Geology Reports⌕ Search

Geology topics

Edward H. Field

Publications and source records attributed to Edward H. Field.

At least 37 records · Page 2Linked to original sources
USGS2021

The 2018 update of the US National Seismic Hazard Model: Where, why, and how much probabilistic ground motion maps changed

The 2018 US Geological Survey National Seismic Hazard Model (NSHM) incorporates new data and updated science to improve the underlying earthquake and ground motion forecasts for the conterminous United States. The NSHM considers many new data and component input models: (1) new earthquakes between 2013 and 2017 and updated earthquake magnitudes for some earlier earthquakes; (2) two updated smoothed seismicity models to forecast earthquake rates; (3) two suites of new central and eastern US (CEUS) ground motion models (GMMs) to translate ground shaking for various earthquake sizes and source-to-site distances considered in the model; (4) two CEUS GMMs for aleatory variability; (5) two CEUS site-effect models that modify ground shaking based on alternative shallow site conditions; (6) more advanced western US (WUS) lithologic and structural information to assess basin site effects for selected urban regions; and (7) a more comprehensive range of outputs (22 periods and 8 site classes) than in previous versions of the NSHMs. Each of these new datasets and models produces changes in the probabilistic ground shaking levels that are spatially and statistically analyzed. Recent earthquakes or changes to some older earthquake magnitudes and locations mostly result in probabilistic ground shaking levels that are similar to previous models, but local changes can reach up to +80% and −60% compared to the 2014 model. Newly developed CEUS models for GMMs, aleatory variability, and site effects cause overall changes up to ±64%. The addition of the WUS basin amplifications causes changes of up to +60% at longer periods for sites overlying deep soft soils. Across the conterminous United States, the hazard changes in the model are mainly caused by new GMMs in the CEUS, by sedimentary basin effects for long periods (≥1 s) in the WUS, and by seismicity changes for short (0.2 s) and long (1 s) periods for both areas.

Earthquake Spectra↗
USGS2021

Toward physics-based nonergodic PSHA: A prototype fully-deterministic seismic hazard model for southern California

We present a nonergodic framework for probabilistic seismic‐hazard analysis (PSHA) that is constructed entirely of deterministic, physical models. The use of deterministic ground‐motion simulations in PSHA calculations is not new (e.g., CyberShake), but prior studies relied on kinematic rupture generators to extend empirical earthquake rupture forecasts. Fully dynamic models, which simulate rupture nucleation and propagation of static and dynamic stresses, are still computationally intractable for the large simulation domains and many seismic cycles required to perform PSHA. Instead, we employ the Rate‐State earthquake simulator (RSQSim) to efficiently simulate hundreds of thousands of years of M ≥ 6.5 earthquake sequences on the California fault system. RSQSim produces full slip‐time histories for each rupture, which, unlike kinematic models, emerge from frictional properties, fault geometry, and stress transfer; all intrinsic variability is deterministic. We use these slip‐time histories directly as input to a 3D wave‐propagation code within the CyberShake platform to obtain simulated F max = 0.5 Hz ground motions. The resulting 3 s spectral acceleration ground motions closely match empirical ground‐motion model (GMM) estimates of median and variability of shaking. When computed over a range of sources and sites, the variability is similar to that of ergodic GMMs. Variability is reduced for individual pairs of sources and sites that repeatedly sample a single path, which is expected for a nonergodic model. This results in increased exceedance probabilities for certain characteristic ground motions for a source–site pair, while decreasing probabilities at the extreme tails of the ergodic GMM predictions. We present these comparisons and preliminary fully deterministic physics‐based RSQSim–CyberShake hazard curves, as well as a new technique for estimating within‐ and between‐event variability through simulation.

California↗
USGS2020

Generalizing the inversion‐based PSHA source model for an interconnected fault system

This article represents a step toward generalizing and simplifying the procedure for constructing an inversion‐based seismic hazard source model for an interconnected fault system, including the specification of adjustable segmentation constraints. A very simple example is used to maximize understandability and to counter the notion that an inversion approach is only applicable when an abundance of data is available. Also exemplified is how to construct a range of models to adequately represent epistemic uncertainties (which should be a high priority in any hazard assessment). Opportunity is also taken to address common concerns and misunderstandings associated with the third Uniform California Earthquake Rupture Forecast, including the seemingly disproportionate number of large‐magnitude events, and how well hazard is resolved given the overall problem is very underdetermined. However, the main aim of this article is to provide a general protocol for constructing such models.

