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 out-performs 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, paleoseis-mic 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.
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 . 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. 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 ). 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.
1] Here, I present a new absolute plate motion model of the Earth's surface, determined from the ... more 1] Here, I present a new absolute plate motion model of the Earth's surface, determined from the alignment of present-day surface motions with 474 published shear wave (i.e., SKS) splitting orientations. When limited to oceanic islands and cratons, splitting orientations are assumed to reflect anisotropy in the asthenosphere caused by the differential motion between lithosphere and mesosphere. The best fit model predicts a 0.2065°/Ma counterclockwise net rotation of the lithosphere as a whole, which revolves around a pole at 57.6°S and 63.2°E. This net rotation is particularly well constrained by data on cratons and/or in the Indo-Atlantic region. The average data misfit is 19°and 24°for oceanic and cratonic areas, respectively, but the normalized root-mean-square misfits are about equal at 5.4 and 5.2. Predicted plate motions are very consistent with recent hot spot track azimuths (<8°on many plates), except for the slowest moving plates (Antarctica, Africa, and Eurasia). The difference in hot spot propagation vectors and plate velocities describes the motion of hot spots (i.e., their underlying plumes). For most hot spots that move significantly, the motions are considerably smaller than and antiparallel to the absolute plate velocity. Only when the origin depth of the plume is considered can the hot spot motions be explained in terms of mantle flow. The results are largely consistent with independent evidence of subasthenospheric mantle flow and asthenospheric return flow near spreading ridges. The results suggest that, at least where hot spots are, the lithosphere is decoupled from the mesosphere, including in western North America. Citation: Kreemer, C. (2009), Absolute plate motions constrained by shear wave splitting orientations with implications for hot spot motions and mantle flow,
We identify the key design aspects of a GPS-based system (and in the future, GNSS-based systems) ... more We identify the key design aspects of a GPS-based system (and in the future, GNSS-based systems) that could contribute to real-time earthquake source determination and tsunami warning systems. Our approach is based on models of both transient and permanent displacement of GPS stations caused by large earthquakes, while considering the effect of GPS errors on inverted earthquake source parameters. Our main conclusions are that (1) the spatial pattern, magnitude, and timing of permanent displacement of GPS stations can be inverted for the earthquake source and so predict the 3D displacement field of the ocean bottom, thus providing the initial conditions for tsunami models, and (2) there are no inherently limiting factors arising from real-time orbit and positioning errors, provided sufficient near-field GPS stations are deployed. This signal could be readily exploited by GPS networks currently in place, and will be facilitated by the IGS Real-Time Project as it comes to fruition.
We investigate possible driving forces behind the occurrence of the 1998 Antarctic plate earthqua... more We investigate possible driving forces behind the occurrence of the 1998 Antarctic plate earthquake. We determine first a regional strain rate field associated with the accommodation of relative plate motion, and second, a vertically averaged minimum deviatoric stress field associated with lithospheric gravitational potential energy differences and deglaciation of the Antarctic ice cap. We find that the mechanism of this event is inconsistent with strain orientations inferred from kinematic modeling of a diffuse zone of deformation within the triple junction region. Stress perturbations associated with deglaciation cannot be ruled out as a triggering mechanism for this event.
Static coseismic offsets >1 mm are observed up to 7800 km away from the great Sumatra-Andaman ear... more Static coseismic offsets >1 mm are observed up to 7800 km away from the great Sumatra-Andaman earthquake of 26 Dec. 2004 using global GPS network data. We investigate the rupture process based on far-field continuous GPS data. To reduce error in the coseismic offset estimates due to post-seismic deformation in the days following the main shock, we simultaneously fit a model of co-and postseismic offsets for nearby stations SAMP (500 km) and NTUS (900 km). The 3-month cumulative postseismic displacement for station SAMP amounts to 20% of the coseismic displacement, and can be well modeled by velocity-strengthening afterslip. We find that coseismic slip on the northern rupture segment is ∼3 m, which is consistent with seismic estimates. Our best estimate of the moment magnitude is M w = 9.13 if we take into account the expected increase of the shear modulus with depth (for uniform µ = 30 GPa, the moment-magnitude would only be 8.97). Our geodetic results, and thus our inferred rupture model, are different from a similar study using far-field data of . These differences highlight the challenge in earthquake studies on a global scale in terms of the sensitivity of far-field offset estimates to the analysis strategy and reference frame treatment. Our predicted coseismic offsets from this event are at least 1 mm across almost the entire globe. This warrants a reconsideration of how to maintain the global terrestrial reference frame affected by earthquakes of M w > 9.0.
