S33B-1942
Evaluation of Flexible Array Station Performance and Ambient Noise Analysis Using 500 Days of Continuous Recordings
Within the NSF funded EarthScope USArray program, the Flexible Array (FA) is a pool of campaign seismic instruments for Principal Investigator-driven studies to augment the Transportable Array footprint in imaging key geophysical targets at higher resolution. In this study we evaluate the performance of FA stations using data recorded from the EarthScope CAFÉ experiment in western Washington. Using this unique data set, we create a reference point on how well portable broadband stations perform for an extended continuous period of over 500 days (150:2006 - 50:2008) . All instrumentation that comprise the CAFE experiment is essentially new, of the same type and deployed using a uniform installation technique. The performance of 60 stations is analyzed; 46 stations are broadband, equipped with Guralp CMG 3T and Reftek R130's, the remainder equipped with short period Guralp CMG 40T1Hz and the same data acquisition system. The information used for this evaluation is derived from three sources; detailed field service notes kindly provided by the PI's (Ken Creager, Stephane Rondenay, Geoff Abers), data reports from the IRIS Data Management Center, and the actual time series data. The data return (based on data archived at the DMC w/o any problems) for this experiment is calculated to be 94.5% . The various failures through time are segregated into logical categories where trends in deployment techniques and equipment failures are quantified. Using McNamara statistical analysis to characterize background seismic noise, probability density functions were computed for 25 CAFE stations spanning over 500 days of recording beginning in mid 2006. Results from each station were then combined to produce a network wide characterization of the background noise level. For the same time period, PSD's for 35 nearby Transportable Array stations were also computed and combined into a single system wide PSD. Both installation types perform remarkable well with some differences being observed at given frequency ranges and components. Comparison between both data sets reveal that the deeper TA vaults result in lower ambient noise levels of approximately 15 dB in the horizontal component below periods of 15 seconds. However for frequencies between 1 Hz to 10 Hz, the FA stations are approximately 10 dB quieter than the TA. The equipment used for the TA and FA have similar instrument responses. The only major difference between both types of stations in this comparison is the station design.
S33B-1943
Analysis of Station Quality Issues from EarthScope's Transportable Array
160 of the first 400 Earthscope USARRY transportable array (TA) stations have completed their first two-year deployment and are being moved to their next locations. Over the past 4 years the majority of stations have run with few interruptions in the transfer of real time data to the Array Network Facility (ANF) at the Univ of CA San Diego and near real time data to the IRIS Data Management System (DMS). The combination of telemetered data and dedicated people reviewing the waveforms and state of health data have revealed several conditions that can affect the data quality or cause loss of data. The data problems fall into three broad categories; station power, equipment malfunction, and communication failures. Station power issues have been implicated in several types of noise seen in the seismic data (as well as causing station failures and resultant data gaps). The most common type of equipment problem that has been found to degrade data quality is caused by sensor problems, and has affected all 3 types of sensors used in the TA to varying degrees. While communication problems can cause real time data loss, they do not cause a degradation of the quality of the data, and any gaps in the real time data due solely to communications problems are filled in later with the continuous data recorded to disk at each TA station. Over the past 4 years the TA team has recognized a number of noise sources and have made several design changes to minimize the effects on data quality. Design/procedural changes include: stopping water incursion into the stations, power conditioning, changing mass re-center voltage thresholds. Figures that demonstrate examples are provided. Changes have created better data quality and improved the station performance. Vigilance and deployment of service teams to reestablish communications, replace noisy sensors, and troubleshoot problems is also key to maintaining the high-quality TA network.