Bulletin of the Seismological Society of America↗
USGS2020

Pseudo-prospective evaluation of UCERF3-ETAS forecasts during the 2019 Ridgecrest sequence

The 2019 Ridgecrest sequence provides the first opportunity to evaluate Uniform California Earthquake Rupture Forecast v.3 with epidemic‐type aftershock sequences (UCERF3‐ETAS) in a pseudoprospective sense. For comparison, we include a version of the model without explicit faults more closely mimicking traditional ETAS models (UCERF3‐NoFaults). We evaluate the forecasts with new metrics developed within the Collaboratory for the Study of Earthquake Predictability (CSEP). The metrics consider synthetic catalogs simulated by the models rather than synoptic probability maps, thereby relaxing the Poisson assumption of previous CSEP tests. Our approach compares statistics from the synthetic catalogs directly against observations, providing a flexible approach that can account for dependencies and uncertainties encoded in the models. We find that, to the first order, both UCERF3‐ETAS and UCERF3‐NoFaults approximately capture the spatiotemporal evolution of the Ridgecrest sequence, adding to the growing body of evidence that ETAS models can be informative forecasting tools. However, we also find that both models mildly overpredict the seismicity rate, on average, aggregated over the evaluation period. More severe testing indicates the overpredictions occur too often for observations to be statistically indistinguishable from the model. Magnitude tests indicate that the models do not include enough variability in forecasted magnitude‐number distributions to match the data. Spatial tests highlight discrepancies between the forecasts and observations, but the greatest differences between the two models appear when aftershocks occur on modeled UCERF3‐ETAS faults. Therefore, any predictability associated with embedding earthquake triggering on the (modeled) fault network may only crystalize during the presumably rare sequences with aftershocks on these faults. Accounting for uncertainty in the model parameters could improve test results during future experiments.

California↗
USGS2020

Assessing the value of removing earthquake-hazard-related epistemic uncertainties, exemplified using average annual loss in California

To aid in setting scientific research priorities, we assess the potential value of removing each of the epistemic uncertainties currently represented in the US Geological Survey California seismic-hazard model, using average annual loss (AAL) as the risk metric of interest. Given all the uncertainties, represented with logic-tree branches, we find a mean AAL of $3.94 billion. The modal value is 17.5% lower than the mean, and there is a 78% chance that the true AAL value is more than 10% away from the mean, and a 5% chance that it is a factor 2.1 greater or lower than the mean. We quantify the extent to which resolving each uncertainty improves the AAL estimate. The most influential branch is one that adds additional epistemic uncertainty to ground motion models, but others are found to be influential as well, such as the rate of M ≥ 5 events throughout the region. We discuss the broader implications of our findings, and note that the time dependence caused by spatiotemporal clustering can be much more influential on AAL than the epistemic uncertainties explored here.

California↗
USGS2020

Operational earthquake forecasting during the 2019 Ridgecrest, California, earthquake sequence with the UCERF3-ETAS model

The first Uniform California Earthquake Rupture Forecast, Version 3–epidemic‐type aftershock sequence (UCERF3‐ETAS) aftershock simulations were running on a high‐performance computing cluster within 33 min of the 4 July 2019 M 6.4 Searles Valley earthquake. UCERF3‐ETAS, an extension of the third Uniform California Earthquake Rupture Forecast (UCERF3), is the first comprehensive, fault‐based, epidemic‐type aftershock sequence (ETAS) model. It produces ensembles of synthetic aftershock sequences both on and off explicitly modeled UCERF3 faults to answer a key question repeatedly asked during the Ridgecrest sequence: What are the chances that the earthquake that just occurred will turn out to be the foreshock of an even bigger event?

