the fault in the last approximately one to two thousand years (Figure 7). It is unclear if this paucity of earthquake timing information is due to the apparent poor preservation of paleo-earthquakes, due to erosion (removal of evidence), or very low deposition for the sites currently evaluated, or from the overall fault behavior for the RCF. By comparison, the northern Hayward fault has a much more robust earthquake chronology with as many as nine events occurring between ~B.C. 400 and A.D. 1800 (HPEG, 1999). If there is a connectivity between the faults (e.g., Watt et al., 2016), it is possible that the RCF ruptures less frequently than the Hayward fault to the south, or more likely the existing late Holocene record on the RCF is very incomplete. More work is required at high-sedimentation sites to further constrain event timing and recurrence intervals along the RCF. For additional details regarding this research see Givler et al. (2010; 2019).

Acknowledgements

Thanks to our many co-authors and collaborators (Michael Rymer; Rufus Catchings, Robert Sickler, Keith Knudsen, Nora Lewandowski, Eileen Hemphill-Halley, Josh Goodman, and Matt Huebner) and the landowners of the Tolay Marsh and Cline Vineyard sites.

References

Blisniuk, K., and Walker, A., 2018, Determining the distribution of slip across the northern San Andreas fault system: through long-term fault slip rates: US Geological Survey NEHRP Final Technical Report Grant number. Budding, K.E., Schwartz, D.P., and Oppenheimer, D.H., 1991, Slip rate, earthquake recurrence, and seismogenic potential of the Rodgers Creek fault zone, northern California: Initial results: Geophysical Research Letters, v. 18, no. 3, p. 447–450. Field, E.H., Biasi, G.P., Bird, P., Dawson, T.E., Felzer, K.R., Jackson, D.D., Johnson, K.M., Jordan T.H., Madden, C., Michael, A.J., Milner, K.R., Page, M.T., Parsons, T., Powers, P.M., Shaw, B.E., Thatcher, W.R., Weldon, R.J., Seng, Y., 2015, Long-term time-dependent probabilities for the Third Uniform California Earthquake Rupture Forecast (UCERF3), Bulletin of the

Seismological Society of America, v. 105, no. 2a, p. 511–543.

Floyd, M .A., Funning, G .J., Lipovsky, B ., 2009, Geodetic evidence for creep along the Rodgers Creek and Maacama fault zones, northern California: AGU Fall Meeting, abstract #G23B-0692. Funning, G.J., Burgmann, R., Ferretti, A., Novali, F., Fumagalli, A., 2007, Creep on the Rodgers Creek fault, northern San Francisco Area from a 10-year PS-InSAR dataset: Geophysical Research Letters, v. 34, L19306, doi:10.1029/2007GL030836 Givler, R.W., Baldwin, J., Knudsen, K., 2010, Structural characterization and pilot paleoseismic investigation of the southernmost Rodgers Creek fault, northern San Pablo Bay, California: Final Technical Report for USGS NEHRP Award No. G09AP00022. Givler, R.W., and Baldwin, J., 2016, Pilot paleoseismic investigation of the southernmost Rodgers Creek fault, northern San Pablo Bay, California: Final Technical Report for USGS NEHRP Award No. G13AP00036. Givler, R.W., Baldwin, J.N., Bloszies, C., 2019, Paleoseismic investigation of the southernmost Rodgers Creek fault at Cline Vineyards Site, Sonoma, California: Final Technical Report for USGS NEHRP Award No. G17AP00102. Harris, R. A., Barall, M., Lockner, D. A., Moore, D. E., Ponce, D. A., Graymer, R.W., et al., 2021, A geology and geodesy based model of dynamic earthquake rupture on the Rodgers Creek-Hayward- Calaveras fault system, California: Journal of Geophysical Research: Solid Earth, 126, e2020JB020577. https://doi.org/10.1029/2020JB020577. Hayward fault Paleoearthquake Group (HFPG), 1999, Timing of paleoearthquakes on the Northern Hayward fault—preliminary evidence in El Cerrito, California: US Geological Survey Open-File Report 99–318. Hecker, S., and Randolph Loar, C.E., 2018, Map of recently active traces of the Rodgers Creek fault, Sonoma County, California: US Geological Survey Investigations Map 3410, 7 p., 1 sheet, https://doi.org/10.3133/sim3410. Hecker, S., Pantosti, D.P., Schwartz, D.P., Hamilton, J.C., Reidy, L.M., Powers, T.J., 2005, The most recent large earthquake on the Rodgers Creek fault, San Francisco Bay area: Bulletin of the Seismological Society of America, Vol. 95, No. 3, pp. 844–860, June 2005. Knudsen, K. L., R. C. Witter, C. E. Garrison-Laney, J. N. Baldwin, G. A. Carver, L.B. Grant, and W.R. Lettis, 2002, Past earthquake-induced rapid subsidence along the northern San Andreas fault: a paleoseismological method for investigating strike-slip faults: Bulletin of the Seismological Society of America, V. 92, No. 7, pp. 2612-2636. Lienkaemper, J.J., McFarland, F.S., Simpson, R.W., Caskey, S.J., 2014, Using surface creep rate to infer fraction locked for sections of the San Andreas fault system in northern California from alignment array and GPS data:

