Peter Bird - Academia.edu (original) (raw)
Papers by Peter Bird
Geosphere
Fault traces and offsets, cross-section length changes, paleomagnetic inclination and declination... more Fault traces and offsets, cross-section length changes, paleomagnetic inclination and declination anomalies, and stress-direction indicators with ages back to 90 Ma are collected from the geologic literature on the western United States and northern Mexico. Finite-element program Restore simulates paleokinematics by weighted least squares and integrates displacements, strains, and rotations back in time, producing paleogeologic maps, as well as maps of velocity, heave rate, strain rate, and stress direction at 6 m.y. intervals. After calibrating three program parameters against neotectonic velocities from geodesy, all classes of data except inclination anomalies are fit reasonably well. The kink in the San Andreas fault near San Gorgonio Pass has been gradually restored by slip on adjacent faults and automated smoothing. Piercing-point pairs successfully restored along the San Andreas–Gulf of California plate boundary include the Pelona and Orocopia Schists at 6 Ma, the Pinnacles an...
Bulletin of the Seismological Society of America, Dec 1, 1999
Seismic-hazard estimates require merging datasets from seismology, geodesy, and geology that have... more Seismic-hazard estimates require merging datasets from seismology, geodesy, and geology that have different footprints, which works best when there is a global model that relates them. The rigid-plate model serves this purpose for plate-boundary hazard estimation but fails for intraplate hazard. The global thin-shell finite-element model of the lithosphere by Bird (1998) is a first attempt at a model with realistic intraplate regions. It is based on stress equilibrium and assumed flow laws for the lithosphere, incorporates topographic and geothermal variations and plate-boundary faults and predicts long-term anelastic strain rates in all plate interiors. For comparison, the historical seismic catalog JISSIG (Utsu, 1996) and the instrumental catalog CMT (Dziewonski et al. , 1997) are converted to seismic strain rates in intraplate regions corresponding to the finite elements. Spatial correlations are tested between the logarithms of strain rates in order to reduce the dominance of the greatest earthquakes. The finite-element model has correlations of 0.391 with JISSIG and 0.338 with CMT, neither of which is as high as the correlation of JISSIG with CMT (0.476). This may mean that the model is defective or that intraplate seismicity contains a large fraction of aftershocks and other triggered events. If we take the CMT dataset as representative of the short-term intraplate seismicity that we might try to forecast, then multiple regression shows that this is best correlated (0.504) with a 56/44% mixture of the JISSIG and finite-element model strain rates. The contribution of the finite-element model, though small, is significant with 99.9% confidence.
1. Statistical modeling of geologic offset-rates with program Slippery. Direct evidence from date... more 1. Statistical modeling of geologic offset-rates with program Slippery. Direct evidence from dated offset features on each f ult, and/or indirect evidence from dated offset features on other faults of the same t yp in the same tectonic province, are combined to obtain the “pure-geologic” probability density function (PDF) for each component of offset-rate on each fault. Additional constraints such as geometric compatibility, plate tectonics, geodesy, and stress directions are not used in this phase of the study.
* * Vice-Chair Science Working Groups & Planning Committee (PC) Disciplinary Com. Interdi... more * * Vice-Chair Science Working Groups & Planning Committee (PC) Disciplinary Com. Interdisciplinary Focus Groups Special Projects
Seismic Hazard Harmonization in Europe (SHARE) published new seismic hazard maps of Europe in 201... more Seismic Hazard Harmonization in Europe (SHARE) published new seismic hazard maps of Europe in 2013. Seismic source models are the basis for hazard calculations. SHARE constructed three seismic source models based on historical earthquakes and geological fault data. The SHARE source models provided parameters from which magnitude-frequency distributions can be specified for each of 437 seismic source zones covering most of Europe. To evaluate the SHARE seismic source models, we construct an earthquake potential model of Southern Europe using the Global Strain Rate Map released in 2014. Because the individual SHARE area source zones are too small to have sufficient data for accurate estimates, we combine the source zones into five groups according to SHARE’s estimates of maximum magnitude. Using the strain rates, we calculate tectonic moment rates for each group. Then, we infer seismicity rates and probable maximum earthquake magnitudes from the tectonic moment rates. For two groups, ...