California↗
USGS2020

Probabilistic seismic hazard analysis at regional and national scale: State of the art and future challenges

Seismic hazard modelling is a multi-disciplinary science that aims to forecast earthquake occurrence and its resultant ground shaking. Such models consist of a probabilistic framework that quantifies uncertainty across a complex system; typically, this includes at least two model components developed from Earth science: seismic-source and ground-motion models. Although there is no scientific prescription for the forecast length, the most common probabilistic seismic hazard analyses consider forecasting windows of 30 to 50 years, which are typically an engineering demand for building code purposes. These types of analyses are the topic of this review paper. Although the core methods and assumptions of seismic hazard modelling have largely remained unchanged for more than 50 years, we review the most recent initiatives which face the difficult task of meeting both the increasingly sophisticated demands of society and keeping pace with advances in scientific understanding. A need for more accurate and spatially precise hazard forecasting must be balanced with increased quantification of uncertainty and new challenges such as moving from time-independent hazard to forecasts that are time-dependent and specific to the time-period of interest. Meeting these challenges requires the development of science-driven models which integrate all information available, the adoption of proper mathematical frameworks to quantify the different types of uncertainties in the hazard model, and the development of a proper testing phase of the model to quantify its consistency and skill. We review the state-of-the-art of the national seismic hazard modeling, and how the most innovative approaches try to address future challenges.

Reviews of Geophysics↗
USGS2020

An exploration of parametric earthquake risk transfer solutions that dynamically adapt to seismicity changes

(Re)insurance companies rely on earthquake risk models to estimate the frequency and severity of their potential financial losses. To protect themselves, they sometimes use parametric risk transfer solutions, which are derivative-form agreements that provide compensation as a function of routine measurable earthquake characteristics. These mechanisms typically remain in force for one to three years and assume seismic conditions—and our estimates of them—remain unchanged during this period. However, seismic risk estimates evolve continuously due to changes in nearby seismicity, sudden ruptures, slower redistributions of stress, or improvements in our own understanding of these phenomena. As a consequence, the likelihood of some loss-causing events might decrease and make the protection superfluous (wasted money), or, more problematically, it might increase and render the protection insufficient (increased risk). This paper explores the construction of parametric earthquake risk transfer mechanisms that adapt efficiently (i.e., near real-time) to changes in seismicity throughout the lifetime of the transaction. The mechanism proposes the periodic adjustment of the payment conditions of the parametric agreement in harmony with the evolving probabilities of event occurrence. This, we hypothesize, may result in a more efficient allocation of premiums that reflects the changing nature of seismic risk. To build the proposed dynamic risk transfer mechanism, we first employ one of the earthquake models commonly used in the (re)insurance industry to assess the risk of a portfolio of assets. The modeling exercise yields the expected frequency distribution of loss, which a standard (re)insurance transaction would typically consider constant for the entire coverage period. Here, we use these results simply as a baseline for the initial time step of reference. Next, we construct a retrospective update loop, which consists of two parts: (1) we obtain the earthquake occurrence rate conditions at a previous time step taking into account the changes in seismicity observed in the interim period; and (2) we use the modeled losses and adjusted frequencies at the new time step to build a parametric risk transfer solution. This parametric solution remains in force until it is updated at the next iteration. We also track the effects on the efficiency of the risk transfer solution and its premium if these continuous updates were not implemented. We apply the proposed mechanism to California and find that changes in seismicity can cause swings in the frequency of parametric payments (which is related to the premium paid for the cover) in average of 16% and up to 36% in any three-year period from 1986 to 2020. We also find that avoiding an update of the parametric solution on a yearly basis to match the new risk profile can decrease the efficiency of the cover (measured as the relative contribution to the average annual loss of the events covered) in the same time period by 13% on average and up to 35%.

Conference Paper↗
USGS2020

2018 and 2023 U. S. National Seismic Hazard Models

During 2017-2018, the National Seismic Hazard Model (NSHM) was updated by incorporating (1) new median ground motion models, new estimates of their epistemic uncertainties and aleatory variabilities, and new soil amplification factors for the central and eastern U.S., (2) amplification of long-period ground motions in deep sedimentary basins in the Los Angeles, San Francisco, Seattle, and Salt Lake City areas, (3) an updated seismicity catalog, which includes new earthquakes that occurred between 2012 and 2017, and (4) improved computer code and implementation details. Results show significantly increased ground shaking in many (but not all) locations across the central and eastern U.S., and increased ground shaking in the four urban areas (listed above) that overlie deep sedimentary basins in the western U.S. During 2019-2023, the NSHM will consider additional updates for the 2023 NSHM including: 3D simulations in urban areas of Los Angeles and Seattle, additional soil amplification models for other urban areas (e.g., eastern and southern coastal plains of the U.S., Reno, and Las Vegas), non-ergodic ground motion models that reconsider epistemic and aleatory uncertainty, topographic amplification, and new geological and geodetic models for faults. These maps will be considered by the Building Seismic Safety Council, American Society of Civil Engineers, and International Building Code committees for inclusion in upcoming building codes. Due to population growth, more people live and work in areas of moderate or high seismic hazard than ever before, leading to higher risk of undesirable consequences from future ground shaking.