ric synthetic aperture radar (InSAR) study, and paleoseismic data to evaluate the postseismic res... more ric synthetic aperture radar (InSAR) study, and paleoseismic data to evaluate the postseismic response of historic earthquakes in the Central Nevada seismic belt. We assume that GPS velocity has contributions from time-invariant (i.e., steady permanent crustal deformation) and transient (i.e., time varying and associated with the seismic cycle) processes that are attributable to postseismic viscoelastic relaxation of the crust and upper mantle. In order to infer the viscosity structure of Basin and Range lower crust, η LC , and upper mantle, η UM , we apply three objective criteria to identify rheological models that fi t both geodetic and geologic data. The model must (1) improve the apparent mismatch between geodetically and geologically inferred slip rates, (2) explain the InSAR-inferred vertical uplift rate, and (3) not imply timeinvariant contractions anywhere in the extending province. It is not required for the postseismic deformation fi eld to resemble the time-invariant velocity fi eld in pattern, rate, or style. We fi nd that the InSAR and horizontal GPS velocities form complementary constraints on the viscoelastic structure, excluding different parts of the model space. The best-fi tting model has a lower crust that is stronger than the uppermost mantle, with η LC = 10 20.5 Pa·s and η UM = 10 19 Pa·s, a fi nding consistent with the majority of similar studies in the Basin and Range. The best-fi tting viscosity model implies that the majority of Central Nevada seismic belt deformation is attributable to postseismic relaxation, and hence that western Basin and Range time-invariant deformation north of 39°N latitude is more tightly focused into the northern Walker Lane than would be inferred from uncorrected GPS velocities. However, signifi cant deformation remains after correction for postseismic effects, consistent with Central Nevada seismic belt faults slipping at rates intermediate between the Walker Lane belt and the central Basin and Range.
Eos, Transactions American Geophysical Union, 2005
The Global Strain Rate Map project II-8, initiated in 1998 by the International Lithosphere Progr... more The Global Strain Rate Map project II-8, initiated in 1998 by the International Lithosphere Program (ILP), provides constraints for understanding continental dynamics and for quantifying seismic hazards in general. To date, the Global Strain Rate Map (GSRM) model is a numerical velocity gradient tensor field solution (i.e.,spatial variations of horizontal strain rate tensor components and rotation rates) for the entire Earth surface [Kreemer et al., 2003].The global model consists of 25 rigid spherical plates and ˜25,000 0.6°×0.5° deformable grid areas within the diffuse plate boundary zones which lie between these plates (e.g., western North America, central Asia, the Alpine-Himalaya belt). This model provides an estimate of the horizontal strain rates, rotation rates, and velocity fields for the diffuse plate boundary zones as well as an estimate of the motions of the spherical caps.
