S23E-01
The USGS National Earthquake Information Center's Response to the Wenchuan, China Earthquake
Immediately after detecting the May 12th, 2008 Mw 7.9 Wenchuan Earthquake, the USGS National
Earthquake Information Center (NEIC) began a coordinated effort to understand and communicate the
earthquake's seismological characteristics, tectonic context, and humanitarian impact. NEIC's initial estimates
of magnitude and location were distributed within 30 minutes of the quake by e-mail and text message to
70,000 users via the Earthquake Notification System. The release of these basic parameters automatically
triggered the generation of more sophisticated derivative products that were used by relief and government
agencies to plan their humanitarian response to the disaster. Body-wave and centroid moment tensors
identified the earthquake's mechanism. Predictive ShakeMaps provided the first estimates of the geographic
extent and amplitude of shaking. The initial automated population exposure estimate generated and
distributed by the Prompt Assessment of Global Earthquakes for Response (PAGER) system stated that 1.2
million people were exposed to severe-to-extreme shaking (Modified Mercalli Intensity VIII or greater),
indicating a large-scale disaster had occurred. NEIC's modeling of the mainshock and aftershocks was
continuously refined and expanded. The length and orientation of the fault were determined from
aftershocks, finite-fault models, and back-projection source imaging. Firsthand accounts of shaking intensity
were collected and mapped by the "Did You Feel It" system. These results were used to refine our
ShakeMaps and PAGER exposure estimates providing a more accurate assessment of the extent and
enormity of the disaster. The products were organized and distributed in an event-specific summary poster
and via the USGS Earthquake Program web pages where they were viewed by millions and reproduced by
major media outlets (over 1/2 billion hits were served that month). Rather than just a point showing
magnitude and epicenter, several of the media's schematic maps showed both intensity distribution and
population exposure, achieving a significant communication goal of the PAGER project and the Advanced
National Seismic System (ANSS).
http://earthquake.usgs.gov/eqcenter/recenteqsww/Quakes/us2008ryan.php
S23E-02
Rupture history of 2008 May 12 Mw 8.0 Wen-Chuan earthquake: Evidence of slip interaction
We will present the rupture process of the May 12, 2008 Mw 8.0 Wenchuan earthquake using all available data. The current model, using both teleseismic body and surface waves and interferometric LOS displacements, reveals an unprecedented complex rupture process which can not be resolved using either of the datasets individually. Rupture of this earthquake involved both the low angle Pengguan fault and the high angle Beichuan fault, which intersect each other at depth and are separated approximately 5-15 km at the surface. Rupture initiated on the Pengguan fault and triggered rupture on the Beichuan fault 10 sec later. The two faults dynamically interacted and unilaterally ruptured over 270 km with an average rupture velocity of 3.0 km/sec. The total seismic moment is 1.1x1021 Nm (Mw 8.0), roughly equally partitioned between the two faults. However, the spatiotemporal evaluations of the two faults are very different. This study will focus on the evidence for fault interactions and will analyze the corresponding uncertainties, in preparation for future dynamic studies of the same detailed nature.
S23E-03 INVITED
Dense Seismic Array Observation in the Wenchuan M 8.0 Earthquake Region
Beginning in October, 2006, a high-density passive seismic array consisted of 300 portable broadband seismometers was deployed in Western Sichuan region(the Chuan-xi seismic array) and will continue to operate until the first half of 2009. Western Sichuan and its adjacent region is one of the most important regions for studying the dynamic process between the Tibetan Plateau and its surrounding tectonic blocks. It is also an ideal region for understanding the dynamic process of large earthquakes at active plate boundaries. On May 12, 2008, the great Wenchuan Ms 8.0 earthquake occurred on the Longmen Shan fault zone during our seismic observation, within the northeast corner of the Chuan-xi array. This provides a rare opportunity to penetrate deeply on the genesis and occurring process of strong earthquakes. The stations of the Chuan-xi array are equipped with Reftek 130B digitizers and Guralp CMG-3ESPC broadband sensors with the frequency range from 0.0167 to 40Hz. 40 sps continuous and 100 sps triggered recording is configured to work at the same time. In the Wenchuan earthquake region(29° ~ 32° E , 100° ~ 105° N), the average station spacing along main profiles is about 20km, and the distance from the nearest station to the epicenter of the main shock of the Wenchuan M 8.0 earthquake is only several kilometers. Using the data collected by the Chuan-xi array before and after the Wenchuan earthquake, structural background for the Wenchuan earthquake region and the earthquake tectonic is obtained. Teleseismic receiver functions for all the stations are estimated from the waveform data and utilized to both the grid scan and receiver function inversion methods for crustal thickness, average Poisson's ratio and crustal velocity structures. Results show an abruptly Moho depth and velocity structure variation on the two sides of the Longmenshan fault. Strong lateral velocity structure variation and Poisson's ratio anomaly along the Longmenshan is also found near where the main shock of the Wenchuan earthquake occurred. Relocation result of the Wenchuan earthquake sequence shows closely relationship between the source distribution and the deep structure.
