Walnut Creek, CA, United States
Walnut Creek, CA, United States

Time filter

Source Type

Graehl N.A.,Lettis Consultants International Inc. | Kelsey H.M.,Humboldt State University | Witter R.C.,U.S. Geological Survey | Hemphill-Haley E.,Humboldt State University | Engelhart S.E.,University of Rhode Island
Bulletin of the Geological Society of America | Year: 2015

The Sallys Bend swamp and marsh area on the central Oregon coast onshore of the Cascadia subduction zone contains a sequence of buried coastal wetland soils that extends back ~4500 yr B.P. The upper 10 of the 12 soils are represented in multiple cores. Each soil is abruptly overlain by a sandy deposit and then, in most cases, by greater than 10 cm of mud. For eight of the 10 buried soils, times of soil burial are constrained through radiocarbon ages on fi ne, delicate detritus from the top of the buried soil; for two of the buried soils, diatom and foraminifera data constrain paleoenvironment at the time of soil burial. We infer that each buried soil represents a Cascadia subduction zone earthquake because the soils are laterally extensive and abruptly overlain by sandy deposits and mud. Preservation of coseismically buried soils occurred from 4500 yr ago until ~500- 600 yr ago, after which preservation was compromised by cessation of gradual relative sea-level rise, which in turn precluded drowning of marsh soils during instances of coseismic subsidence. Based on grain-size and microfossil data, sandy deposits overlying buried soils accumulated immediately after a subduction zone earthquake, during tsunami incursion into Sallys Bend. The possibility that the sandy deposits were sourced directly from landslides triggered upstream in the Yaquina River basin by seismic shaking was discounted based on sedimentologic, microfossil, and depositional site characteristics of the sandy deposits, which were inconsistent with a fl uvial origin. Biostratigraphic analyses of sediment above two buried soils-in the case of two earthquakes, one occurring shortly after 1541-1708 cal. yr B.P. and the other occurring shortly after 3227-3444 cal. yr B.P.-provide estimates that coseismic subsidence was a minimum of 0.4 m. The average recurrence interval of subduction zone earthquakes is 420-580 yr, based on an ~3750-4050-yr-long record and seven to nine interearthquake intervals. The comparison of the Yaquina Bay earthquake record to similar records at other Cascadia coastal sites helps to defi ne potential patterns of rupture for different earthquakes, although inherent uncertainty in dating precludes defi nitive statements about rupture length during earthquakes. We infer that in the fi rst half of the last millennia, the northern Oregon part of the subduction zone had a different rupture history than the southern Oregon part of the subduction zone, and we also infer that at ca. 1.6 ka, two earthquakes closely spaced in time together ruptured a length of the megathrust that extends at least from southwestern Washington to southern Oregon. © 2014 Geological Society of America.


Kwong N.S.,University of California at Berkeley | Chopra A.K.,University of California at Berkeley | Mcguire R.K.,Lettis Consultants International Inc.
Earthquake Engineering and Structural Dynamics | Year: 2015

In this short communication, we respond to the comments made by Dr Brendon A. Bradley and provide additional context to our paper under discussion. © 2015 John Wiley & Sons, Ltd.


Kwong N.S.,University of California at Berkeley | Chopra A.K.,University of California at Berkeley | Mcguire R.K.,Lettis Consultants International Inc.
Earthquake Engineering and Structural Dynamics | Year: 2015

This study presents a novel approach for evaluating ground motion selection and modification (GMSM) procedures in the context of probabilistic seismic demand analysis. In essence, synthetic ground motions are employed to derive the benchmark seismic demand hazard curve (SDHC), for any structure and response quantity of interest, and to establish the causal relationship between a GMSM procedure and the bias in its resulting estimate of the SDHC. An example is presented to illustrate how GMSM procedures may be evaluated using synthetic motions. To demonstrate the robustness of the proposed approach, two significantly different stochastic models for simulating ground motions are considered. By quantifying the bias in any estimate of the SDHC, the proposed approach enables the analyst to rank GMSM procedures in their ability to accurately estimate the SDHC, examine the sufficiency of intensity measures employed in ground motion selection, and assess the significance of the conditioning intensity measure in probabilistic seismic demand analysis. © 2015 John Wiley & Sons, Ltd.


Kwong N.S.,University of California at Berkeley | Chopra A.K.,University of California at Berkeley | Mcguire R.K.,Lettis Consultants International Inc.
Earthquake Engineering and Structural Dynamics | Year: 2015

This study develops a framework to evaluate ground motion selection and modification (GMSM) procedures. The context is probabilistic seismic demand analysis, where response history analyses of a given structure, using ground motions determined by a GMSM procedure, are performed in order to estimate the seismic demand hazard curve (SDHC) for the structure at a given site. Currently, a GMSM procedure is evaluated in this context by comparing several resulting estimates of the SDHC, each derived from a different definition of the conditioning intensity measure (IM). Using a simple case study, we demonstrate that conclusions from such an approach are not always definitive; therefore, an alternative approach is desirable. In the alternative proposed herein, all estimates of the SDHC from GMSM procedures are compared against a benchmark SDHC, under a common set of ground motion information. This benchmark SDHC is determined by incorporating a prediction model for the seismic demand into the probabilistic seismic hazard analysis calculations. To develop an understanding of why one GMSM procedure may provide more accurate estimates of the SDHC than another procedure, we identify the role of 'IM sufficiency' in the relationship between (i) bias in the SDHC estimate and (ii) 'hazard consistency' of the corresponding ground motions obtained from a GMSM procedure. Finally, we provide examples of how misleading conclusions may potentially be obtained from erroneous implementations of the proposed framework. © 2014 John Wiley & Sons, Ltd.