Bulletin of the Seismological Society of America, V. 104, no. 6, p. 1–21. McFarland, F. S., Lienkaemper, J.J., Caskey, S .J., Grove, K., 2016, Data from theodolite measurements of creep rates on the San Francisco Bay region faults, California: 1979 -2007, US Geological Survey Open-file report 20071367, 14 p. (updated version 1.8 March 2016). McLaughlin, R.J., Sarna-Wojcicki, A.M., Wagner, D.L., Fleck, R.J., Langenheim, V.E., Jachens, R.C., Clahan, K., Allen, J. R., 2012, Evolution of the Rodgers Creek–Maacama right-lateral fault system and associated basins east of the northward-migrating Mendocino Triple Junction, northern California:

Geosphere, v. 8; no. 2; p. 1-32. Randolph-Loar, C.E., 2002, Neotectonics of the southern Rodgers Creek fault, Sonoma County, California: Masters Thesis, San Francisco State University, 154 p. Randolph-Loar, C. E., Witter, R. C., and Lettis, W. R., 2004, Paleoseismic investigation of the southern Rodgers Creek fault, Martinelli Ranch, Sonoma County, CA: Final Technical Report for the US Geological National Earthquake Hazards Program, 14 p. Schwartz, D.P., Lienkaemper, J.J., Hecker, S., Kelson, K.I., Fumal, T.E., Baldwin, J.N., Seitz, G.G., Niemi, T.M., 2014, The earthquake cycle in the San Francisco Bay region: A.D. 1600-2012: Bulletin of the Seismological

Society of America, v., 104, no. 3, p. 1-30. Schwartz, D.P., Pantosti, D., Hecker, S., Okumura, K., Budding, K.E., and Powers, T., 1992, Late Holocene behavior and seismogenic potential of the Rodgers Creek fault zone, Sonoma County, California: Borchardt, G., et al., eds., Proceedings of the Second Conference on Earthquake Hazards in the Eastern San Francisco Bay Area: California Division of Mines and Geology Special Publication 113, p. 393-398.

Toppozada, T.R., Bennett, J.H., Hallstrom, C.L., and Youngs, L.G., 1992, 1898 “Mare Island” earthquake at the southern end of the Rodgers Creek fault: Borchardt, G., et al., Proceedings of the Second Conference on Earthquake Hazards in the Eastern San Francisco Bay Area: California Division of Mines and Geology Special Publication 113, p. 385-392. Wagner, D.L., Saucedo, G.J., Clahan, K.B., Fleck, R.J., Langenheim, V.E., McLaughlin, R.J., Sarna-Wojcicki, A.M., Allen, J.R., Deino, A.L., 2011, Geology, geochronology, and paleogeography of the southern Sonoma volcanic field and adjacent areas, northern San Francisco Bay region, California: Geosphere, v. 7, no 3, p. 658-683. Watt, J., Ponce, D., Parsons, T., Hart, P., 2016, Missing link between the Hayward and Rodgers Creek faults: Science Advances, V. 2, e1601441. Wong, I.G., 1991, Contemporary seismicity, active faulting and seismic hazards of the Coast Ranges between San Francisco Bay and Healdsburg, California: Journal of Geophysical Research, v. 96, no. B12, p.19, 891-19,904.