In this paper we present a global model (GSRM-1) of both horizontal velocities on the Earth's sur... more In this paper we present a global model (GSRM-1) of both horizontal velocities on the Earth's surface and horizontal strain rates for almost all deforming plate boundary zones. A model strain rate field is obtained jointly with a global velocity field in the process of solving for a global velocity gradient tensor field. In our model we perform a least-squares fit between model velocities and observed geodetic velocities, as well as between model strain rates and observed geological strain rates. Model velocities and strain rates are interpolated over a spherical Earth using bi-cubic Bessel splines. We include 3000 geodetic velocities from 50 different, mostly published, studies. Geological strain rates are obtained for central Asia only and they are inferred from Quaternary fault slip rates. For all areas where no geological information is included a priori constraints are placed on the style and direction (but not magnitude) of the model strain rate field. These constraints are taken from a seismic strain rate field inferred from (normalized) focal mechanisms of shallow earthquakes. We present a global solution of the second invariant of the model strain rate field as well as strain rate solutions for a few selected plate boundary zones. Generally, the strain rate tensor field is consistent with geological and seismological data. With the assumption of plate rigidity for all areas other than the plate boundary zones we also present relative angular velocities for most plate pairs. We find that in general there is a good agreement between the present-day plate motions we obtain and long-term plate motions, but a few significant differences exist. The rotation rates for the Indian, Arabian and Nubian plates relative to Eurasia are 30, 13 and 50 per cent slower than the NUVEL-1A estimate, respectively, and the rotation rate for the Nazca Plate relative to South America is 17 per cent slower. On the other hand, Caribbean–North America motion is 76 per cent faster than the NUVEL-1A estimate. While crustal blocks in the India–Eurasia collision zone move significantly and self-consistently with respect to bounding plates, only a very small motion is predicted between the Nubian and Somalian plates. By integrating plate boundary zone deformation with the traditional modelling of angular velocities of rigid plates we have obtained a model that has already been proven valuable in, for instance, redefining a no-net-rotation model of surface motions and by confirming a global correlation between seismicity rates and tectonic moment rates along subduction zones and within zones of continental deformation.
We report on an initial assessment of a new method to target potential geothermal resources on th... more We report on an initial assessment of a new method to target potential geothermal resources on the regional scale. The method is based on seeking relationships between geologic structures and geodetic observations of regional tectonic strain. The working hypothesis is that geothermal plumbing systems might in some regions be controlled by fault planes acting as conduits that are continuously being stressed apart by tectonic activity. Specifically, this type of model would predict that (1) geothermal activity would be spatially correlated with areas of high inter-seismic strain accumulation, (2) especially when faults are favorably oriented with respect to the strain-rate tensor field. We tested these predictions in the Great Basin using tectonic velocity gradients derived from a network of Global Positioning System (GPS) sites, producing a regional strain-rate map with precision approaching a few parts per billion per year in some regions. Our initial conclusions are that regions of high inter-seismic strain accumulation correlate significantly with high geothermal well temperatures and the locations of geothermal power-producing sites. The next step is to test the prediction that fault orientation is also important, and that this information can be used to enhance the predictive capability of regional geodetic strain to target potential geothermal resources.
Previous studies have shown that ~5% of the Pacifi c-North America relative plate motion is accom... more Previous studies have shown that ~5% of the Pacifi c-North America relative plate motion is accommodated in the eastern part of the Great Basin (western United States). Near the Wasatch fault zone and other nearby faults, deformation is currently concentrated within a narrow zone of extension coincident with the eastern margin of the northern Basin and Range. Farther south, the pattern of active deformation implied by faulting and seismicity is more enigmatic. To assess how present-day strain is accommodated farther south and how this relates to the regional kinematics, we analyze data from continuous global positioning system (GPS) stations and model the strain rate tensor fi eld using the horizontal GPS velocities and earthquake focal mechanisms. The results indicate an ~100-km-wide zone of ~3.3 mm/yr extension at 40.5°N that broadens southward from the Wasatch fault zone to a width of >400 km at 36°N. This broadening involves at least one zone of localized extension in northwestern Arizona that encroaches into the southwestern plateau, and an eastnortheast-trending sinistral shear zone (the Pahranagat shear zone) through southern Nevada. This shear zone may accommodate as much as 1.8 mm/yr, and is a key feature that enables westward transfer of extension, thereby providing a kinematic connection between the Wasatch fault zone and the Eastern California shear zone.
