S41C-01
3-D Structure of the Moho Interface beneath South Korea from Regional Seismic Observations
The current project is concerned with the collection and processing of seismic waveform data to perform 3-D tomographic inversions and produce high-resolution 3-D crustal P- and S-wave velocity models for the South Korean peninsula. At present, we have analyzed and archived Korean Meteorological Administration (KMA) waveform data from 2001 through 2006 and mapped of the Moho discontinuity below South Korea. Phase arrival information from both, velocity and accelerometer sensors were collected. The analysis included 226 events throughout the region producing a total of 6,275 phase picks including Pg, Pn, Sg/Lg, and Sn phases. A total of 3,550 P-wave and 2,725 S-wave phases were identified. Using the combination of all available velocity and accelerometer data it was possible to estimate depth locations for 198 KMA events. The hypocenters were subsequently used to derive travel-time distance curves to appraise the quality of the travel-time picks. Static corrections were calculated for each seismic station within the KMA network to remove the effects of local inhomogeneities in the vicinity of each station. After applying static corrections to the observed travel-times, refracted P-wave phases along the Moho boundary were selected from the dataset to estimate the depth and topography of the Moho discontinuity beneath South Korea. In total, 526 Pn phases were collected from the KMA data with hypocentral distances from 130 km to over 650 km. The resulting Moho topography reveals a slightly undulating interface with a large-scale dip from north (31 km) to south (38 km) and a depth range from 32 km in the east to 39 km in the south-west. On a smaller scale, a more pronounced depression is evident in the south- central part of the mapped area, which opens to the south. The presented results are corroborated by other studies which mapped the Moho interface using surface-wave dispersion and receiver-function analysis. The present study is complementary to these earlier works, in that a different method and a different dataset generated comparable results.
S41C-02
Lateral Variations in Attenuation Beneath Japan
Improved images of the lateral variations in elastic properties of the mantle in terms of isotropic and anisotropic seismic wave speeds are generated almost routinely by various research groups. These models provide key information on our understanding of the mantle composition and dynamics. Another piece of information that would be very useful in advancing our knowledge of the mantle is constraints on the heterogeneity of anelasticity or attenuation. Lateral variations in attenuation are particularly of interest in subduction zones where complex chemical and dynamical signals, such as presence of water/volatiles, melting, and wedge flow, are directly related to attenuation of seismic waves. Japan is one of the most suitable subduction zones for modelling of attenuation structure. There are numerous local earthquakes with depths ranging from the surface to nearly 600-km depth, and there are dense seismic networks to record signals generated by these events. In particular, the bore-hole instrument network, the High Sensitivity Seismograph Network (Hi-net), provides high- quality data at about 20-km spacing. The first arrivals recorded by the Hi-net array from local earthquakes show strong location dependence in their frequency content. We use relative frequency content and arrival times to model variations in attenuation in the upper mantle beneath Japan. Combined analysis of elastic and anelastic structure of the mantle may shed light on our understanding of subduction processes.
S41C-03
Attenuation in the Upper Mantle Beneath the Northern Apennines (Italy) from Teleseismic P- and S-Wave Spectra
We present preliminary results for seismic attenuation in the mantle beneath the Italian region. We estimate P- and S-wave spectral ratios from teleseisms recorded at the temporary broadband seismic network deployed during the RETREAT (Retreating-TRrench, Extension, and Accretion Tectonics) project. We examine body-wave attenuation variation across the northern part of the Apennines mountain belt, which represents the accretionary wedge exposed during recent episodes of the subduction process in Italy. The data recorded during the three-year seismic campaign were analyzed using an ad hoc semi-automated procedure based on the cross-correlation analysis of a single phase across all the stations for each event. The seismic phases analyzed (P, S, SKS) display different patterns of seismic attenuation. Furthermore, we observe systematic variations in the distribution of the attenuation values as function of both the azimuth and the incidence angle of the seismic rays. Relatively high attenuation values are found on the Tyrrhenian side by seismic rays coming from the SW for both P- and S-phases. For NE-approaching rays the pattern of high attenuation values varies considerably, depending on the seismic phases: for P-waves it grossly corresponds to the mountain belt, while for S-waves it extends over almost the whole study area. By correlating attenuation estimates and the velocity structure from the existing tomographic models, we can make some inferences on the thermal state of the sublithospheric mantle, and on the physical properties of the tectonic elements which constitute the subduction system in the region. From the analysis of the P-phases we can clearly distinguish three main areas with different attenuation values, corresponding to the back-arc mantle (high attenuation), to the slab (low attenuation) and to the retro-slab mantle (high attenuation). The correspondence between the identified elements of the subduction system and the S- waves attenuation is not straightforward, and need to be further investigated.