Seismological Research Letters, 2018
Bulletin of the Seismological Society of America, 2014
A major upgrade of the Global Strain Rate Map, version 2.1, uses far more geodetic data, systemat... more A major upgrade of the Global Strain Rate Map, version 2.1, uses far more geodetic data, systematic data processing, more modeled plates and plate boundaries, an improved algorithm, and a finer spatial grid than version 1 (Kreemer et al., 2014). We convert this model to an indefinite-term tectonic forecast of shallow seismicity on a fine global grid, using the assumptions of Bird et al. (2010) and similar algorithms. One new feature is smoothing of model strain rates, and thus forecast seismicity, around offshore plate boundaries where strain rates are not controlled by local geodetic data. Another is incorporation of velocity-dependent seismic coupling in subduction zones and continental convergent boundaries (Bird et al., 2009). The seismicity model is constructed in six progressively more complex versions. Only catalog years 1977-2004 are used for calibration, leaving years 2005-2012 available for retrospective tests. Tests of forecast success that are independent of total earthquake rate and that use no declustering show success comparable to that of one mature, optimized smoothed-seismicity algorithm.
Bulletin of the Seismological Society of America, 2015
In our previous article (Bird and Kreemer, 2015), we presented a set of six long‐term stationary ... more In our previous article (Bird and Kreemer, 2015), we presented a set of six long‐term stationary forecasts of global shallow seismicity, identified as the model series SHIFT‐GSRM2a–f. As preliminary validation, we also presented retrospective tests against global shallow seismicity in the years 2005–2012, as recorded in the Global Centroid Moment Tensor catalog. A smoothed‐seismicity forecast computed by Yan Kagan and David Jackson (Kagan and Jackson, 1994, 2010a,b, 2011, 2014) was …
Geophysical Journal International, 2017
Agu Fall Meeting Abstracts, Dec 1, 2003
A new plate model is used to analyze the mean seismicities of seven types of plate boundary (CRB,... more A new plate model is used to analyze the mean seismicities of seven types of plate boundary (CRB, continental rift boundary; CTF, continental transform fault; CCB, continental convergent boundary; OSR, oceanic spreading ridge; OTF, oceanic transform fault; OCB, oceanic convergent boundary; SUB, subduction zone). We compare the platelike (nonorogen) regions of model PB2002 (Bird, 2003) with the centroid moment tensor (CMT) catalog to select apparent boundary half-widths and then assign 95% of shallow earthquakes to one of these settings. A tapered Gutenberg-Richter model of the frequency/moment relation is fit to the subcatalog for each setting by maximum likelihood. Best-fitting b values range from 0.53 to 0.92, but all 95% confidence ranges are consistent with a common value of 0.61-0.66. To better determine some corner magnitudes we expand the subcatalogs by (1) inclusion of orogens and (2) inclusion of years 1900-1975 from the catalog of Pacheco and Sykes (1992). Combining both earthquake statistics and the platetectonic constraint on moment rate, corner magnitudes include the following:
Bulletin of the Seismological Society of America, 2015
Global earthquake activity rate model 1 (GEAR1) estimates the rate of shallow earthquakes with ma... more Global earthquake activity rate model 1 (GEAR1) estimates the rate of shallow earthquakes with magnitudes 6-9 everywhere on Earth. It was designed to be reproducible and testable. Our preferred hybrid forecast is a log-linear blend of two parent forecasts based on the Global Centroid Moment Tensor (CMT) catalog (smoothing 4602 m ≥ 5:767 shallow earthquakes, 1977-2004) and the Global Strain Rate Map version 2.1 (smoothing 22,415 Global Positioning System velocities), optimized to best forecast the 2005-2012 Global CMT catalog. Strain rate is a proxy for fault stress accumulation, and earthquakes indicate stress release, so a multiplicative blend is desirable, capturing the strengths of both approaches. This preferred hybrid forecast outperforms its seismicity and strain-rate parents; the chance that this improvement stems from random seismicity fluctuations is less than 1%. The preferred hybrid is also tested against the independent parts of the International Seismological Centre-Global Earthquake Model catalog (m ≥ 6:8 during 1918-1976) with similar success. GEAR1 is an update of this preferred hybrid. Comparing GEAR1 to the Uniform California Earthquake Rupture Forecast Version 3 (UCERF3), net earthquake rates agree within 4% at m ≥ 5:8 and at m ≥ 7:0. The spatial distribution of UCERF3 epicentroids most resembles GEAR1 after UCERF3 is smoothed with a 30 km kernel. Because UCERF3 has been constructed to derive useful information from fault geometry, slip rates, paleoseismic data, and enhanced seismic catalogs (not used in our model), this is encouraging. To build parametric catastrophe bonds from GEAR1, one could calculate the magnitude for which there is a 1% (or any) annual probability of occurrence in local regions. Online Material: Discussion of forecast-scoring metrics, tables of scoring results, and source code and data files needed to reproduce forecast.
