U21C-01 INVITED
A geodynamic and tectonic framework for the Wenchuan Earthquake of May 12, 2008
The Wenchuan earthquake of May 12, 2008 occurred on a west-dipping reverse fault with a pronounced right-slip component. The epicenter was located along the steep eastern edge of the Tibetan plateau. Such earthquakes with thrust mechanisms are typical of active orogenic systems, but many geological features of this region are atypical of active orogenic settings where there is steep, recently developed, high topography and thick crust. These include a small magnitude of young (Late Cenozoic) upper crustal shortening observed and the slow GPS-determined rates of convergence. In our interpretation, the earthquake of May 12 probably reflects long-term uplift of the eastern plateau relative to the Sichuan foreland to the east, without substantial upper crustal shortening, probably several tens of kilometers or less. This is consistent with the interpretation that this margin of the plateau, as well as most of eastern Tibet, has been uplifted by post ~15 Ma thickening and shortening of a weak middle to lower crust beneath a largely undeforming upper crust. We suggest that the Wenchuan earthquake fault continues at depth into the weak (low viscosity) of deep crustal thickening, and that the earthquake occurred in response to near vertical uplift of surface layers as an isostatic response to thickening in the deeper crust.
U21C-02 INVITED
Geologic Setting of the May 12, 2008, Sichuan Earhquake
The May 12, 2008, Sichuan earthquake was located in the Longmen Shan, mountains that form the eastern margin of the Tibetan plateau and border the western side of the Sichuan basin. The region of the Longmen Shan consists of very complex structures that vary greatly along strike. Geologically, the Longmen Shan lie along the western margin of the Yangtze craton and was first deformed in late Triassic-early Jurassic (Indosinian) time when the margin was telescoped by large scale eastward thrusting and development of an adjacent Mesozoic foredeep. These units can be followed from their join with the Qinling Mountains in the north, south into central Yunnan. Although there are suggestions of younger Mesozoic deformation, the next major deformation that can be documented is late Cenozoic (beginning ~12 Ma) with the formation of diverse structures. South of the Sichuan basin folds and thrust faults form a multiply deformed area that involves Precambrian basement rocks and Phanerozoic sedimentary rocks with a modified dome and basin interference pattern. These folds and thrust fault project northward into the southern Sichuan basin where most of them end within the basin, but the Longchuan anticline on the east and several folds on the west continue north; the Longchuan anticline curves west into the middle of the Longmen Shan and the western structures form the eastern margin of the NE-trending Longmen Shan. At this latitude the structures involve only Mesozoic rocks above a decollement that underlies the Chengdu plain, but ramps down to the west beneath Precambrian cored structures that form the core of the Longmen Shan. These basement involved structures have a modified fault propagation fold geometry. Northward, the margin of the Tibetan plateau curves to the west along the N-S trending Min Shan formed by active east-vergent steep thrust faults. These structures may be localized by crustal anisotropy between the Xue Shan plateau and Song Pan Ganze flysch basin. A lower range continues northeast along the west side of the Sichuan basin into the Qinling Mountains and is marked by transpressional right-lateral strike-slip faults. The May 12 earthquake was localized along the ramp part of the thrusts beneath the Longmen Shan and after shocks propagated eastward where they have an increasingly greater right-lateral strike-slip component. The epicenter is located where greater shortening may have been localized because of the northward ending of structures to the east. The Longmen Shan was not considered a major seismic hazard because: 1. They show little horizontal shortening, 2. Lack a foredeep related to shortening, and 3. GPS data show very slow structure normal shortening and strike-slip (~1 mm/yr) confirmed by neotectonic studies of Densmore et al. (2007). However, closer examination suggests the area may have been a seismic gap for large earthquakes and the very steep eastern Longmen Shan front is suggestive of very active uplift. The Sichuan earthquake suggests that we need to examine (or reexamine) carefully criteria for seismic hazards.
U21C-03
Widespread low velocity zone in mid-lower crust in Tibetan Plateau: Implications for crustal channel flow model
The question of how the Tibetan Plateau (TP) grows and deforms after the India-Eurasia collision is both a fundamental one and one that has direct implications for seismic hazards in the eastern and southeastern margin with a large population. One model is the growth by deep crustal channel flow in which the upper crust deformation is decoupled from deformation at depth. The crustal channel flow coupled with ductile extrusion and surface denudation was also proposed to explain the exhumation of high-grade metamorphic sequence observed in the Himalaya. Studies in the last decade have shown evidence for mid-lower crust melting in some parts of the TP. However, the idea has been hotly debated with inconsistent observations and the seismic observations have been made only along certain profiles or small areas. Here we show evidence for wide spread mid-lower crust slow anomalies in the central and eastern parts of the TP (The western part is not well sampled). We have recently obtained a new 3D shear-wave velocity model of China from inverting surface wave dispersion maps obtained from ambient noise tomography. One of the most remarkable features in our 3D model is the extremely slow S velocities in the mid-crust in much of the TP (central, eastern, and southeastern regions). These slow velocities are generally connected and seem to reach to the surface at certain localities in western, southern, and southeastern margins. In a separate inversion with P travel times, we found that mid and lower crust in eastern TP has slow P velocities compared with Sichuan basin and Yangtze block. However, P velocities are not as anomalous as the S velocities in the eastern TP, suggesting anomalous Poisson's ratios in the mid-lower crust. These observations are consistent with the crustal channel flow throughout much of the TP. However, the deformation in the upper crust may not be decoupled from the mantle as the patterns of surface deformation and anisotropy from S splitting and Pn waves in eastern and southern TP seem to be coherent.