S41C-04
Crustal Thickness Estimates for Tuscany and the Northern Apennines from Surface Wave Dispersion Curves
The tectonic processes that define the Northern Apennines and Tuscany are more complicated than a simple subduction model and thus subject to significant controversy. While tomographic images seem to show a descending slab below Northern Italy, a lack of earthquakes below 100 km is inconsistent with ordinary subduction. Some have proposed that subduction may have ceased, and the slab beginning to founder. Another view is that that subduction is not the primary process, but rather the lower lithosphere is detaching to cause the surface deformation that characterizes the Apennines. In the case of subduction, Tuscany should represent the extensional back arc and have thinned crust due to stretching. If delamination has occurred the crust could also be thinned, but by the removal of the lower crust. However, if the crust beneath Tuscany is thick, then some other process must also be at work. Previous active-source and receiver-function studies have found that the crust under Tuscany to be thin (about 20-25 km) while the crust near the crest of the Apennines is believed to be 30-40 km thick. We will report estimates of crustal thickness based on dispersion curves determined for a variety of two-station pairs using wavelet correlation. The stations used were part of the RETREAT deployment, a broadband seismological experiment supported by INGV (Italy), NSF (USA) and the Czech Academy of Sciences. The measured dispersion curves were compared with theoretical dispersion curves for a variety of crustal thicknesses to find a best fit thickness that could serve as a constraint for future crustal studies. In preliminary analyses, dispersion curves measured between stations in the Apennines orogen suggest a thick crust in that region. The dispersion curves for station pairs within Tuscany are less conclusive, but include examples where thin crust can be argued.
S41C-05
New evidence for and against a retreating subduction zone model for Northern Apennines.
We investigate crustal structure of the northern Apennines using broadband seismic data from a dense linear transect. Receiver functions reveal significant differences between the responses of the Tyrrhenian and Adriatic sides of the Apennines, and determine the spatial location of a change in this response, just west of the high crest of the mountains. We find a clear signature of the crust-mantle transition on the eastern Tyrrhenian side, and no obvious signature on the western Adriatic side. We also use relative timing of likely direct and multiply-reflected P-S converted waves in the crust to develop estimates of crustal thickness and Vp/Vs ratio within it. A nearly-uniform thickness of about 25 km is seen between the island of Elba and the crest of the Apennines, while further east crustal thickness increases to 35+ km over 10-30 km lateral distance. Crustal thickness is much less certain east of the Apennines due to complex structure or the interference of near-surface reverberations. Within the Po plain P-SV response of the crust is dominated by the resonance in the sediments, which limits our ability to determine crustal parameters. The abrupt change in crustal properties coincides with a major change in SKS birefringence, with areas south-west and northeast of the Apennines' crest showing distinct patterns. Shear-wave birefringence integrates seismic anisotropy along the wave's path, and likely documents a change in the upper mantle texture, rather than the crust. The spatially coincident change in the crustal structure and the upper mantle texture is expected for the subduction zone retreat model. However, the nature of the upper mantle deformation suggested by anisotropy observations presents a challenge for this interpretation, as we infer flow (extension) along the Apennines beneath the thin crust of the Tyrrhenian coast, i.e. in the back-arc area, not beneath the retreating slab. We speculate that crustal structure in northern Apennines is a relict of the subduction zone retreat that has (nearly or completely) ceased here, while upper mantle texture reflects an ongoing 3-dimensional deformation of the region. http://earth.geology.yale.edu/RETREAT/
S41C-06
Teleseismic imaging of Northern Cascadia upper mantle in the Explorer region