1] The 26 December 2004 Sumatra earthquake (M w 9.2-9.3) generated the most deadly tsunami in his... more 1] The 26 December 2004 Sumatra earthquake (M w 9.2-9.3) generated the most deadly tsunami in history. Yet within the first hour, the true danger of a major oceanwide tsunami was not indicated by seismic magnitude estimates, which were far too low (M w 8.0-8.5). This problem relates to the inherent saturation of early seismicwave methods. Here we show that the earthquake's true size and tsunami potential can be determined using Global Positioning System (GPS) data up to only 15 min after earthquake initiation, by tracking the mean displacement of the Earth's surface associated with the arrival of seismic waves. Within minutes, displacements of >10 mm are detectable as far away as India, consistent with results using weeks of data after the event. These displacements imply M w 9.0 ± 0.1, indicating a high tsunami potential. This suggests existing GPS infrastructure could be developed into an effective component of tsunami warning systems.
Rapid GPS-based determination of earthquake displacement field and magnitude for tsunami propagation modeling and warning
Reliable tsunami early warning requires a rapid assessment of the tsunamigenic potential of an ea... more Reliable tsunami early warning requires a rapid assessment of the tsunamigenic potential of an earthquake as well as a prediction of the likely propagation pattern of the tsunami. Low-latency availability of the coseismic Earth's surface displacements can support the assessment of the tsunamigenic potential of an earthquake and improve predictions of the propagation pattern of the tsunami. We have developed a fingerprint methodology for the rapid determination of the surface displacement field from GNSS-determined displacements. The fingerprint methodology depends on a priori knowledge of the faults potentially involved in a rupture. The known faults are parametrized with standard elements and for each element so-called fingerprint functions are computed for unit strike and dip slips. After an event, the model space of all reasonable fault-element combinations is searched for the element-slip combination best fitting the observed displacements. This combination provides the best estimate of the displacement field and earthquake magnitude consistent with observations. Here we describe the fingerprint methodology and the main architectural elements of a prototype for the rapid determination of magnitude and displacement field.
One of the consequences of plate tectonics is that a spreading ridge will eventually approach a s... more One of the consequences of plate tectonics is that a spreading ridge will eventually approach a subduction zone. The problem whether the possible break-up of the approaching ridge will lead to the development of independent micro-plates, or not, is still unresolved. Some 4 million years ago the interaction between the Juan de Fuca Ridge and the Cascadia subduction zone resulted in ridge fragmentation in the Explorer region. There are two proposed post-Miocene kinematic models: one that proposes the presence of a micro-plate and the other that treats the region as a transform deformation zone, or so-called pseudo-plate. We use earthquake strain rates derived from 74 events since 1948 to estimate a long-term velocity field for the region. By comparing this result with the predicted velocity fields for both models we try to discriminate between the two. The earthquake strain rates indicate the presence of a transform deformation zone between the North American (NAM) and Pacific (PAC) plates. The velocity field derived from the inversion of the earthquake strain rates indicates that seismic activity takes up 50 š 30% .1¦ / of the PAC-NAM relative motion (NUVEL-1A; DeMets et al., 1994. Effect of recent revisions to the geomagnetic reversal time scale on estimates of current plate motions. Geophys. Res. Lett. 21, 2191Lett. 21, -2194 within the seismogenic layer of this zone and is indistinguishable in direction from the NUVEL-1A PAC-NAM model. The presence of this Explorer transform zone is consistent with the strain rate and velocity field for the 'pseudo-plate model' and indicates that seismicity defines a (new) plate boundary zone between the Pacific and North American plates. Earthquake-derived strain rates are low along the Nootka Transform, which accommodates relative motion between the Juan de Fuca and North American plates. The cause for the absence of significant seismic slip along this transform is unclear and may be closely linked to the fact that the adjacent Cascadia subduction zone is locked. The 'micro-plate model', which we reject, predicts SW-NE convergence in the eastern Explorer region and this is inconsistent with the earthquake strain rates as well as with surface fault observations. S 0 0 4 0 -1 9 5 1 ( 9 8 ) 0 0 0 8 9 -4
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 out-performs 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, paleoseis-mic 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.