About the Authors

Robert Givler is a certified engineering geologist (CEG) at Lettis Consultants International, Inc. with over 20 years of professional experience conducting geologic and geotechnical investigations focused on geologic hazard evaluations throughout the California and the United States. He has recently completed paleoseismic studies on the Rodgers Creek Fault (described herein) and the Pajarito Fault System in northern New Mexico. John Baldwin is a CEG at Lettis Consultants International, Inc. with over 25 years of experience in conducting seismic hazard assessments and performing paleoseismic research in the United States and internationally. He currently is conducting

L to R: Authors Chris Blozies, Robert Givler, and John Baldwin

earthquake-related research on the San Andreas and Quien Sabe faults in northern California. Chris Bloszies is a Quaternary geomorphologist at Lettis Consultants International, Inc. with over nine years of experience in surficial geologic mapping and paleoseismic investigation. His interests lie in understanding how tectonic and/or climatological perturbations influence landscape evolution, hydrologic systems, and soil formation. His recent earthquake-related research includes faults in Northern California, the New Madrid Seismic Zone, and Taiwan, and paleoclimate studies in East Africa and in mid-continental United States.

The workshop is co-sponsored by AEG and USSD and will be two days of presentations followed by an optional field trip to Herbert Hoover Dike. Presentations will cover geotechnical investigations and dam modifications best practices and risk management, followed by a round table discussion with all the speakers. The goals of the Workshop are to:

l Highlight best practices in geotechnical exploration and construction in dams and levees l Present case histories with a range of exploration and construction techniques l Discuss issue resolution and decision making in a risk framework Registration is limited to the first 100 people, so sign up now! www.aegmeetings.org

The 1811–12 New Madrid Earthquakes – The Backstory

Phyllis Steckel

eginning on December 16, 1811 and continuing through February 1812, at least four large earthquakes occurred in what is now northeast B

Arkansas and southeast Missouri. Each of these four large earthquakes has since been estimated to have been about magnitude 7½ or larger. They occurred on December 16, 1811; and January 23, February 7, and February 23, 1812.

Many strong aftershocks continued for months, finally tapering off after several years. The New Madrid (MAD-rid—accent on the first syllable) fault system has since been recognized as a complex, buried fault zone that has been intermittently active for millions of years.

But the risk it poses today to the people, property, infrastructure, and economy that has been added to the region since the early 1800s is difficult to understand and mitigate.

OZARK PLATEAU

St. Francois Mountains ILLINOIS BASIN

The Regional Setting

To the northwest of New Madrid are the St. Francois Mountains and the Ozark Plateau (Figure 1). In the St. Francois Mountains region, Precambrian basement rocks are locally exposed at the surface, while Paleozoic strata underlie the surrounding regions of the Ozark Plateau physiographic province. The southern part of the Illinois Basin is located northeast of New Madrid. The Mississippi River and its floodplain flow from the north, a rough boundary between the Ozark Plateau to the west and the Illinois Basin to the east. The river empties into the Mississippi Embayment, a bedrock trough that deepens and widens from the northeast to the southwest. From about the junction of the Mississippi and Ohio rivers and to the south-southwest for about 300 miles (380 km), the axis of the Mississippi Embayment is roughly parallel with today’s Mississippi River.

Over the last tens of millions of years, the Mississippi Embayment has filled with sediment. Most of this sediment was scraped off the northern and central parts of the North American continent by many glacial pulses, transported, and neatly sorted by the Mississippi River and its Missouri and Ohio tributaries system. Reworked glacial and alluvial material carried south by these rivers has built up the surface of the Mississippi Alluvial Plain section of the Gulf Coastal Plain. Glacial sediments from many hundreds of miles to the north were deposited into the Mississippi Embayment, creating a nearly-flat topography all the way to the Gulf of Mexico. These layers of sediments, several thousands of feet thick, amplify and prolong ground-shaking and expand the area affected by New Madrid earthquakes. Compared to earthquakes of similar magnitude in California, a New Madrid earthquake will affect four to six times as much area.

Mississippi Embayment

Figure 1. Today, the New Madrid Seismic Zone includes southeast Missouri, northeast Arkansas, southern Illinois, and westernmost Kentucky and Tennessee. It includes three main faults, all of which probably were displaced during the 1811–12 series of earthquakes. Today, this epicentral area includes the second-largest steel production county in the US, one of the last aluminum smelters in the US, a major electric-power plant, the largest single consumer of electricity in Missouri, critical petroleum and natural gas pipelines that supply the northeastern US, the critical infrastructure of river-barge traffic on the Mississippi and Ohio rivers, and engineered drainage systems that are especially vulnerable to earthquake hazards. Other vulnerable industries include warehousing and overnight shipping in Memphis and Louisville, critical river port facilities, commodity transportation and storage, and multi-modal (rail, barge, and truck) hubs.