Kwong N.S.,University of California at Berkeley | Chopra A.K.,University of California at Berkeley | Mcguire R.K.,Lettis Consultants International Inc.
Earthquake Engineering and Structural Dynamics | Year: 2015

This paper develops a procedure to select unscaled ground motions for estimating seismic demand hazard curves (SDHCs) in performance-based earthquake engineering. Currently, SDHCs are estimated from a probabilistic seismic demand analysis, where several ensembles of ground motions are selected and scaled to a user-specified scalar conditioning intensity measure (IM). In contrast, the procedure developed herein provides a way to select a single ensemble of unscaled ground motions for estimating the SDHC. In the context of unscaled motions, the proposed procedure requires three inputs: (i) database of unscaled ground motions, (ii) IM, the vector of IMs for selecting ground motions, and (iii) sample size, n; in the context of scaled motions, two additional inputs are needed: (i) a maximum acceptable scale factor, SFmax, and (ii) a target fraction of scaled ground motions, γ. Using a recently developed approach for evaluating ground motion selection and modification procedures, the proposed procedure is evaluated for a variety of inputs and is demonstrated to provide accurate estimates of the SDHC when the vector of IMs chosen to select ground motions is sufficient for the response quantity of interest. © 2015John Wiley & Sons, Ltd.


Seifried A.E.,Lettis Consultants International Inc. | Baker J.W.,Stanford University
NCEE 2014 - 10th U.S. National Conference on Earthquake Engineering: Frontiers of Earthquake Engineering | Year: 2014

Conditional spectral dispersion (CSD) is a measure of response spectrum variability that implicitly characterizes the variety of spectral shapes within a suite of ground motions. It is used here to explain the discrepancy between median structural demands estimated from different suites of scaled ground motions as well as those that have been spectrum-matched. Performing response history analyses with spectrum-matched ground motions is known to result in unconservatively biased median demand estimates in some cases. Herein, several suites of scaled ground motions with equivalent median intensities and varying levels of CSD are selected. A single suite of spectrum-matched ground motions is also created. These records are used to analyze inelastic single-degree-of-freedom and multiple-degree-of-freedom structural systems. A consistent trend among responses fully attributes the bias phenomenon to the asymmetric relationship between conditional spectral ordinates at periods affecting inelastic behavior and the resulting inelastic response, suggesting that no further explanation for the bias is needed.


Amos C.B.,Western Washington University | Lutz A.T.,Lettis Consultants International Inc. | Jayko A.S.,U.S. Geological Survey | Mahan S.A.,U.S. Geological Survey | And 2 more authors.
Bulletin of the Seismological Society of America | Year: 2013

Recent upward revision of the 1872 Owens Valley earthquake from Mw 7.4-7.5 to 7.7-7.9 implies either additional unrecognized rupture length or anomalously strong ground motions associated with this event.We investigate the first possibility through paleoseismic trenching south of the mapped surface rupture in the Haiwee area, where historical accounts suggest significant surface deformation following the earthquake. Trenching focused on a prominent north-striking scarp, herein termed the Sage Flat fault, expressed in Pleistocene alluvial fans east of Haiwee Reservoir. Surficial mapping and ground-based Light Detection and Ranging (lidar) surveying suggest that this fault accommodates east-down normalmotion, and possibly a comparable amount of dextral slip.Trenching and luminescence dating brackets the timing of the most recent surfacerupturing earthquake between ~25.7 and 30.1 ka, and provides evidence for an earlier event predating this time. In combination with scarp profiling, these dates also suggest a maximum rate of normal, dip-slip fault motion up to ~0.1 mm=yr over this period. Although we discovered no evidence for recent surface rupture on the Sage Flat fault, a series of subvertical fractures and fissures cut across young trench stratigraphy, consistent with secondary deformation associated with seismic shaking. As such,we suggest that possible ground disturbance in the Haiwee area during the 1872 event primarily reflected ground shaking or liquefaction-related deformation rather than triggered slip. In addition, we infer a structural and kinematic connection between the Owens Valley fault and oblique-dextral faults north of Lower Cactus Flat in the northwestern Coso Range, rather than awest-step into northern or western Rose Valley. Consideration of these structures in the total extent of the Owens Valley fault suggests a length of 140 km, of which at least 113 km ruptured during the 1872 event.