Journal of Geophysical Research: Solid Earth, 2015
We present a neotectonic model of ongoing lithosphere deformation and a corresponding estimate of... more We present a neotectonic model of ongoing lithosphere deformation and a corresponding estimate of long-term shallow seismicity across the Africa-Eurasia plate boundary, including the eastern Atlantic, Mediterranean region, and continental Europe. GPS and stress data are absent or inadequate for the part of the study area covered by water. Thus, we opt for a dynamic model based on the stress-equilibrium equation; this approach allows us to estimate the long-term behavior of the lithosphere (given certain assumptions about its structure and physics) for both land and sea areas. We first update the existing plate model by adding five quasi-rigid plates (the Ionian Sea, Adria, Northern Greece, Central Greece, and Marmara) to constrain the deformation pattern of the study area. We use the most recent datasets to estimate the lithospheric structure. The models are evaluated in comparison with updated datasets of geodetic velocities and the most compressive horizontal principal stress azimuths. We find that the side and basal strengths drive the present-day motion of the Adria and Aegean Sea plates, whereas lithostatic pressure plays a key role in driving Anatolia. These findings provide new insights into the neotectonics of the greater Mediterranean region. Finally, the preferred model is used to estimate long-term shallow seismicity, which we retrospectively test against historical seismicity. As an alternative to reliance on incomplete geologic data or historical seismic catalogs, these neotectonic models help to forecast long-term seismicity, although requiring additional tuning before seismicity rates are used for seismic hazard purposes.
Maurice Ewing Series, 1977
Plate Boundary Zones, 2013
We use the Harvard CMT catalog to separate ocean-ridge seismicity into spreading and transform su... more We use the Harvard CMT catalog to separate ocean-ridge seismicity into spreading and transform sub-catalogs. We use the tapered Gutenberg-Richter distribution to estimate the total seismic moment rates of plate-boundary zones from limited catalogs of large events. We present the plate boundary model PB1999 and use it to associate marine earthquakes with particular plate boundary segments. We then combine these tools to estimate corner magnitudes (c m), spectral slopes (β), and coupled lithosphere thicknesses for all spreading ridges and oceanic transform faults. The distribution of spreading earthquakes is consistent with "normal" 2 3 β = (although β is not well constrained) and with uniform c 5.8 m =. Coupled lithosphere thickness along ridges decreases quasi-exponentially (from about 500 m to under 50 m) as spreading rate increases. Oceanic transform faults also have "normal" 2 3 β ≅ , but their corner magnitudes decrease from about 7.1 to about 6.3 with increasing relative plate velocity. Oceanic transform faults also show a quasi-exponential decrease in coupled lithosphere thickness (from about 3000 m to about 300 m) as relative plate velocity increases. Perhaps this is due to formation of serpentine along slow ridges and transforms and its absence from fast ridges and transforms. Spreading ridges and oceanic transform faults both have imperfect seismic coupling because: (i) all detailed local studies of seismogenic lithosphere thickness exceed our mean values for coupled thickness, and (ii) if coupling were perfect, and seismogenic lithosphere thickness were as small as our estimated coupled thickness, it would require unreasonable stress drops or rupture shapes to explain the moments of the largest earthquakes.