U21C-04
Active tectonics of eastern Tibet and the Szechwan Basin: the importance of simple vs. pure shear deformation
Significant variations in surface morphology and tectonic style are present along the eastern margin of the Tibetan Plateau, which are likely to be due to differences in the lower boundary condition imposed on the deformation within the crust. Where the thick, strong lithosphere of the Szechwan Basin bounds the eastern margin of the Tibetan plateau, vertical planes deform by simple shear as material moves over the rigid basin, resulting in relatively low surface velocities, a steep topographic front, and thrust faulting in the brittle upper crust (such as the 12th May 2008 event). To the north and south, where the lowlands appear to be weaker, vertical planes can deform by pure shear, which means that the surface gradient is more gentle and the surface velocities are larger (as observed in GPS data). The variations in active tectonics and the evolution of topography in eastern Tibet can therefore be understood in terms of differences in the geological history of the lowlands on the plateau margins.
U21C-05
PALEOSEISMOLOGIC STUDY OF BEICHUAN-YINGXIU FAULT ON YINGXIU TOWN, WENCHUAN COUNTY OF SICHUAN, CHINA
Yingxiu Town is one of the epicenters of the Great 2008 Wenchuan Earthquake. Some work on paleoseismology and geomorphology is in process in order to find out the seismic period of the Central Fault (Beichuan-Yingxiu Fault), which is causative structure of the earthquake. The seismic period can be well limited with the help of chronology samples acquired from the tectonic terraces and the trench which stretches across the structure scarp at Yingxiu. At the same time, age of the third terrace (T3) of Minjiang River can be fixed by pollen sample sequence. The uplift amount of T2 and T3 can be acquired by geomorphologic surveying, while T1 is only displaced by the Wenchuan Earthquake. Consequently, vertical slip rate of the Central Fault can be obtained, and paleoseismologic sequence will be established after further research on the trench. Deformation characteristics of thrust fault, that flexure of strata is very obvious while fracture surface is rare even when seismic uplift amount is large, are very different from normal fault and slip fault. The trenches at Leigu Town, Pintong Town where research is done by others also testify it.
U21C-06
Magnitudes of Cenozoic Upper-crustal Shortening across the Frontal Zone of the Longmen Shan Thrust Belt: Implications for the Development of the Eastern Tibetan Plateau
Although the Ms 7.9 Wenchuan earthquake of May 12, 2008 is a clear expression of upper-crustal deformation, the total magnitude of Cenozoic upper-crustal shortening across the 150-200 km wide Longmen Shan thrust belt remains poorly known. Balanced cross-sections of its frontal zone based on subsurface data suggest ~30-km or 40-50% shortening. As the thrust zone started at 30-10 Ma, our shortening estimate implies a long-term shortening rate of 1-3 mm/yr, comparable to the short-term GPS rate across the Longmen Shan front that accounts for about 20% of the total 15-mm/yr active shortening across the whole thrust belt. The consistency between the long- and short-term shortening rates implies steady-state development of the Longmen Shan thrust belt over the past 30-10 m.y., which in turn requires >150-km (or >50%) upper-crustal shortening for the development of the Longmen Shan thrust belt. The inferred >150-km total shortening may have been accommodated by both active and recently active Cenozoic thrusts and folds within the thrust belt, most of which involve only pre-Cenozoic strata such as the Yingxiu- Beichuan thrust along which the Wenchuan earthquake occurred. Discounting contributions from inactive Cenozoic structures within pre-Cenozoic strata can lead to severe underestimates of actual upper-crustal shortening across the Longmen Shan thrust belt. Our inferred large magnitude of shortening across the Longmen Shan leads to the often-asked question: why is there no pronounced Cenozoic foreland sedimentation in the western Sichuan basin? We attribute this to three factors: (1) an extremely low flexural rigidity across the Longmen Shan thrust belt that prevents effective transfer of its load onto the stronger Sichuan block and thus failing to produce a pronounced foredeep, (2) the development of a Cenozoic passive-roof thrust system and the remnant Mesozoic topography have caused basin-ward tilting of the western Sichuan block that hinders the formation of a topographic depression in the Longmen Shan foreland, and (3) fluvial systems traversing the western Sichuan basin are all externally drained transverse rivers (cf., the longitudinal Ganges in the Himalayan foreland), making it difficult to trap sediments eroded from the Tibetan plateau in the foreland. Our tentative conclusion is that the Longmen Shan thrust belt may have deformed at a relatively slow rate, but its total magnitude of crustal shortening is appreciable and possibly sufficient to explain its current topography.