Lienkaemper J.J.,U.S. Geological Survey | Baldwin J.N.,Lettis Consultants International Inc. | Turner R.,William Lettis and Associates | Sickler R.R.,U.S. Geological Survey | Brown J.,U.S. Geological Survey
Bulletin of the Seismological Society of America | Year: 2013

We document evidence for surface-rupturing earthquakes (events) at two trench sites on the southern Green Valley fault, California (SGVF). The 75-80 km long dextral SGVF creeps ~1-4 mm/yr. We identify stratigraphic horizons disrupted by upward-flowering shears and infilled fissures unlikely to have formed from creep alone. The Mason Rd site exhibits four events from ~1013 CE to the present. The Lopes Ranch site (LR, 12 km to the south) exhibits three events from 18 BCE to present including the most recent event (MRE), 1610 ± 52 yr CE (1σ) and a two-event interval (18 BCE-238 CE) isolated by a millennium of low deposition. Using OxCal to model the timing of the four-event earthquake sequence from radiocarbon data and the LR MRE yields a mean recurrence interval (RI or μ) of 199 ± 82 yr (1σ) and ±35 yr (standard error of the mean), the first based on geologic data. The time since the most recent earthquake (open window since MRE) is 402 yr ± 52 yr, well past μ ~ 200 yr. The shape of the probability density function (PDF) of the average RI from OxCal resembles a Brownian passage time (BPT) PDF (i.e., rather than normal) that permits rarer longer ruptures potentially involving the Berryessa and Hunting Creek sections of the northernmost GVF The model coefficient of variation (cv, σ/μ) is 0.41, but a larger value (cv ~ 0.6) fits better when using BPT. A BPT PDF with μ of 250 and cv of 0.6 yields 30 yr rupture probabilities of 20%-25% versus a Poisson probability of 11%-17%.


Schwartz D.P.,U.S. Geological Survey | Lienkaemper J.J.,U.S. Geological Survey | Hecker S.,U.S. Geological Survey | Kelson K.I.,URS Corporation | And 4 more authors.
Bulletin of the Seismological Society of America | Year: 2014

Stress changes produced by the 1906 San Francisco earthquake had a profound effect on the seismicity of the San Francisco Bay region (SFBR), dramatically reducing it in the twentieth century. Whether the SFBR is still within or has emerged from this seismic quiescence is an issue of debate with implications for earthquake mechanics and seismic hazards. Historically, the SFBR has not experienced one complete earthquake cycle (i.e., the accumulation of stress, its release primarily as coseismic slip during surface-faulting earthquakes, its re-accumulation in the interval following, and its subsequent rerelease). The historical record of earthquake occurrence in the SFBR appears to be complete at aboutM5.5 back to 1850 (Bakun, 1999). For large events, the record may be complete back to 1776, which represents about half a cycle. Paleoseismic data provide a more complete view of the most recent pre- 1906 SFBR earthquake cycle, extending it back to about 1600. Using these, we have developed estimates of magnitude and seismic moment for alternative sequences of surface-faulting paleoearthquakes occurring between 1600 and 1776 on the region's major faults. From these we calculate seismic moment and moment release rates for different time intervals between 1600 and 2012. These show the variability in moment release and suggest that, in the SFBR regional plate boundary, stress can be released on a single fault in great earthquakes such as that in 1906 and in multiple ruptures distributed on the regional plate boundary fault system on a decadal time scale.


Madugo C.M.,Earth Consultants International | Dolan J.F.,University of Southern California | Hartleb R.D.,University of Southern California | Hartleb R.D.,Lettis Consultants International Inc.
Bulletin of the Seismological Society of America | Year: 2012

New paleoseismic investigations on the western segment of the Garlock fault at Twin Lakes, California, reveal evidence for up to six surface ruptures in the past ~5600 years. Calibrated radiocarbon dates from accelerator mass spectrometer analysis of detrital charcoal constrain the timing of three well-defined events at Twin Lakes to post-A.D. 1450 (event A), 720-395 B.C. (event G), and 3425-2200 B.C. (event I); two probable events are also constrained to 625-1525 A.D. (event C), and 155 B.C.-A.D. 615 (event E), and a possible additional event to 3425-2200 B.C. and prior to event I. Our findings offer new insights into mid- to late-Holocene behavior of the western and central segments of the Garlock fault, and regional fault interactions. The timing of the most recent event (MRE) on the western segment likely correlates with the MRE at paleoseismic sites on the central segment, suggesting that both segments do sometimes rupture together during large earthquakes. Evidence for events during periods of seismic quiescence on adjacent segments demonstrates that the western and central segments also sometimes rupture independently of one another. The occurrence of event G during a lull in seismic strain release at 2-5 ka on faults in the eastern California shear zone (ECSZ), contrasts with other studies that suggest the Garlock fault ruptures in phase with ECSZ faults. Our data suggest that seismic strain release on the Garlock fault may actually be more in phase with moment release on the Mojave section of the San Andreas fault and the Transverse Ranges faults of the Los Angeles region.

Loading Lettis Consultants International Inc. collaborators
Loading Lettis Consultants International Inc. collaborators