Bulletin of the Seismological Society of America, 2015
The 2014 Working Group on California Earthquake Probabilities (WGCEP 2014) presents time-dependen... more The 2014 Working Group on California Earthquake Probabilities (WGCEP 2014) presents time-dependent earthquake probabilities for the third Uniform California Earthquake Rupture Forecast (UCERF3). Building on the UCERF3 time-independent model published previously, renewal models are utilized to represent elasticrebound-implied probabilities. A new methodology has been developed that solves applicability issues in the previous approach for unsegmented models. The new methodology also supports magnitude-dependent aperiodicity and accounts for the historic open interval on faults that lack a date-of-last-event constraint. Epistemic uncertainties are represented with a logic tree, producing 5760 different forecasts. Results for a variety of evaluation metrics are presented, including logic-tree sensitivity analyses and comparisons to the previous model (UCERF2). For 30 yr M ≥ 6:7 probabilities, the most significant changes from UCERF2 are a threefold increase on the Calaveras fault and a threefold decrease on the San Jacinto fault. Such changes are due mostly to differences in the time-independent models (e.g., fault-slip rates), with relaxation of segmentation and inclusion of multifault ruptures being particularly influential. In fact, some UCERF2 faults were simply too long to produce M 6.7 size events given the segmentation assumptions in that study. Probability model differences are also influential, with the implied gains (relative to a Poisson model) being generally higher in UCERF3. Accounting for the historic open interval is one reason. Another is an effective 27% increase in the total elastic-rebound-model weight. The exact factors influencing differences between UCERF2 and UCERF3, as well as the relative importance of logic-tree branches, vary throughout the region and depend on the evaluation metric of interest. For example, M ≥ 6:7 probabilities may not be a good proxy for other hazard or loss measures. This sensitivity, coupled with the approximate nature of the model and known limitations, means the applicability of UCERF3 should be evaluated on a case-by-case basis.
Geosphere
Fault traces and offsets, cross-section length changes, paleomagnetic inclination and declination... more Fault traces and offsets, cross-section length changes, paleomagnetic inclination and declination anomalies, and stress-direction indicators with ages back to 90 Ma are collected from the geologic literature on the western United States and northern Mexico. Finite-element program Restore simulates paleokinematics by weighted least squares and integrates displacements, strains, and rotations back in time, producing paleogeologic maps, as well as maps of velocity, heave rate, strain rate, and stress direction at 6 m.y. intervals. After calibrating three program parameters against neotectonic velocities from geodesy, all classes of data except inclination anomalies are fit reasonably well. The kink in the San Andreas fault near San Gorgonio Pass has been gradually restored by slip on adjacent faults and automated smoothing. Piercing-point pairs successfully restored along the San Andreas–Gulf of California plate boundary include the Pelona and Orocopia Schists at 6 Ma, the Pinnacles an...
Bulletin of the Seismological Society of America, Dec 1, 1999
Seismic-hazard estimates require merging datasets from seismology, geodesy, and geology that have... more Seismic-hazard estimates require merging datasets from seismology, geodesy, and geology that have different footprints, which works best when there is a global model that relates them. The rigid-plate model serves this purpose for plate-boundary hazard estimation but fails for intraplate hazard. The global thin-shell finite-element model of the lithosphere by Bird (1998) is a first attempt at a model with realistic intraplate regions. It is based on stress equilibrium and assumed flow laws for the lithosphere, incorporates topographic and geothermal variations and plate-boundary faults and predicts long-term anelastic strain rates in all plate interiors. For comparison, the historical seismic catalog JISSIG (Utsu, 1996) and the instrumental catalog CMT (Dziewonski et al. , 1997) are converted to seismic strain rates in intraplate regions corresponding to the finite elements. Spatial correlations are tested between the logarithms of strain rates in order to reduce the dominance of the greatest earthquakes. The finite-element model has correlations of 0.391 with JISSIG and 0.338 with CMT, neither of which is as high as the correlation of JISSIG with CMT (0.476). This may mean that the model is defective or that intraplate seismicity contains a large fraction of aftershocks and other triggered events. If we take the CMT dataset as representative of the short-term intraplate seismicity that we might try to forecast, then multiple regression shows that this is best correlated (0.504) with a 56/44% mixture of the JISSIG and finite-element model strain rates. The contribution of the finite-element model, though small, is significant with 99.9% confidence.