FIGURE FROM USGS.

The Local Setting

In the early 1800s, much of the northern part of the Mississippi Embayment was swampland at the surface. Flooding could occur at any time of year, fed by runoff from the Mississippi, Missouri, and Ohio river watersheds—nearly half of today’s continental United States. The entire Mississippi Embayment was thickly forested and essentially impenetrable for wagons and livestock at that time. As a result, keelboats, flatboats, rafts, and canoes were transportation necessities. The few settlements that were established were on the banks of the Mississippi River. It was the interstate highway of the 1800s.

The town of New Madrid, established in 1789, is located within the Mississippi Embayment, on a sliver of slightly higher

Figure 2. Instrumentally located seismicity 1974 to June 2021 (source CERI online catalog) highlights the active part of the New Madrid Seismic Zone (black oval).

FIGURE FROM MITCHELL M WITHERS, CENTER FOR EARTHQUAKE RESEARCH AND INFORMATION (CERI), UNIVERSITY OF MEMPHIS.

ground. This sliver is the southernmost part of the Sikeston Ridge, the “crest” of which is only about 20-feet above the normal level of the Mississippi River. In those days, the Mississippi River level varied widely from raging flood to nearly dry, depending on weather and other conditions upstream. Thick forests, which were sometimes flooded and usually wet, surrounded the townsite. Clearings were limited to the townsite and small garden plots tended by the various households. Trees were plentiful, and logs from local clearing provided construction material.

The Town of New Madrid, Louisiana Territory, in 1811

In 1811, the town of New Madrid was a loose community of about 1,000 people. Great Britain, France, Spain, and the fledgling United States all had laid claim to the area at various times within the previous 20 years or so. The United States had just purchased the Louisiana Territory from France in 1803. Before then, previous governments were allegiant to England, and even earlier, to Spain. (The State of Missouri would not exist for another nine years or so.)

In the early 1800s, the residents of New Madrid reflected that diversity. Depending on who was speaking, local conversations were in French, Spanish, English, native languages, or a muddle of two or more of these. The government du jour, political allegiance, and administrative authority were much less of a concern to New Madrid residents than an outbreak of illness, a serious broken bone, or a bear in the garden. Life was difficult, the location was remote, resources were limited. Life-or-death challenges occurred almost daily. Government and impractical rules were just not all that important to the locals.

Even so, New Madrid was an important and busy place. At that time, it was the largest and most important settlement along the Mississippi River between St. Louis and New Orleans. Memphis would not even be founded for another eight or ten

years. Cape Girardeau and Ste. Genevieve (now both in Missouri) were neither as large nor as important at that time. River traffic stopped in New Madrid to rest their crews, make repairs to their vessels, get news of river conditions ahead, deliver mail and months-old newspapers, and resupply provisions.

Several known persons were in New Madrid at the time the earthquakes began on December 16, 1811. Eliza Bryan was about 31 years old and ran a boarding house or public lodging in New Madrid. To date, no record has been found of Eliza’s husband, so she probably was a widow. Her son, Fred Bryan, was about six years old. A newcomer to New Madrid was William Bratton, a veteran of Lewis and Clark’s Corps of Discovery venture in 1804–06. He had made the overland trek from St. Louis to the Pacific Ocean and back.

The Earthquakes

Sometime around 2:00 am on the morning of December 16, 1811, a large earthquake occurred on the New Madrid fault system. Neither the exact time nor the epicenter location is known. In the New Madrid area, the weather was unseasonably warm, muggy, and still: it might have been in the 40s or maybe even the low 50s (degrees Fahrenheit). The shaking was intense and long-lasting, causing significant damage or destroying every structure and household in New Madrid. An acrid sulfur smell was released into the air—probably from decaying organics that were disturbed in the surrounding swamplands.