The transition from subduction along the northern Cascadia margin to transcurrent motion between the Pacific and North America plates occurs in the vicinity of northern Vancouver Island. While there is clear evidence for active subduction of the Juan de Fuca plate beneath central and southern Vancouver Island, the situation is more complicated beneath the northern part of the island, where the small Explorer plate is not actively subducting, but rather being over-ridden by North America. It is not clear whether megathrust earthquakes in this area will involve only the Juan de Fuca plate (with a northern limit near Nootka Island), or if they will also include the Explorer plate region (to the vicinity of the Brooks Peninsula). Addressing this question has important implications for seismic hazard on Vancouver Island. To investigate the velocity structure of the upper mantle in the Explorer region, a cross-shaped array composed of 24 portable broad-band seismic stations was deployed on northern Vancouver Island as part of the POLARIS-BC experiment (NVI array). One arm of the array trends NW-SE in the direction parallel to strike and straddles the assumed northern end of the subduction zone, and the second arm trends SW-NE in the direction perpendicular to strike, just north of the extension of the Nootka fault beneath Vancouver Island where convergence is observed. Here we present an analysis of teleseismic data recorded between June 2005 and June 2007 from the entire NVI array. We show preliminary images obtained using receiver function analysis and P-wave traveltime tomography.
S41C-07
Shear-wave splitting and mantle dynamics beneath the Rivera and Cocos subduction zone
Shear-wave splitting measurements are determined using data collected from MARS (Mapping the Rivera subduction zone) project to study the origin of seismic anisotropy in the mantle beneath the wedge of the subduction zone. The MARS project consists of the deployment of 50 broadband temporary stations in Mexico covering the Jalisco block from the coast to the Tepic-Chapala rift in the north and about 150 km to the west of the Colima rift. Results show that the fast directions are oriented in a SSW-NNE direction beneath the western Jalisco block, while the fast directions beneath the eastern Jalisco block show a predominantly N-S oriented fast direction. If the divergence of the fast direction is a measure of the mantle flow within the wedge of the subduction zone, then the overriding continental margin is being pulled apart. The Colima rift and active volcano are situated above the tear. Fast splitting directions in the vicinity of the volcano are generally trending N-S with a small delay time between the fast and slow split SKS waves.
S41C-08
Upper mantle of the Bohemian Massif (Central Europe) studied by surface waves from Kurile Islands M8.1 and M8.3 earthquakes
Two strong earthquakes of M8.3 and M8.1 occurred near Kurile Islands in December 2006 and January 2007. Broadband surface wave records were obtained at many stations in Central Europe in the epicentral distance of about 8500 km. We have analyzed broadband seismograms from the Czech Regional Seismic Network together with twelve temporary stations of the Institute of Rock Structure and Mechanics. We present a surface wave phase velocity analysis of several interstation paths of length around 150 km crossing the Bohemian Massif. This geological unit located in Central Europe is of Variscan age and its upper mantle seismic structure is still under investigation. Both Love and Rayleigh waves in the period range from 20 to 180 s have been studied. We provide polarization analysis to show the true backazimuths of surface wave propagation paths. We use also an array approach of our 7-stations broadband array placed near the center of the Bohemian Massif. We present Love- Rayleigh wave discrepancy as well as vertical-radial Rayleigh wave discrepancy pronounced in different period bands. The most striking feature is the low Love wave phase velocity around the period of 130 s. 1-D inversion of the phase velocity dispersion curves has been computed. Phase velocity dispersion curve anomalies are interpreted as a consequence of low velocity channel in the middle crust and upper mantle anisotropy. http://www.irsm.cas.cz