conterminous United States↗
USGS2019

The 2018 update of the US National Seismic Hazard Model: Overview of model and implications

During 2017–2018, the National Seismic Hazard Model for the conterminous United States was updated as follows: (1) an updated seismicity catalog was incorporated, which includes new earthquakes that occurred from 2013 to 2017; (2) in the central and eastern United States (CEUS), new ground motion models were updated that incorporate updated median estimates, modified assessments of the associated epistemic uncertainties and aleatory variabilities, and new soil amplification factors; (3) in the western United States (WUS), amplified shaking estimates of long-period ground motions at sites overlying deep sedimentary basins in the Los Angeles, San Francisco, Seattle, and Salt Lake City areas were incorporated; and (4) in the conterminous United States, seismic hazard is calculated for 22 periods (from 0.01 to 10 s) and 8 uniform V S30 maps (ranging from 1500 to 150 m/s). We also include a description of updated computer codes and modeling details. Results show increased ground shaking in many (but not all) locations across the CEUS (up to ~30%), as well as near the four urban areas overlying deep sedimentary basins in the WUS (up to ~50%). Due to population growth and these increased hazard estimates, more people live or work in areas of high or moderate seismic hazard than ever before, leading to higher risk of undesirable consequences from forecasted future ground shaking.

Earthquake Spectra↗
USGS2019

Relaxing segmentation on the Wasatch Fault Zone: Impact on seismic hazard

The multisegment Wasatch fault zone is a well-studied normal fault in the western United States that has paleoseismic evidence of recurrent Holocene surface-faulting earthquakes. Along the 270-km-long central part of the fault, four primary structural complexities provide possible along-strike limits to these ruptures and form the basis for models of fault segmentation. Here, we assess the impact that the Wasatch fault segmentation model has on seismic hazard by evaluating the time-independent long-term rate of ruptures on the fault that satisfy fault slip rates and paleoseismic event rates, adapting standard inverse theory used in the Uniform California Earthquake Rupture Forecast 3 (UCERF3), and implementing a segmentation constraint where ruptures across primary structural complexities are penalized. We define three models with varying degrees of rupture penalization: (1) segmented (ruptures confined to individual segments), (2) penalized (multi-segment ruptures allowed, but penalized), and (3) unsegmented (all ruptures allowed). Seismic-hazard results show that on average, hazard is highest for the segmented model, where seismic moment is accommodated by frequent moderate (moment magnitude, M w 6.2–6.8) earthquakes. The unsegmented model yields the lowest average seismic hazard because part of the seismic moment is accommodated by large (M w 6.9–7.9), but infrequent ruptures. We compare these results to model differences derived from other inputs such as slip rate and magnitude scaling relationships and conclude that segmentation exerts a primary control on seismic hazard. This study demonstrates the need for additional geologic constraints on rupture extent and methods by which these observations can be included in hazard-modeling efforts.

Utah↗
USGS2019

How physics‐based earthquake simulators might help improve earthquake forecasts

Questions have persisted on the usefulness of physics‐based earthquake simulators with respect to forecasting earthquakes, due mostly to the inevitable assumptions, approximations, and uncertainties. Whether any model is reliable or trustworthy depends entirely on what questions we are asking of it, so the point of this article is to outline a number of currently anticipated and desired inferences, informed largely by recent forecasting efforts in California. This article does not provide an in‐depth review of physics‐based simulators, nor does it render judgment on the usefulness question. The intent, rather, is to provide an explicit list of potential uses, or inferences, to enable more informed discussions among physics‐based modeling experts.

Seismological Research Letters↗
USGS2018

Improving earthquake rupture forecasts using California as a guide

This article discusses ways in which earthquake rupture forecast models might be improved. Because changes are most easily described in the context of specific models, the third Uniform California Earthquake Rupture Forecast (UCERF3) and its presumed successor, UCERF4, is used as a basis for discussion. Virtually all of the issues and possible improvements discussed are nevertheless general and should therefore be applicable to other regions as well. Two common themes are a need for better epistemic uncertainty representation and the potential utility of physics‐based simulators. Given the large number of possible improvements, coupled with challenges in defining the potential value of each, which will vary among uses, community feedback is invaluable in terms of setting priorities. We should also strive to define more objective valuation metrics.