1. Statistical modeling of geologic offset-rates with program Slippery. Direct evidence from date... more 1. Statistical modeling of geologic offset-rates with program Slippery. Direct evidence from dated offset features on each f ult, and/or indirect evidence from dated offset features on other faults of the same t yp in the same tectonic province, are combined to obtain the “pure-geologic” probability density function (PDF) for each component of offset-rate on each fault. Additional constraints such as geometric compatibility, plate tectonics, geodesy, and stress directions are not used in this phase of the study.
* * Vice-Chair Science Working Groups & Planning Committee (PC) Disciplinary Com. Interdi... more * * Vice-Chair Science Working Groups & Planning Committee (PC) Disciplinary Com. Interdisciplinary Focus Groups Special Projects
Seismic Hazard Harmonization in Europe (SHARE) published new seismic hazard maps of Europe in 201... more Seismic Hazard Harmonization in Europe (SHARE) published new seismic hazard maps of Europe in 2013. Seismic source models are the basis for hazard calculations. SHARE constructed three seismic source models based on historical earthquakes and geological fault data. The SHARE source models provided parameters from which magnitude-frequency distributions can be specified for each of 437 seismic source zones covering most of Europe. To evaluate the SHARE seismic source models, we construct an earthquake potential model of Southern Europe using the Global Strain Rate Map released in 2014. Because the individual SHARE area source zones are too small to have sufficient data for accurate estimates, we combine the source zones into five groups according to SHARE’s estimates of maximum magnitude. Using the strain rates, we calculate tectonic moment rates for each group. Then, we infer seismicity rates and probable maximum earthquake magnitudes from the tectonic moment rates. For two groups, ...
Seismological Research Letters, 2018
Bulletin of the Seismological Society of America, 2014
A major upgrade of the Global Strain Rate Map, version 2.1, uses far more geodetic data, systemat... more A major upgrade of the Global Strain Rate Map, version 2.1, uses far more geodetic data, systematic data processing, more modeled plates and plate boundaries, an improved algorithm, and a finer spatial grid than version 1 (Kreemer et al., 2014). We convert this model to an indefinite-term tectonic forecast of shallow seismicity on a fine global grid, using the assumptions of Bird et al. (2010) and similar algorithms. One new feature is smoothing of model strain rates, and thus forecast seismicity, around offshore plate boundaries where strain rates are not controlled by local geodetic data. Another is incorporation of velocity-dependent seismic coupling in subduction zones and continental convergent boundaries (Bird et al., 2009). The seismicity model is constructed in six progressively more complex versions. Only catalog years 1977-2004 are used for calibration, leaving years 2005-2012 available for retrospective tests. Tests of forecast success that are independent of total earthquake rate and that use no declustering show success comparable to that of one mature, optimized smoothed-seismicity algorithm.