S33B-1944
Infrasound Observations at USArray Sites – A Preliminary Investigation of the Utility of Seismo-Acoustic Observations on a Regional Scale
An exploratory investigation is underway to quantify acoustic to seismic coupling and infrasound signals at four USArray Transportable Array sites in Colorado. Three new data channels were added to each of four US Array sites (TA-025A, TA-P25A, TA-N25A, and TA-N24A) and consist of a Paroscientific microbarometer (10 s/sample) in the seismometer vault, a Validyne DP250 or DP350 acoustic gauge (0.025 s/sample) in the vault, and a Chaparral 2.5 acoustic gauge (0.025 s/sample) connected to ten, twenty-five foot porous hoses at the surface for noise reduction. The three gauges record infrasound signals in overlapping bands from periods of days to 1/10 second with resolutions down to expected infrasound noise levels and record pressure levels both in the vault and on the surface. In addition to these four installations, a six-element infrasound array was deployed in the same region to assess infrasound phase velocities and back azimuths for correlation with the single station observations. The data from these sites is being telemetered and will be stored at the IRIS DMC. While there have been previous studies of seismo-acoustic arrays, little is known about the difference between infrasound observations within a shallow underground vault and those recorded in the open at surface. Data from these new installations are being used to address two initial questions: (1) What is the utility of acoustic and barometric measurements for identifying and removing infrasound noise from long-period seismic data and (2) Can adiabatic generated infrasound vortices in severe storm cells be recognized and tracked on infrasound arrays and stations. Preliminary analysis has quantified correlated long-period barometric signals between neighboring USArray sites with periods of hours. Higher frequency signals are not correlated as might be expected for signals dominated by either local noise or regional signals.
S33B-1945
Observations of Infrasound-to-Seismic Coupling at Earthscope Stations Using Co Located Infrasound Microphones
In August 2007, infrasound microphones were added at Earthscope stations P13A, M13A, M14A, and N12A as part of an experiment to record 40,000 lb explosions at the Utah Test and Training Range on Hill AFB. These multiple observations provide insight into the infrasound-to-seismic transfer functions during this time. Additional observations on infrasound-to-seismic coupling are provided by thirteen seismic/infrasound stations in the University of Utah Regional Seismic Network. Strong infrasound-to-seismic coupling is observed at most stations and in most cases can be described with a few term pole-zero empirical transfer function. Using this transfer function, the direct infrasound contribution may be removed from the recorded seismogram leaving a small residual. The characteristic infrasound-to-seismic signals suggest simple filters that may be used to enhance such coupled signals on seismograms. Details of the transfer function may give insight into the relationship of infrasound and wind generated noise to recorded seismic noise at each site. At some stations, there are additional infrasound-to-seismic converted signals where the conversion occurs at significant distances from the site. These converted signals seem to be site dependent and may also be source dependent. Combining the arrival time differences between the more distant converted signals and the direct acoustic signals with polarization directions determined from particle motion analysis provides clues to where the coupling occurs. Understanding of where the coupling occurs is useful for understanding infrasound propagation and should provide information about the very shallow velocity structure in the region.
S33B-1946
Using USArray to Determine Lithospheric Thickness in the Central Andes From Teleseismic Depth-Phase Precursors
The Central Andean margin is a natural laboratory for understanding orogenic processes associated with plateau uplift and subduction of oceanic lithosphere beneath continental margins. Here we analyze seismograms of intermediate to deep Andean earthquakes recorded at USArray stations in a search for underside reflections from lithospheric discontinuities. The large number of USArray stations allows the use of cross-correlation and slant-stacking techniques to identify arrivals that are likely associated with the Moho, the lithosphere-asthenosphere boundary (LAB) and other deep interfaces in the central Andes. Our results suggest the MOHO beneath Puna-Altiplano varies in thickness from 50 to 80 km, consistent with estimates from other techniques. We also observe, for the first time, precursors that seem to represent reflections from the underside of the lithospheric-asthenosphere boundary. Importantly, this technique utilizes higher frequencies than other LAB-imaging methods such as S-wave receiver functions. Through using these higher frequencies we imply that this transition may be much sharper than previously documented. Earthquakes in the northeastern Puna (23-24S; 67-66.5W) indicate LAB precursors at depths of 135-150 km whereas further northeast under the Cordillera Oriental (22S-20S; 66-65W), the LAB deepens to 175-190 km. A relatively abrupt transition occurs between these two locales. At the southern end of the Subandean fold-thrust belt near 22.5S at 66W the LAB deepens markedly from 150 to 175km. To the east in the foreland, a north-south trending profile at 65W shows a deep, stable lithospheric thickness of ~200 km from 28S to 24S. Further south above the Pampean flatslab between 69W and 67W at 31S, we also observe the LAB. Southeast of Pie de Palo (68W) lithospheric thickness is measured to be 80 km. The LAB is not seen beneath Pie de Palo and reappears to the east in the Sierras Pampeanas at 67W at a depth of 100 km.