In the area of New Madrid and much of the Mississippi Embayment, fountains of sandy water—sand blows or liquefaction features—spurted up from beneath the ground surface. Some continued to build up for hours to days after the main ground shaking stopped. These sand blows actively built up the immediate area around each vent with a deposit of fresh sand a few inches to a few feet thick. Linear crevasses opened in places that, before the earthquake, had been a flat and continuous surface. The banks of the Mississippi River had slumped into the channel from both sides of the river, dragging large trees and other debris with it. At least one person died in New Madrid, a probable heart attack. But dozens to perhaps hundreds more were probably lost; river travelers would simply disappear, and those of African and Native American ancestry were specifically not accounted.

One of the most puzzling effects of the earthquake at New Madrid was that the Mississippi River quickly ran dry. The flow of water in the river stopped and completely drained the channel area. This interruption of the river lasted only a short time—perhaps only a few hours or so. Then it slowly recovered its murky flow and again began to drain along its previous course. This phenomenon has become an urban myth and has been repeated incorrectly as the “river ran backwards.” (Incorrect, but hard to remove from urban legend.) This was probably from Eliza Bryan’s description of “…the current of which [the Mississippi River] ran retrograde….”

In reality, a surface fault rupture running nearly north-south was located just east and south of New Madrid. The west side of this surface rupture was uplifted, perhaps 30 to 50 feet. As a result, the course of the Mississippi River here—a distinct loop in the river known as “Kentucky Bend”—had been bisected by the river. This created a natural dam made of the loose sediments of the river’s natural levee. This natural dam briefly interrupted the river’s flow, causing water from upstream to divert to the southeast and pond in an area now known as Reelfoot Lake. Because the uplifted material was easily eroded, the river soon cut through the soft mud and quickly returned to its former channel.

The New Madrid earthquakes were felt in southern Canada, Massachusetts, South Carolina, and Georgia. Several of the larger earthquakes were felt in the White House in Washington DC by then-President James Madison. Literally thousands of aftershocks were recorded in Louisville, Kentucky by Dr. Jared Brooks. Historians have gleaned felt-reports from newspapers, personal journals, insurance logs, and many other sources. It was felt strongly at Fort Osage (near present-day Kansas City); in Savannah, Georgia; Charleston, South Carolina; and many towns along the Ohio and Lower Mississippi valleys. Felt-reports from the west and southwest have not yet been found. Very few literate English- or French-speakers lived in the region at that time, except for a few French-speaking Jesuit missionaries and Spanish-speaking traders traveling the Santa Fe Trail.

In 1811, Abraham Lincoln was almost three years old, and he and his family lived only about 200 miles from New Madrid. They most certainly felt the earthquakes, but specific information on the impact to their household has not yet been found. A few months later, the Territorial Governor William Clark, who, with Meriwether Lewis, had led the Corps of Discovery to the Pacific Ocean and back in 1804-1806, led a successful effort to have Congress pass “A Resolution for the Relief of the Inhabitants of the County of New Madrid”—the first federal disaster funding bill.

The New Madrid earthquake series was just that: a series of earthquakes. At least four earthquakes of about magnitude 7½ occurred in a little over two months. The ground-shaking, liquefaction, surface fault rupture, and landslides would repeat after each large event. Aftershocks began immediately and continued incessantly for many months. New Madrid was destroyed, and almost all residents left the area as soon as possible. Travel on the Mississippi River was difficult to impossible.

After the Earthquakes

Several years later, the Reverend Lorenzo Dow, urged Eliza Bryan to write a descriptive account of what she saw and experienced during that time. In 1816, Eliza Bryan, a gifted writer, had completed her letter to Lorenzo Dow www.memphis.edu/ceri/compendium/pdfs/bryan.pdf

A quick look at Eliza Bryan’s account suggests that she was wildly imaginative, if not fantastical, with her descriptions. However, a closer review of her writing proves her observations to be amazingly accurate. Numerous still-visible earthquake scars in the New Madrid region verify Eliza’s various descriptions that were written more than 200 years ago. Other plausible and proven geologic and physical explanations support her observations. Eliza Bryan was a gifted writer and keen observer. Without her meticulous account, the New Madrid story would not be what it is today.

Earthquake Insight Field Trip

Since 2005, the annual Earthquake Insight Field Trip has traveled a route along the Mississippi River from the St. Louis area, south into the Mississippi Embayment. Stops are made at historical sites that tell the story of the 18111812 New Madrid earthquakes. The field trip includes a look at earthquake hazards and earthquake risks to the people, infrastructure, and economy of the region.

