S34A-01
A high velocity layer at depth beneath Bezymyannii Volcano, a possible source of seismicity?
Volcanoes of Central Kamchatka (Klyuchevskoi, Bezymyannii, Tolbachik) are offset westward relative to the strike of the remainder of the Kamchatka volcanic arc. On the basis of local seismicity the depth of the subducting Pacific plate beneath these volcanoes is estimated to be around 150 km, while the crustal thickness estimates for Kamchatka fall into the 30-40 km range. We apply receiver function analysis to data collected by temporary broadband seismic stations installed in the vicinity of Bezymyannii Volcano, (Kamchatka, Russia) by the PIRE group in 2006 and 2007. We find evidence for a sharply defined impedance contrast in the 80-100 km depth range beneath the Klyuchevskoi group of volcanoes. The signature of this feature corresponds to an abrupt increase of seismic velocity with depth, however its depth range is not consistent with either the crust-mantle transition or the top of the subducting Pacific Plate. A review of data collected in central Kamchatka a decade ago by the Side Edge of Kamchatka Slab project suggests similar features at upper-mantle depths both south and west of the Bezymyannii. We note rare, but consistent earthquake events associated with the depth range of 75-100 km recorded in both global and local catalogs. Presence of these events within the supra-slab mantle wedge is intriguing, as it would signify a presence of brittle (cold?) material in the same depth range where we find an abrupt impedance contrast with receiver function analysis. We use data from newly available local earthquake catalogs of Kamchatka Branch of Geophysical Service to verify the accuracy of these oddly located earthquakes. We attempt to further constrain the position of these events, both relative to the main Benioff- Wadati zone seismicity and with respect to the velocity contrast we image.
S34A-02
The fingerprints of Changbaishan Volcano on Regional Seismic Waveforms
Located in the southeastern margin of Jilin Province, China, Changbaishan Volcano, a Cenozoic composite volcano, has been dormant for about 300 years. In this study we investigate the seismic structure beneath this volcano by comparing waveforms for rays that pass through it with rays that do not. Our seismic sources are composed mainly of chemical and nuclear explosions, including the 9 October 2006 North Korea underground nuclear test. All were recorded by 16 broadband stations installed by Tongji University along the southeastern boundary of Jilin Province. Here we focus our attention on seismograms from the 9 October 2006 North Korea event and a magnitude 2.3 local chemical explosion. Our preliminary results show that shear waves disappear abruptly when great-circle paths intersect the volcanic region. Our observations suggest that a magma body is present beneath the Changbaishan and that it is responsible for the severe attenuation of shear waves. We are currently expanding our dataset to pin down the location and geometry of the magma body. Our study highlights the dramatic effects a tectonic structure may exert upon the observed wavefield and hence the necessity to take them into account when carrying out source discrimination analysis.
S34A-03
Intrinsic and Scattering Attenuation the Mt. Fuji Region, Japan
Mount(Mt.) Fuji, typical strato-volcano located near Tokyo metropolitan area, have anticipated potential hazard in view of the geologic and historic activity. Based on the extensive data of Mt. Fuji, we measured the coda quality factor (QC-1) in the single scattering model hypothesis, and the intrinsic and scattering quality factor (Qi-1 and QS-1 ) by the Multiple Lapse Time Window Analysis (MLTW) method. We classified near Fuji, within the 5 km radius (R < 5) from the summit of Mt. Fuji, and far-Fuji area, outer region from the 20 km radius (R > 20) of the summit. This classification show the largest discrepancy of at a range of sampling volumes corresponding to overlapped sampling depth of about 80 km, and also show the reliably greatest difference of and at the hypocentral distance of 80 km. In particular, the large difference of at 2 - 4, and 4 - 8Hz indicates lithospheric heterogeneity beneath the Mt. Fuji with a characteristic heterogeneity scale length about 1 km. The results, with small error range due to abundant data, show that all values in near-Fuji area are predominant over those of far-Fuji area, and for both near and far-Fuji area is over at high frequencies. The and values of far-Fuji reflect the values of typical non-volcanic area. The values of the near-Fuji area is lower than those of the other volcanic area, while values of is not. The low for the volcanic region of near-Fuji may suggest the magmatic inactiveness compared to the Hawaii volcano.