Bulletin of the Seismological Society of America, 2015
In our previous article (Bird and Kreemer, 2015), we presented a set of six long‐term stationary ... more In our previous article (Bird and Kreemer, 2015), we presented a set of six long‐term stationary forecasts of global shallow seismicity, identified as the model series SHIFT‐GSRM2a–f. As preliminary validation, we also presented retrospective tests against global shallow seismicity in the years 2005–2012, as recorded in the Global Centroid Moment Tensor catalog. A smoothed‐seismicity forecast computed by Yan Kagan and David Jackson (Kagan and Jackson, 1994, 2010a,b, 2011, 2014) was …
Geophysical Journal International, 2017
Agu Fall Meeting Abstracts, Dec 1, 2003
A new plate model is used to analyze the mean seismicities of seven types of plate boundary (CRB,... more A new plate model is used to analyze the mean seismicities of seven types of plate boundary (CRB, continental rift boundary; CTF, continental transform fault; CCB, continental convergent boundary; OSR, oceanic spreading ridge; OTF, oceanic transform fault; OCB, oceanic convergent boundary; SUB, subduction zone). We compare the platelike (nonorogen) regions of model PB2002 (Bird, 2003) with the centroid moment tensor (CMT) catalog to select apparent boundary half-widths and then assign 95% of shallow earthquakes to one of these settings. A tapered Gutenberg-Richter model of the frequency/moment relation is fit to the subcatalog for each setting by maximum likelihood. Best-fitting b values range from 0.53 to 0.92, but all 95% confidence ranges are consistent with a common value of 0.61-0.66. To better determine some corner magnitudes we expand the subcatalogs by (1) inclusion of orogens and (2) inclusion of years 1900-1975 from the catalog of Pacheco and Sykes (1992). Combining both earthquake statistics and the platetectonic constraint on moment rate, corner magnitudes include the following:
Bulletin of the Seismological Society of America, 2015
Global earthquake activity rate model 1 (GEAR1) estimates the rate of shallow earthquakes with ma... more Global earthquake activity rate model 1 (GEAR1) estimates the rate of shallow earthquakes with magnitudes 6-9 everywhere on Earth. It was designed to be reproducible and testable. Our preferred hybrid forecast is a log-linear blend of two parent forecasts based on the Global Centroid Moment Tensor (CMT) catalog (smoothing 4602 m ≥ 5:767 shallow earthquakes, 1977-2004) and the Global Strain Rate Map version 2.1 (smoothing 22,415 Global Positioning System velocities), optimized to best forecast the 2005-2012 Global CMT catalog. Strain rate is a proxy for fault stress accumulation, and earthquakes indicate stress release, so a multiplicative blend is desirable, capturing the strengths of both approaches. This preferred hybrid forecast outperforms its seismicity and strain-rate parents; the chance that this improvement stems from random seismicity fluctuations is less than 1%. The preferred hybrid is also tested against the independent parts of the International Seismological Centre-Global Earthquake Model catalog (m ≥ 6:8 during 1918-1976) with similar success. GEAR1 is an update of this preferred hybrid. Comparing GEAR1 to the Uniform California Earthquake Rupture Forecast Version 3 (UCERF3), net earthquake rates agree within 4% at m ≥ 5:8 and at m ≥ 7:0. The spatial distribution of UCERF3 epicentroids most resembles GEAR1 after UCERF3 is smoothed with a 30 km kernel. Because UCERF3 has been constructed to derive useful information from fault geometry, slip rates, paleoseismic data, and enhanced seismic catalogs (not used in our model), this is encouraging. To build parametric catastrophe bonds from GEAR1, one could calculate the magnitude for which there is a 1% (or any) annual probability of occurrence in local regions. Online Material: Discussion of forecast-scoring metrics, tables of scoring results, and source code and data files needed to reproduce forecast.
Journal of Geophysical Research: Solid Earth, 2015
We present a neotectonic model of ongoing lithosphere deformation and a corresponding estimate of... more We present a neotectonic model of ongoing lithosphere deformation and a corresponding estimate of long-term shallow seismicity across the Africa-Eurasia plate boundary, including the eastern Atlantic, Mediterranean region, and continental Europe. GPS and stress data are absent or inadequate for the part of the study area covered by water. Thus, we opt for a dynamic model based on the stress-equilibrium equation; this approach allows us to estimate the long-term behavior of the lithosphere (given certain assumptions about its structure and physics) for both land and sea areas. We first update the existing plate model by adding five quasi-rigid plates (the Ionian Sea, Adria, Northern Greece, Central Greece, and Marmara) to constrain the deformation pattern of the study area. We use the most recent datasets to estimate the lithospheric structure. The models are evaluated in comparison with updated datasets of geodetic velocities and the most compressive horizontal principal stress azimuths. We find that the side and basal strengths drive the present-day motion of the Adria and Aegean Sea plates, whereas lithostatic pressure plays a key role in driving Anatolia. These findings provide new insights into the neotectonics of the greater Mediterranean region. Finally, the preferred model is used to estimate long-term shallow seismicity, which we retrospectively test against historical seismicity. As an alternative to reliance on incomplete geologic data or historical seismic catalogs, these neotectonic models help to forecast long-term seismicity, although requiring additional tuning before seismicity rates are used for seismic hazard purposes.