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 . 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. 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 ). 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.
1] Here, I present a new absolute plate motion model of the Earth's surface, determined from the ... more 1] Here, I present a new absolute plate motion model of the Earth's surface, determined from the alignment of present-day surface motions with 474 published shear wave (i.e., SKS) splitting orientations. When limited to oceanic islands and cratons, splitting orientations are assumed to reflect anisotropy in the asthenosphere caused by the differential motion between lithosphere and mesosphere. The best fit model predicts a 0.2065°/Ma counterclockwise net rotation of the lithosphere as a whole, which revolves around a pole at 57.6°S and 63.2°E. This net rotation is particularly well constrained by data on cratons and/or in the Indo-Atlantic region. The average data misfit is 19°and 24°for oceanic and cratonic areas, respectively, but the normalized root-mean-square misfits are about equal at 5.4 and 5.2. Predicted plate motions are very consistent with recent hot spot track azimuths (<8°on many plates), except for the slowest moving plates (Antarctica, Africa, and Eurasia). The difference in hot spot propagation vectors and plate velocities describes the motion of hot spots (i.e., their underlying plumes). For most hot spots that move significantly, the motions are considerably smaller than and antiparallel to the absolute plate velocity. Only when the origin depth of the plume is considered can the hot spot motions be explained in terms of mantle flow. The results are largely consistent with independent evidence of subasthenospheric mantle flow and asthenospheric return flow near spreading ridges. The results suggest that, at least where hot spots are, the lithosphere is decoupled from the mesosphere, including in western North America. Citation: Kreemer, C. (2009), Absolute plate motions constrained by shear wave splitting orientations with implications for hot spot motions and mantle flow,
We identify the key design aspects of a GPS-based system (and in the future, GNSS-based systems) ... more We identify the key design aspects of a GPS-based system (and in the future, GNSS-based systems) that could contribute to real-time earthquake source determination and tsunami warning systems. Our approach is based on models of both transient and permanent displacement of GPS stations caused by large earthquakes, while considering the effect of GPS errors on inverted earthquake source parameters. Our main conclusions are that (1) the spatial pattern, magnitude, and timing of permanent displacement of GPS stations can be inverted for the earthquake source and so predict the 3D displacement field of the ocean bottom, thus providing the initial conditions for tsunami models, and (2) there are no inherently limiting factors arising from real-time orbit and positioning errors, provided sufficient near-field GPS stations are deployed. This signal could be readily exploited by GPS networks currently in place, and will be facilitated by the IGS Real-Time Project as it comes to fruition.
We investigate possible driving forces behind the occurrence of the 1998 Antarctic plate earthqua... more We investigate possible driving forces behind the occurrence of the 1998 Antarctic plate earthquake. We determine first a regional strain rate field associated with the accommodation of relative plate motion, and second, a vertically averaged minimum deviatoric stress field associated with lithospheric gravitational potential energy differences and deglaciation of the Antarctic ice cap. We find that the mechanism of this event is inconsistent with strain orientations inferred from kinematic modeling of a diffuse zone of deformation within the triple junction region. Stress perturbations associated with deglaciation cannot be ruled out as a triggering mechanism for this event.
Static coseismic offsets >1 mm are observed up to 7800 km away from the great Sumatra-Andaman ear... more Static coseismic offsets >1 mm are observed up to 7800 km away from the great Sumatra-Andaman earthquake of 26 Dec. 2004 using global GPS network data. We investigate the rupture process based on far-field continuous GPS data. To reduce error in the coseismic offset estimates due to post-seismic deformation in the days following the main shock, we simultaneously fit a model of co-and postseismic offsets for nearby stations SAMP (500 km) and NTUS (900 km). The 3-month cumulative postseismic displacement for station SAMP amounts to 20% of the coseismic displacement, and can be well modeled by velocity-strengthening afterslip. We find that coseismic slip on the northern rupture segment is ∼3 m, which is consistent with seismic estimates. Our best estimate of the moment magnitude is M w = 9.13 if we take into account the expected increase of the shear modulus with depth (for uniform µ = 30 GPa, the moment-magnitude would only be 8.97). Our geodetic results, and thus our inferred rupture model, are different from a similar study using far-field data of . These differences highlight the challenge in earthquake studies on a global scale in terms of the sensitivity of far-field offset estimates to the analysis strategy and reference frame treatment. Our predicted coseismic offsets from this event are at least 1 mm across almost the entire globe. This warrants a reconsideration of how to maintain the global terrestrial reference frame affected by earthquakes of M w > 9.0.