U21C-07
Seismic moment balance in the Ganze-Songpan region, eastern Tibet, and implications for the great Wenchuan earthquake
The 12 May 2008 Wenchuan earthquake (Mw 7.9), which killed more than 80,000 people and injured millions, occurred on the Longmenshan fault zone, where the slip rates are low and earthquakes infrequent in comparison with other major fault zones in the Ganze-Songpan region of eastern Tibet. We have studied the cause of the great Wenchuan earthquake and the potential seismic hazard in the surround regions by balancing the accumulation and release of the seismic moments. We first calculated the slip rates on the Longmenshan and other major faults in the region using a three-dimensional regional-scale block model, constrained by the updated GPS data. The predicted slip rates on the major faults are comparable to previous geodetic results. On the Longmenshan fault, the right-lateral and dip slips are respectively 3.2±1.3 mm/yr and 0.5±2.0 mm/yr along the southwestern segments, 0.3±1.1 mm/yr and 2.6±1.3mm/yr on the northeastern segments. In contrast, slips rates on the Xianshuihe and other faults in the region are much higher (up to 17 mm/a). The predicted rate of total scalar moment accumulation in the Ganze-Songpan region is about 9.85±1.30×1018 Nm/yr. Using the earthquake catalog, we estimated that the total scalar moment released in the Ganze-Songpan region is 3.56-6.77×1020 Nm from 1879 to 2007, approximately half of the regional scalar moment accumulated during this period. The moment deficits were located mainly in the western parts of the Xiangshuihe and Kunlun faults. The eastern segment of the Xianshuihe fault has been a focus of studies because of the high slip rates and frequent earthquakes, but the recent quakes here have reduced the moment deficit to 5.19×1019 Nm, barely enough for a Mw 7.0 event. The moment deficit on the Longmenshan fault during this period is only 8.85×1019 Nm, insufficient for the Wenchuan earthquake. Hence we conclude that no large earthquakes similar to the Wenchuan event occurred on the Longmenshan fault in the past 1300 years before May 12th, 2008.
U21C-08
The 2008 M7.9 Wenchuan earthquake - Result of Local and Abnormal Mass Imbalances?
The May 12, 2008 M7.9 Wenchuan earthquake occurred along the Longmen Shan margin of the eastern
Tibetan plateau in the Sichuan province of the People's Republic of China. A complex and NNW dipping
reverse fault system including the Beichuan fault ruptured 250-300 km parallel to the Longmen Shan thrust
belt. This region has been tectonically loaded for >10kyr. It has low deformation rates of less than
1.0±1.0 mm yr-1 resulting in no major seismic activity during the Quaternary period. Several
geophysical observations suggest that this M7.9 earthquake was triggered by local and abnormal mass
imbalances on the surface of the Earth's crust. These observations include (1) elastostatic response of the
crust to the mass changes (2) slip distribution of the main rupture, and (3) aftershock distribution. Initially,
approximately 2 years prior the nucleation of the mainshock, at least 320 million tonnes of water accumulated
within the upper Min river valley. It enters the Chengdu plain of the Sichuan basin, a stable continental
region (SCR). The water volume amplified the strain energy on the Earth's crust. Shear stresses increased
by >1kPa on the Beichuan fault at the nucleation point in about 20km depth. Normal stresses decreased
by <-4kPa and weakened the fault strength. Pore pressure increases might have additionally destabilized
the fault locally due to pore pressure diffusion. This effect, however, might be minor in 20km depth, because
of low lateral fracture connectivity and permeability between the area of water accumulation and the
Beichuan fault. Overall, the stress alterations within a 120±70km2 large area resulted in the
Beichuan fault coming closer to failure. Such an area ruptured would account for a M7.2±0.1
earthquake assuming only 10 MPa stress drop. Secondly, a reverse fault focal mechanism dominated, in
particular, during the first 50 seconds of the main M7.9 rupture. The Beichuan fault slipped up to 7m upward
peaking at shallow depth (<7km) (Nishimura and Yagi 2008). A third reason is that the aftershock
distribution (M>3) along the ruptured fault zone pointed toward a very shallow seismicity in the uppermost
part of crust. Data show that about 87% of the total seismic moment has been released in the upper third
part of the crust (<10km), 1% in the middle third (10-20km), and 12% in the lower third (>20km).
Such a bimodal depth distribution with an aseismic mid crust is typical for earthquakes in stable continental
regions (Klose and Seeber 2007, SRL 78:554-562). Thus, high spatiotemporal correlatives can be observed
at shallow depth between both the seismic moment release of the aftershocks and the rupture slip distribution
of the mainshock. This indicates that Mohr-Coulomb failure stress states were much higher in the uppermost
part of crust than near a possible nucleation point of the mainshock in 19km depth. In conclusion, the
ensemble of geophysical observations suggests that the root cause of triggering the M7.9 Wenchuan
earthquake may have stemmed from local and rapid mass changes on the surface.
http://www.geol.tsukuba.ac.jp/~nisimura/20080512/