S23E-04
Imaging the Wenchuan Earthquake and its aftershocks using back-projection of Teleseismic P-Waves and Point-Spread Function Deconvolution
Applications of seismic wave back-projection and reverse-time methods to earthquake finite-source rupture imaging have increased with ready availability of large digital data sets and expanded computer processing capabilities. For large earthquakes, these approaches offer potential for explicitly resolving rupture attributes that are treated parametrically in conventional modeling and inversion procedures. Teleseismic P-wave back- projection source imaging using large aperture networks in Alaska, Europe and Japan is applied to the May 12, 2008 Wenchuan earthquake. Our images show clear rupture propagation towards the northeast extending for approximately 220 km. This computationally fast method uses information contained in the direct teleseismic P-waveform, an important attribute for real time monitoring applications. However, time resolution, network aperture, rate of wave slowness variation across the network, and signal coherence are key issues that limit imaging resolution. Back-projected wavefields have many space-time artifacts due to intrinsically limited resolution associated with the data acquisition geometry and seismic wave periods relative to wavefront curvature. Complexity of the back-projected wavefield for a large event produced by the source-receiver geometry can be reduced by deconvolving empirical or theoretical point-spread functions (the space-time image formed from back-projection of signals from the same network for an effective point- source) from the data images. This facilitates measurement in back-projection images of key rupture attributes such as average rupture velocity and minimum rupture extent. These faulting characteristics can then be used for rapid assessment of shaking and provide a priori constraints on finite-fault slip model inversions. Within only a few hours of the main shock, hundreds of aftershocks followed. Identifying aftershocks following a large earthquake is a significant challenge for real time monitoring applications. We modified the back-projection technique to act as an earthquake detector, similar to spotlight techniques used for CTBT monitoring. We show that this computationally fast method successfully identifies most aftershocks in the USGS earthquake catalog as well as some that are not in the catalog.
S23E-05 INVITED
Heterogeneity and Azimuthal Anisotropy of the Crust in SE Tibet and Sichuan Province from Surface Wave Array Tomography
On 12 May 2008 a destructive magnitude 7.9 earthquake hit Wenchuan, Sichuan Province, SW China. The main shock and many aftershocks occurred along the Longmenshan fault zone, which marks the eastern Tibetan Plateau margin and the boundary with Sichuan Basin. With data from temporary seismograph arrays and local permanent stations in SE Tibet we have performed high-resolution surface wave array tomography of the crust and uppermost mantle in SW China. Our tomographic images of heterogeneity and anisotropy put important constraints on the structural framework of the crust near the May 12, 2008 epicenters and will help improve our understanding of the deformation and (tectonic) block motions in this earthquake prone area. Using both ambient noise and teleseismic surface wave data, we performed high-resolution surface wave array tomography in SE Tibet. The horizontal resolution is about 100 km in the crust and the data can resolve wavespeed variations and azimuthal anisotropy in the upper, middle, and lower crust. Because of the possible effect of mechanically weak layers on regional deformation, of our particular interest is the existence and geometry of low (shear) velocity layers (LVLs). In some regions prominent LVLs occur in the middle crust, in others they may appear in the lower crust. In some cases the lateral transition of shear wavespeed coincides with major fault zones. Azimuthal anisotropy of the upper crust reveals a pattern that is similar to the surface strain field and which is consistent with clockwise rotation of crustal material around the eastern Himalayan syntaxis. At larger depth the pattern of azimuthal anisotropy is different from that of shallow crust, implying changes of deformation pattern from the crust to the upper mantle. The widespread crustal LVLs as well as the clockwise rotational pattern of azimuthal anisotropy support models of ductile flow of crustal material beneath SE Tibet. However, the spatial variation in strength and depth of crustal LVLs and in the pattern of azimuthal anisotropy suggests that the 3-D geometry of crustal weak layers is complex and that unhindered crustal flow over large regions may not occur. Consideration of such complexity may be the key to a better understanding of relative block motion and patterns of regional seismicity. In close collaboration with colleagues in China (at China Earthquake Administration and Sichuan Province Earthquake Administration) we are in the process of analyzing data from the (permanent) Sichuan Province seismograph network and temporary array deployments in the region, with the objective to establish relationships among the crustal structure, surface deformation, regional tectonics, and local seismicity in SW China.