ric synthetic aperture radar (InSAR) study, and paleoseismic data to evaluate the postseismic res... more ric synthetic aperture radar (InSAR) study, and paleoseismic data to evaluate the postseismic response of historic earthquakes in the Central Nevada seismic belt. We assume that GPS velocity has contributions from time-invariant (i.e., steady permanent crustal deformation) and transient (i.e., time varying and associated with the seismic cycle) processes that are attributable to postseismic viscoelastic relaxation of the crust and upper mantle. In order to infer the viscosity structure of Basin and Range lower crust, η LC , and upper mantle, η UM , we apply three objective criteria to identify rheological models that fi t both geodetic and geologic data. The model must (1) improve the apparent mismatch between geodetically and geologically inferred slip rates, (2) explain the InSAR-inferred vertical uplift rate, and (3) not imply timeinvariant contractions anywhere in the extending province. It is not required for the postseismic deformation fi eld to resemble the time-invariant velocity fi eld in pattern, rate, or style. We fi nd that the InSAR and horizontal GPS velocities form complementary constraints on the viscoelastic structure, excluding different parts of the model space. The best-fi tting model has a lower crust that is stronger than the uppermost mantle, with η LC = 10 20.5 Pa·s and η UM = 10 19 Pa·s, a fi nding consistent with the majority of similar studies in the Basin and Range. The best-fi tting viscosity model implies that the majority of Central Nevada seismic belt deformation is attributable to postseismic relaxation, and hence that western Basin and Range time-invariant deformation north of 39°N latitude is more tightly focused into the northern Walker Lane than would be inferred from uncorrected GPS velocities. However, signifi cant deformation remains after correction for postseismic effects, consistent with Central Nevada seismic belt faults slipping at rates intermediate between the Walker Lane belt and the central Basin and Range.
Eos, Transactions American Geophysical Union, 2005
The Global Strain Rate Map project II-8, initiated in 1998 by the International Lithosphere Progr... more The Global Strain Rate Map project II-8, initiated in 1998 by the International Lithosphere Program (ILP), provides constraints for understanding continental dynamics and for quantifying seismic hazards in general. To date, the Global Strain Rate Map (GSRM) model is a numerical velocity gradient tensor field solution (i.e.,spatial variations of horizontal strain rate tensor components and rotation rates) for the entire Earth surface [Kreemer et al., 2003].The global model consists of 25 rigid spherical plates and ˜25,000 0.6°×0.5° deformable grid areas within the diffuse plate boundary zones which lie between these plates (e.g., western North America, central Asia, the Alpine-Himalaya belt). This model provides an estimate of the horizontal strain rates, rotation rates, and velocity fields for the diffuse plate boundary zones as well as an estimate of the motions of the spherical caps.