The Earthquake Insight Field Trips were originally sponsored by the US Geological Survey. More recently, the Missouri Seismic Safety Commission and the Missouri State Emergency Management Agency are sponsors. Earthquake Insight Field Trips are led by the author, Phyllis Steckel, RG.

Due to Covid-19, Earthquake Insight Field Trips were not held in 2020. However, one is tentatively planned for the fall of 2021. Contact Phyllis Steckel at psteckel@charter.net for more information.

CEUs and/or PDHs are available from the University of Missouri – Columbia.

Eliza Bryan and her son, Fred Bryan, lived most of the rest of their lives in New Madrid. Fred died in 1865 and Eliza died in 1866. They are buried next to each other in the old New Madrid cemetery, near the corner of St. Isadore and St. Paul drives in New Madrid. Unfortunately, the names and experiences of the hundreds of other New Madrid residents in 1811–12 have been lost to history—or, at least, remain unfound to date. Most who felt these earthquakes were not able to read nor write in any language. But if they could, their recollections may have been in Spanish or French, or even in Native American records, such as winter counts—potential information sources that have yet to be sought out.

The Town of New Madrid, Missouri, in 2021

Today, New Madrid is home to about 3,000 people and the county seat of New Madrid County. Although some of the street names have changed over the years, many of the streets in the oldest part of town are still on the same grid pattern as when they were laid out in the late 1700s. At New Madrid, the Mississippi River flows along the south edge of town, from east to west, across the northern arc of the Kentucky Bend. Being on the outside of this river meander has exposed the townsite to continuous riverbank erosion over more than 200 years. As a result, several streets shown on the earliest maps of New Madrid that were closest to the Mississippi River have been washed away during the many floods that affected the area.

Almost all of the land in New Madrid County is now under intense corporate agricultural production, mostly soybeans, cotton, corn, and commercial vegetable crops. The swampy areas around the original townsite have all been drained and cleared by the Little River Drainage District, an ambitious, private, engineered-drainage project installed in southeast Missouri in the early 1900s. Paid for by the landowners within its project area, its installation actually required moving more material than the Panama Canal and employed some of the same engineers and workers that had just finished up on that project in Central America. Now, there are several large industrial facilities, mostly agriculture-related. One of the United States’ largest and last aluminum smelters is nearby. Now the town of New Madrid is an exit off Interstate 55, and travel that once took days can be completed in less than an hour.

The New Madrid earthquake series of 1811–12 is usually not much more than a footnote in most significant conversations, plans, and decisions made in the central United States. The local museum features a small exhibit about the earthquakes, and the telephone exchange building in New Madrid has been retrofit with an exterior steel frame. Local public schools hold earthquake drills annually for students and staff, and the local first-responders (mostly volunteers) receive special earthquake-response training. But overall, the hazard from the New Madrid fault is often given minor and token consideration by elected leaders and industry decision-makers in remote locations. In 2019, earthquake insurance premiums increased by 700–1,000 percent, and the number of residences having earthquake insurance has decreased from more than 60 percent to less than 14 percent in southeast Missouri. The State of Missouri still has no required building code in southeast Missouri; as a result, almost no structures have been built to any seismic-design standard.

The land surface on both sides of the Mississippi River has been engineered for intense agriculture; now, any minor perturbation from an earthquake will permanently flood some dry areas and interrupt the precarious engineered drainages.

This story of both the geology and the history of the New Madrid earthquakes of 1811–12 is a call to action. Over a million people now live in this epicentral area, including Memphis. Tens of millions now live in the distant areas that were also affected by the earthquakes—St. Louis, Cincinnati, New Orleans, Little Rock, Louisville, and more. When the next significant earthquake, or series of earthquakes, from the New Madrid fault occurs, we can expect to see every problem seen by Eliza Bryan, or President James Madison, or Dr. Jared Brooks who documented what they saw or felt or experienced in 1811–12. Our losses will be larger and our recoveries will take longer. We have much more to lose.

About the Author

Phyllis Steckel, RG, Earthquake Insight LLC, Washington, Missouri, has been an active member of the AEG St. Louis Chapter for about 20 years. Currently, she serves as the AEG Region 7 Director, and the lead author of the Geology of the Cities of the World – St. Louis, Missouri USA. Mostly retired now, she has focused her work on geologic hazards in the central US. Previously, she was involved with the AEG San Francisco Section.