Seismological Research Letters↗
USGS2018

Candidate products for operational earthquake forecasting illustrated using the HayWired planning scenario, including one very quick (and not‐so‐dirty) hazard‐map option

In an effort to help address debates on the usefulness of operational earthquake forecasting (OEF), we illustrate a number of OEF products that could be automatically generated in near‐real time. To exemplify, we use an M "> M 7.1 mainshock on the Hayward fault, which is very similar to the U.S. Geological Survey (USGS) HayWired earthquake planning scenario. Given that there is always some background level of hazard or risk, we emphasize that probability gains (the ratio of short‐term to long‐term‐average estimates) might be of particular interest to users. We also illustrate how such gains are highly sensitive to forecast duration and latency, with the latter representing how long it takes to generate the forecast and/or to take action. The influence of fault‐based information, which has traditionally been ignored in OEF, is also evaluated using the newly developed the third Uniform California Earthquake Rupture Forecast epidemic‐type aftershock sequence (UCERF3‐ETAS) model. We find that the inclusion of faults only makes a difference for hazard and risk metrics that are dominated by large‐event likelihoods. We also show how the ShakeMap of a mainshock represents a decent estimate of the ground motions that have a 6% chance of being exceeded due to aftershocks in the week that follows. The ultimate value of these types of OEF products can only be determined in the context of specific uses, and because these vary widely, institutions responsible for providing OEF products will depend heavily on user feedback, especially when making resource‐allocation decisions.

Seismological Research Letters↗
USGS2018

A physics-based earthquake simulator replicates seismic hazard statistics across California

Seismic hazard models are important for society, feeding into building codes and hazard mitigation efforts. These models, however, rest on many uncertain assumptions and are difficult to test observationally because of the long recurrence times of large earthquakes. Physics-based earthquake simulators offer a potentially helpful tool, but they face a vast range of fundamental scientific uncertainties. We compare a physics-based earthquake simulator against the latest seismic hazard model for California. Using only uniform parameters in the simulator, we find strikingly good agreement of the long-term shaking hazard compared with the California model. This ability to replicate statistically based seismic hazard estimates by a physics-based model cross-validates standard methods and provides a new alternative approach needing fewer inputs and assumptions for estimating hazard.

California↗
USGS2017

A prototype operational earthquake loss model for California based on UCERF3-ETAS – A first look at valuation

We present a prototype operational loss model based on UCERF3-ETAS, which is the third Uniform California Earthquake Rupture Forecast with an Epidemic Type Aftershock Sequence (ETAS) component. As such, UCERF3-ETAS represents the first earthquake forecast to relax fault segmentation assumptions and to include multi-fault ruptures, elastic-rebound, and spatiotemporal clustering, all of which seem important for generating realistic and useful aftershock statistics. UCERF3-ETAS is nevertheless an approximation of the system, however, so usefulness will vary and potential value needs to be ascertained in the context of each application. We examine this question with respect to statewide loss estimates, exemplifying how risk can be elevated by orders of magnitude due to triggered events following various scenario earthquakes. Two important considerations are the probability gains, relative to loss likelihoods in the absence of main shocks, and the rapid decay of gains with time. Significant uncertainties and model limitations remain, so we hope this paper will inspire similar analyses with respect to other risk metrics to help ascertain whether operationalization of UCERF3-ETAS would be worth the considerable resources required.

Earthquake Spectra↗
USGS2017

Trimming a hazard logic tree with a new model-order-reduction technique

The size of the logic tree within the Uniform California Earthquake Rupture Forecast Version 3, Time-Dependent (UCERF3-TD) model can challenge risk analyses of large portfolios. An insurer or catastrophe risk modeler concerned with losses to a California portfolio might have to evaluate a portfolio 57,600 times to estimate risk in light of the hazard possibility space. Which branches of the logic tree matter most, and which can one ignore? We employed two model-order-reduction techniques to simplify the model. We sought a subset of parameters that must vary, and the specific fixed values for the remaining parameters, to produce approximately the same loss distribution as the original model. The techniques are (1) a tornado-diagram approach we employed previously for UCERF2, and (2) an apparently novel probabilistic sensitivity approach that seems better suited to functions of nominal random variables. The new approach produces a reduced-order model with only 60 of the original 57,600 leaves. One can use the results to reduce computational effort in loss analyses by orders of magnitude.

Earthquake Spectra↗