In this paper we present a global model (GSRM-1) of both horizontal velocities on the Earth's sur... more In this paper we present a global model (GSRM-1) of both horizontal velocities on the Earth's surface and horizontal strain rates for almost all deforming plate boundary zones. A model strain rate field is obtained jointly with a global velocity field in the process of solving for a global velocity gradient tensor field. In our model we perform a least-squares fit between model velocities and observed geodetic velocities, as well as between model strain rates and observed geological strain rates. Model velocities and strain rates are interpolated over a spherical Earth using bi-cubic Bessel splines. We include 3000 geodetic velocities from 50 different, mostly published, studies. Geological strain rates are obtained for central Asia only and they are inferred from Quaternary fault slip rates. For all areas where no geological information is included a priori constraints are placed on the style and direction (but not magnitude) of the model strain rate field. These constraints are taken from a seismic strain rate field inferred from (normalized) focal mechanisms of shallow earthquakes. We present a global solution of the second invariant of the model strain rate field as well as strain rate solutions for a few selected plate boundary zones. Generally, the strain rate tensor field is consistent with geological and seismological data. With the assumption of plate rigidity for all areas other than the plate boundary zones we also present relative angular velocities for most plate pairs. We find that in general there is a good agreement between the present-day plate motions we obtain and long-term plate motions, but a few significant differences exist. The rotation rates for the Indian, Arabian and Nubian plates relative to Eurasia are 30, 13 and 50 per cent slower than the NUVEL-1A estimate, respectively, and the rotation rate for the Nazca Plate relative to South America is 17 per cent slower. On the other hand, Caribbean–North America motion is 76 per cent faster than the NUVEL-1A estimate. While crustal blocks in the India–Eurasia collision zone move significantly and self-consistently with respect to bounding plates, only a very small motion is predicted between the Nubian and Somalian plates. By integrating plate boundary zone deformation with the traditional modelling of angular velocities of rigid plates we have obtained a model that has already been proven valuable in, for instance, redefining a no-net-rotation model of surface motions and by confirming a global correlation between seismicity rates and tectonic moment rates along subduction zones and within zones of continental deformation.
We report on an initial assessment of a new method to target potential geothermal resources on th... more We report on an initial assessment of a new method to target potential geothermal resources on the regional scale. The method is based on seeking relationships between geologic structures and geodetic observations of regional tectonic strain. The working hypothesis is that geothermal plumbing systems might in some regions be controlled by fault planes acting as conduits that are continuously being stressed apart by tectonic activity. Specifically, this type of model would predict that (1) geothermal activity would be spatially correlated with areas of high inter-seismic strain accumulation, (2) especially when faults are favorably oriented with respect to the strain-rate tensor field. We tested these predictions in the Great Basin using tectonic velocity gradients derived from a network of Global Positioning System (GPS) sites, producing a regional strain-rate map with precision approaching a few parts per billion per year in some regions. Our initial conclusions are that regions of high inter-seismic strain accumulation correlate significantly with high geothermal well temperatures and the locations of geothermal power-producing sites. The next step is to test the prediction that fault orientation is also important, and that this information can be used to enhance the predictive capability of regional geodetic strain to target potential geothermal resources.
Previous studies have shown that ~5% of the Pacifi c-North America relative plate motion is accom... more Previous studies have shown that ~5% of the Pacifi c-North America relative plate motion is accommodated in the eastern part of the Great Basin (western United States). Near the Wasatch fault zone and other nearby faults, deformation is currently concentrated within a narrow zone of extension coincident with the eastern margin of the northern Basin and Range. Farther south, the pattern of active deformation implied by faulting and seismicity is more enigmatic. To assess how present-day strain is accommodated farther south and how this relates to the regional kinematics, we analyze data from continuous global positioning system (GPS) stations and model the strain rate tensor fi eld using the horizontal GPS velocities and earthquake focal mechanisms. The results indicate an ~100-km-wide zone of ~3.3 mm/yr extension at 40.5°N that broadens southward from the Wasatch fault zone to a width of >400 km at 36°N. This broadening involves at least one zone of localized extension in northwestern Arizona that encroaches into the southwestern plateau, and an eastnortheast-trending sinistral shear zone (the Pahranagat shear zone) through southern Nevada. This shear zone may accommodate as much as 1.8 mm/yr, and is a key feature that enables westward transfer of extension, thereby providing a kinematic connection between the Wasatch fault zone and the Eastern California shear zone.