S34A-04
Time Analysis of Hydrogeodeformative Processes in the Territory of Armenia
In the present work the analysis of data of hydrogeodynamic supervision of hydrogeodeformative chinks is given taking into account seismic mode. The characteristics of the locations of chinks are given, starting with block structure of earth crust of Armenia. Hydrogeodynamic mode was compared with areas of tectonic stress. The character of distribution of hydrogeodeformative processes in time and space is studied. The method of hydrogeodynamic monitoring of earth crust was used and is being used now to supervise the data of level of waters of hydrogeodynamic chinks of the region, hydrodeformative processes and seismic mode. Large-scale zones of the earth crust are allocated in the territory of Armenia, geodynamics of which conditions seismicity of the territory. The folded zone of Armenia is the largest zone which stretches from suburbs of Gjumry to north-west to Megry to south-east. Eight hydrogeodynamic chinks function within a zone. Four chinks are located in the Somheto-Karabakh zone. One chink is located in the Kafansky zone and another one in Priaraksinsky depression. High informativness have chinks, which are located from 0.2 to 3 km from a break [3]. It indicates that in fault structures the influence of such factors, as deformation of environment and tensosensitivity of systems on formation of hydrogeodinamic effects is higher, than in the block. According to the water level of chinks hydrogeodynamic effects were marked in form of bay rising before earthquake as well as variations of water level on other chinks of observed deformation field. Though effects preceding earthquake were also observed they were very weak as there was decrease of tectonic stresses of environment moving off from epicenter of earthquake source. From the above mentioned we may conclude that - The revealed periodic variations of water levels of chinks (¹ 11, 18. 27) may be identified with influence of tide-generating forces. - Low-amplitude periodic variations of underground water caused by sawtooth fluctuations of waters of chinks (¹ 11, 15) are connected with slow rising of water level and its sudden abatement indicating the presence of weak deformations - By changes of water levels in chinks of tectonic blocks it would be possible to use the trends of increase or decline of water table of chinks (¹ 2, 1, 27, 15) as an indicator of definition of activity of tectonic blocks. - Proceeding from seismic mode, the greatest quantity of earthquakes with M>3 has occurred in the north of Armenia and in eastern part of the Sevan lake, as well as in across-track direction (presumably along the Garni break). The spatio-temporal analysis has shown the presence of mechanism of development of hydrogeodeformative processes, both as a distribution of geodynamic tension as well as the character of their influence on underground waters level.
S34A-05
Long-Period Seismic Noise at the Farallon Islands: Island's Tilting due to Infragravity Waves as a Possible Source of Earth's Horizontal Hum
We present the analysis of long-period noise observed at the broadband seismic station FARB (Streckeisen STS-2) located on the South Farallon Island, 43 km off the coast of San Francisco, CA. Results from our previous work showed that long-period noise (20-500 sec) observed at the ocean-bottom broadband station MOBB located offshore Monterey Bay, CA is mainly due to seafloor deformation under the pressure forcing by long-period ocean surface gravity waves (infragravity waves). Similar type of long-period noise is observed on the vertical and both horizontal components of the Farallon Islands station FARB, but not on the nearby mainland stations. The long-period noise at FARB is best observed on stormy days when it extends all the way to 1000 sec and is even stronger than at MOBB. The long-period noise at FARB is stronger on the E-W than on the N-S component, suggesting that it results from the infragravity waves that propagate from the nearshore region into the deeper ocean. The energy in the infragravity wave band at FARB is, as previously observed at MOBB, modulated in phase with tides. The phase of the modulation observed at FARB agrees with the phase of the local tides, suggesting that infragravity waves observed at FARB are locally generated. Analysis of the infragravity signal following the arrival of a dispersed swell shows that the infragravity waves are generated in the nearshore region from the ocean waves with periods shorter than 23 sec. The comparison of the incoming swell dispersion and the frequency of the resulting infragravity waves further shows that nonlinear interaction between a pair of swell components with frequencies f1 and f2 results in infragravity wave with the difference (f2-f1) frequency. Strong horizontal noise in the infragravity wave band observed at FARB suggests that passing infragravity waves tilt the island. We suggest that the swaying of islands (and underwater mounds) driven by tilting due to infragravity waves can contribute to the recently detected horizontal hum of the Earth.