S23E-06
Rupture Propagation and Near Field Ground Shaking From Dynamic Models of the 12 May 2008 Ms 8.0 Wenchun (China) Earthquake
Surface rupture of the 12May 2008 Ms 8.0 Wenchun (China) earthquake shows that the earthquake occurred at least on two thrust faults, which may merge at depths. Waveform inversions by several research groups indicate that fault slip is thrust near the hypocenter and it is oblique (thrust and right-lateral strike slip) to the north. The intensity map released by China Earthquake Administration demonstrates that destruction along the surface rupture is not uniform from south to north, and the decay of intensity away from the faults is faster to the east than to the west. To understand the complex rupture process of the event and resultant destruction pattern, we build dynamic rupture and wave propagation models of the earthquake using a finite element method. As the first attempt, we simulate rupture propagation on a shallow dipping (33 degree) fault and wave propagation in a homogeneous medium with flat topography. We use a depth-dependent normal stress distribution (e.g., increase in normal stress with depth) and assign shear stress on the fault comparable with changes in the rake of fault slip along the fault strike indicated by kinematic inversions. A slip-weakening friction law is used to govern the rupture propagation on the fault. Preliminary results from this simplified dynamic model show several features that are qualitatively consistent with observations of surface rupture and the intensity distribution. In particular, major variations in surface rupture appear to be associated with changes in faulting mechanisms along the fault strike, and different decay rates of intensity away from the faults between the two directions can be explained by difference in ground motion variations on the two walls of the shallow dipping fault. We are working on incorporating more complexities consistent with surface rupture and geologic structure into dynamic models of the earthquake to better understand the rupture process and resultant ground shaking.
S23E-07 INVITED
Six months later: Testing the Coulomb stress change model by examining calculations made immediately after the 12 May, 2008 Ms=8.0 Wenchuan earthquake
On the 12th of May, 2008 a devastating Ms=8.0 earthquake struck the eastern edge of the Tibetan Plateau, collapsing buildings and killing thousands in major cities aligned along the western Sichuan basin in China. After a high-magnitude earthquake like the 12 May event, rearrangement of stresses in the crust commonly causes subsequent damaging earthquakes. The Sichuan basin and surroundings are crossed by major active strike-slip and thrust faults. By 72 hours after the earthquake, coseismic stress changes were calculated on models of those faults, with many showing significant stress increases. Rapid mapping of stress changes was intended to locate fault sections with relatively higher odds of producing the largest aftershocks and to enable prospective testing of the static-stress triggering hypothesis. A recent prospective test of the method was conducted by McCloskey et al. [2005] after the great 2004 Sumatra earthquake, and was validated by a M=8.7 shock that struck three months later in a region calculated to have been stressed by the mainshock. Our test begins at the time peer review was completed, 38 days after the mainshock on 19 June, 2008. Thus aftershocks occurring between that time and the present can be used for prospective testing. As of this writing, in our test region magnitude greater than 4.0 aftershocks have been largely confined to the mainshock rupture zone, with virtually no activity on Sichuan basin faults with calculated stress increases. Examination of magnitude-frequency behavior of the aftershocks suggests either a corner magnitude at about magnitude 6, or a deficiency in the magnitude greater than 6 range. This experiment is ongoing, and time will tell if the Coulomb model is confirmed in the Sichuan region; our conclusion at present is that there has been no validation, and that use of a generalized aftershock forecast model would have been sufficient.
S23E-08
Global survey of earthquakes and non-volcanic tremor triggered by the 2008 Mw7.9 Wenchuan earthquake
We perform a global survey of triggered earthquakes and non-volcanic tremor by the 2008 Mw7.9 Wenchuan earthquake. The analyzed data is obtained from the Global Seismic Network and various local and regional seismic networks around the world. We identify triggered earthquakes as impulsive seismic energies with clear P and S arrivals on 5 Hz high-pass-filtered three-component velocity seismograms, and triggered tremor as bursts of high-frequency, non-impulsive seismic energies that are coherent among many stations and during the passage of teleseismic body and surface waves. We find wide-spread triggering of regular earthquakes within mainland China and elsewhere in the world. The triggered earthquakes mostly occur in tectonically active regions in northwest and northeast China. However, we also find clear evidence of triggered earthquakes in southeast China that is not tectonically active. Our observations are consistent with previous studies of earthquake triggering (e.g., Gomberg et al., 2004; Velasco et al., 2008), indicating that dynamic triggering of earthquakes is ubiquitous and independent of the tectonic environments. In comparison, clear triggered tremor associated with the Wenchuan earthquake is found in the Taiwan Island (Chao and Peng, 2008), southwest Japan, Cascadia (Vidale et al., 2008), and around the Parkfield section of the San Andreas fault (Peng et al., 2008), where regular and/or triggered tremor has been found before. So far we have not found clear evidence of triggered tremor within mainland China. At least part of the reason could be due to severe clippings of the broadband waveforms during large-amplitude surface waves for many stations within 2000 km of the epicenter. Updated results will be presented at the meeting.