Maurice Ewing Series, 1977
Plate Boundary Zones, 2013
We use the Harvard CMT catalog to separate ocean-ridge seismicity into spreading and transform su... more We use the Harvard CMT catalog to separate ocean-ridge seismicity into spreading and transform sub-catalogs. We use the tapered Gutenberg-Richter distribution to estimate the total seismic moment rates of plate-boundary zones from limited catalogs of large events. We present the plate boundary model PB1999 and use it to associate marine earthquakes with particular plate boundary segments. We then combine these tools to estimate corner magnitudes (c m), spectral slopes (β), and coupled lithosphere thicknesses for all spreading ridges and oceanic transform faults. The distribution of spreading earthquakes is consistent with "normal" 2 3 β = (although β is not well constrained) and with uniform c 5.8 m =. Coupled lithosphere thickness along ridges decreases quasi-exponentially (from about 500 m to under 50 m) as spreading rate increases. Oceanic transform faults also have "normal" 2 3 β ≅ , but their corner magnitudes decrease from about 7.1 to about 6.3 with increasing relative plate velocity. Oceanic transform faults also show a quasi-exponential decrease in coupled lithosphere thickness (from about 3000 m to about 300 m) as relative plate velocity increases. Perhaps this is due to formation of serpentine along slow ridges and transforms and its absence from fast ridges and transforms. Spreading ridges and oceanic transform faults both have imperfect seismic coupling because: (i) all detailed local studies of seismogenic lithosphere thickness exceed our mean values for coupled thickness, and (ii) if coupling were perfect, and seismogenic lithosphere thickness were as small as our estimated coupled thickness, it would require unreasonable stress drops or rupture shapes to explain the moments of the largest earthquakes.
Bulletin of the Seismological Society of America, 2015
The 2014 Working Group on California Earthquake Probabilities (WGCEP 2014) presents time-dependen... more The 2014 Working Group on California Earthquake Probabilities (WGCEP 2014) presents time-dependent earthquake probabilities for the third Uniform California Earthquake Rupture Forecast (UCERF3). Building on the UCERF3 time-independent model published previously, renewal models are utilized to represent elasticrebound-implied probabilities. A new methodology has been developed that solves applicability issues in the previous approach for unsegmented models. The new methodology also supports magnitude-dependent aperiodicity and accounts for the historic open interval on faults that lack a date-of-last-event constraint. Epistemic uncertainties are represented with a logic tree, producing 5760 different forecasts. Results for a variety of evaluation metrics are presented, including logic-tree sensitivity analyses and comparisons to the previous model (UCERF2). For 30 yr M ≥ 6:7 probabilities, the most significant changes from UCERF2 are a threefold increase on the Calaveras fault and a threefold decrease on the San Jacinto fault. Such changes are due mostly to differences in the time-independent models (e.g., fault-slip rates), with relaxation of segmentation and inclusion of multifault ruptures being particularly influential. In fact, some UCERF2 faults were simply too long to produce M 6.7 size events given the segmentation assumptions in that study. Probability model differences are also influential, with the implied gains (relative to a Poisson model) being generally higher in UCERF3. Accounting for the historic open interval is one reason. Another is an effective 27% increase in the total elastic-rebound-model weight. The exact factors influencing differences between UCERF2 and UCERF3, as well as the relative importance of logic-tree branches, vary throughout the region and depend on the evaluation metric of interest. For example, M ≥ 6:7 probabilities may not be a good proxy for other hazard or loss measures. This sensitivity, coupled with the approximate nature of the model and known limitations, means the applicability of UCERF3 should be evaluated on a case-by-case basis.