S34A-06
Seasonal Anisotropy of Short-Period Seismic Noise in South Asia
We calculated frequency-wavenumber spectra for 955 samples of seismic noise recorded at the short-period Chiang Mai seismic array (CMAR) from 1995 through 2004. At frequencies above about 1.4 Hz the noise is isotropic and diffuse, but at lower frequencies the noise at CMAR is strongly partitioned by apparent velocity into two categories: teleseismic P wave energy with apparent velocities higher than 25 km/s (ray parameters of 0.0-5.0 s/deg) and higher mode Rayleigh energy with apparent velocities near 4.0 km/s. The ring of slowness space in between, corresponding to P waves turning in the crust and upper mantle, is relatively quiet. The Rayleigh noise is further partitioned by direction, with the strongest signal arriving from the Bay of Bengal at backazimuths of 180-255. A secondary peak in the Rayleigh noise occurs in the direction of the South China Sea at backazimuths of 80-120. The Rayleigh noise is strongly seasonal with annual variations of 10-15 dB in power. The easterly noise has peaks in local winter and troughs in local summer, while the noise from the southwest has the opposite pattern. This behavior is well-matched by the seasonal anti-correlation in significant wave heights in the South China Sea and the Bay of Bengal, as determined from TOPEX/POSEIDON satellite tracks. While propagating Rayleigh waves are often observed in seismic noise, it is less common to observe teleseismic body waves. Nearly all reports of body wave noise document ray parameters of 8-12 s/deg, which correspond to P waves that turn in the upper mantle, therefore our observations of a consistent noise peak with ray parameters of 4.5 s/deg and smaller, equivalent to apparent velocities of 25 km/s and higher, may be unique in the geophysical literature. Like the Rayleigh noise, the P noise observed at CMAR is seasonal. It has an annual power variation of 5-10 dB, with peaks in local winter and troughs in local summer. The seasonality implies that the noise is unrelated to small unidentified earthquakes or other tectonic processes, and instead is created at least indirectly by ocean waves, similar to the Rayleigh noise. Under this assumption there are several geographical regions that could act as sources: the western Atlantic Ocean near the coast of northern Brazil may contribute PKP energy, the Pacific Ocean just north of New Guinea may contribute PcP energy, and central portions of the North Pacific may contribute P waves that turn in the lower mantle. However, none of these sources provides an ideal match to our slowness observations and so none are preferred over the others. In order to resolve the uncertainty of the source location of the P noise at CMAR, a finer comparison of seismic data and ocean wave data is required. In any case, and irrespective of the precise source mechanism for the high velocity noise, our observations point towards a new method of imaging Earth's deep interior. Just as Rayleigh microseismic noise has been used to image Earth's crust it may be possible to use microseismic P-noise to image Earth's lowermost mantle and core. This could be especially beneficial for regions of the deep Earth that are poorly sampled by present-day patterns of seismicity.
S34A-07
Anatomy of T Phases Recorded by An OBS at 5 km Water Depth: Effects of Local Bathymetry and SOFAR Channel Heterogeneity
We have deployed broadband ocean bottom seismometers (OBSs) offshore eastern Taiwan in 2006 and
2007. One OBS at 5 km water depth has routinely recorded abnormal seismic phase after P and S phases.
The attributes of the energy, including frequency band and arrival time, are consistent with that of a T wave.
T waves are phases contained at least some acoustic paths through water bodies. It is still not clear how the
T phase energy leaks out of the SOFAR channel before being recorded at depth. From regional physical
oceanographic data we found heterogeneous SOFAR channel near the OBS site. Thus, some of the energy
can be scattered out of the waveguide. We used an earthquake near the coast to calculate the traveltimes
through different solid-earth to acoustic conversion points along the regional 1000 meter bathymetry contour
line where the regional SOFAR channel axis is located. The beginning of the T wave usually travel with a
path that follows Fermat's principle, instead of the shortest path. The T wave ends when there are no
effective conversion points available in the regional bathymetry. The maximum amplitude of the T phase
usually has a path with the least solid-earth leg. In addition, T phase amplitude decreases when a typhoon
passed between the conversion points and the OBS, possibly due to the strong winds causing mixing of the
upper SOFAR channel, making the waveguide less effective.
http://www.earth.sinica.edu.tw/~chi
S34A-08
Comparsions of accuracy and stability tests among the methods determining hypocentral parameters: VELEST, HYPOSAT, hypoDD, and GA-MHYPO
The accuracy and stability tests of hypocentral parameters determined by VELEST, HYPOSAT, hypoDD, and GA-MHYPO are carried out in this study by applying synthetic data for error-free and for noisy data of a 0.1- second standard deviation. The synthetic data applying true velocity model and hypocentral parameters provide us to compare accuracy and stability among the methods. Computational results show that GA- MHYPO yields most accurate and stable hypocentral parameters among four methods. VELEST and HYPOSAT yield large focal depth and origin time errors generally, but yield relatively small epicenter errors. hypoDD improves accuracy of hypocenter greatly compared with accuracy of initial hypocenter when initial hypocenter has large errors, but does not improve or decrease accuracy of hypocentral parameters when initial hypocenter has small errors. The accuracy of origin time determined by hypoDD is similar to accuracy of initial origin time without depending on accuracy of initial location. We also show the differences of hypocentral parameters determined by these methods using earthquake data occurred in the vicinity of Kyeong-ju area in the Korean Peninsula. The stability test is carried out by applying bootstrapping technique for one event.