1] The 26 December 2004 Sumatra earthquake (M w 9.2-9.3) generated the most deadly tsunami in his... more 1] The 26 December 2004 Sumatra earthquake (M w 9.2-9.3) generated the most deadly tsunami in history. Yet within the first hour, the true danger of a major oceanwide tsunami was not indicated by seismic magnitude estimates, which were far too low (M w 8.0-8.5). This problem relates to the inherent saturation of early seismicwave methods. Here we show that the earthquake's true size and tsunami potential can be determined using Global Positioning System (GPS) data up to only 15 min after earthquake initiation, by tracking the mean displacement of the Earth's surface associated with the arrival of seismic waves. Within minutes, displacements of >10 mm are detectable as far away as India, consistent with results using weeks of data after the event. These displacements imply M w 9.0 ± 0.1, indicating a high tsunami potential. This suggests existing GPS infrastructure could be developed into an effective component of tsunami warning systems.
Rapid GPS-based determination of earthquake displacement field and magnitude for tsunami propagation modeling and warning
Reliable tsunami early warning requires a rapid assessment of the tsunamigenic potential of an ea... more Reliable tsunami early warning requires a rapid assessment of the tsunamigenic potential of an earthquake as well as a prediction of the likely propagation pattern of the tsunami. Low-latency availability of the coseismic Earth's surface displacements can support the assessment of the tsunamigenic potential of an earthquake and improve predictions of the propagation pattern of the tsunami. We have developed a fingerprint methodology for the rapid determination of the surface displacement field from GNSS-determined displacements. The fingerprint methodology depends on a priori knowledge of the faults potentially involved in a rupture. The known faults are parametrized with standard elements and for each element so-called fingerprint functions are computed for unit strike and dip slips. After an event, the model space of all reasonable fault-element combinations is searched for the element-slip combination best fitting the observed displacements. This combination provides the best estimate of the displacement field and earthquake magnitude consistent with observations. Here we describe the fingerprint methodology and the main architectural elements of a prototype for the rapid determination of magnitude and displacement field.
One of the consequences of plate tectonics is that a spreading ridge will eventually approach a s... more One of the consequences of plate tectonics is that a spreading ridge will eventually approach a subduction zone. The problem whether the possible break-up of the approaching ridge will lead to the development of independent micro-plates, or not, is still unresolved. Some 4 million years ago the interaction between the Juan de Fuca Ridge and the Cascadia subduction zone resulted in ridge fragmentation in the Explorer region. There are two proposed post-Miocene kinematic models: one that proposes the presence of a micro-plate and the other that treats the region as a transform deformation zone, or so-called pseudo-plate. We use earthquake strain rates derived from 74 events since 1948 to estimate a long-term velocity field for the region. By comparing this result with the predicted velocity fields for both models we try to discriminate between the two. The earthquake strain rates indicate the presence of a transform deformation zone between the North American (NAM) and Pacific (PAC) plates. The velocity field derived from the inversion of the earthquake strain rates indicates that seismic activity takes up 50 š 30% .1¦ / of the PAC-NAM relative motion (NUVEL-1A; DeMets et al., 1994. Effect of recent revisions to the geomagnetic reversal time scale on estimates of current plate motions. Geophys. Res. Lett. 21, 2191Lett. 21, -2194 within the seismogenic layer of this zone and is indistinguishable in direction from the NUVEL-1A PAC-NAM model. The presence of this Explorer transform zone is consistent with the strain rate and velocity field for the 'pseudo-plate model' and indicates that seismicity defines a (new) plate boundary zone between the Pacific and North American plates. Earthquake-derived strain rates are low along the Nootka Transform, which accommodates relative motion between the Juan de Fuca and North American plates. The cause for the absence of significant seismic slip along this transform is unclear and may be closely linked to the fact that the adjacent Cascadia subduction zone is locked. The 'micro-plate model', which we reject, predicts SW-NE convergence in the eastern Explorer region and this is inconsistent with the earthquake strain rates as well as with surface fault observations. S 0 0 4 0 -1 9 5 1 ( 9 8 ) 0 0 0 8 9 -4
Uploads
Papers by Corne Kreemer