S33B-1947
Modeling of the Byerly's False S Phase for the Earthquakes off the Coast of Northern California
We modeled the observations of the Byerly's False S phase for the earthquakes off the coast of northern California to identify the origin of the seismic phase that was first observed 70 years ago (Byerly, 1937). One of the reasons that False S has not yet been adequately explained is that the seismic station coverage in the region was rather sparse. The recently installed USArray stations that complemented the existing Northern California Seismic Network (NCSN) and Berkeley Digital Seismic Network (BDSN) have provided ample observations of the False S phase in just a couple of years. We identified offshore events that produced the False S phase, relocated them using a double-difference algorithm, and then inverted for the seismic moment tensor. With the location and source parameters constrained, we modeled the broadband waveforms using 2D and 3D velocity structures in order to find the origin of the False S arrivals. Our preliminary results are showing that the geometry and velocity structure of the subducting Gorda plate play an important role in the generation of strong phases with travel times between direct P and S. We modeled the False S observations by developing a network of 2D structures to explain False S on all the stations leading to 3D modeling. We also studied possible source mechanism dependence of False S. The modeling results provide constraints on the geometry and velocity structure of the subducting Gorda plate to the north of the Mendocino Triple Junction (MTJ) as well as on the structure of the overlying Franciscan crust. Future modeling of the observations to the south of the MTJ could provide constraints on the depth, shape, and velocity structure of Gorda plate remnants.
S33B-1948
Quasi-Love Surface Wave Observations on USArray: Evidence for Upper Mantle Anisotropy Along the North American Plate Boundary
Strong evidence for mantle anisotropy exists along the North American and Pacific plate boundary, likely a result of deformation of upper-mantle rocks from the applied stress of tectonic motion. We have observed anisotropy-sensitive Love-to-Rayleigh scattered waves, also known as Quasi-Love waves, on the USArray component of EarthScope, on propagation paths that cross the North American plate boundary. These observed Quasi-Love waves may indicate the location and orientation of this anisotropy and help understand the tectonic motion along the North American plate boundary on the Pacific coast of the US. We have performed quantitative and qualitative tests to support the interpretation of these Quasi-Love waves against the misnomer of a surface-reflected body wave or Rayleigh overtone. All of our Quasi-Love wave observations are made after applying a 100 second low-pass to the data in effort to mute higher frequency Rayleigh wave overtones. The Quasi-Love waveforms that we have observed are the largest deviation between the USArray recorded wave trains and long-period synthetic seismograms computed using an isotropic Earth model. The observed amplitudes of the Quasi-Love waves fall roughly within 5-10% of the incident Love wave amplitudes, which is consistent with a 3-5% lateral anisotropic gradient along the great circle path [Yu et al., 1995]. Two example observations that we have made that support the proposed hypothesis are from a 3/31/2006 earthquake source near the Kermadec Islands region, where the observed Quasi-Love wave arrives from the west closely after the arrival of the Love wave, suggesting that the anisotropic scattering occurs relatively close to USArray stations in the Pacific Northwest. By contrast, no Quasi-Love waves are observed from a 4/5/2007 earthquake near the Azores Islands, which arrives at the same stations from the east. A particular point of interest is how the anisotropic scattering mechanism changes, if at all, between the Juan de Fuca subduction zone along the coast of Washington and Oregon and the transform plate boundary in California. Though the exact cause of the anisotropy, whether LPO, parallel fracturing, or a combination of both, may not be known, the orientation of the anisotropy will illuminate the stress and strain in the upper mantle and aid in the investigation of tectonic motion along the boundary.
S33B-1949
Upper Mantle Shear Wave Structure Beneath North America From Multi-mode Surface Wave Tomography
The upper mantle structure beneath the North American continent has been investigated from measurements of multi-mode phase speeds of Love and Rayleigh waves. To estimate fundamental-mode and higher-mode phase speeds of surface waves from a single seismogram at regional distances, we have employed a method of nonlinear waveform fitting based on a direct model-parameter search using the neighbourhood algorithm (Yoshizawa & Kennett, 2002). The method of the waveform analysis has been fully automated by employing empirical quantitative measures for evaluating the accuracy/reliability of estimated multi-mode phase dispersion curves, and thus it is helpful in processing the dramatically increasing numbers of seismic data from the latest regional networks such as USArray. As a first step toward modeling the regional anisotropic shear-wave velocity structure of the North American upper mantle with extended vertical resolution, we have applied the method to long-period three-component records of seismic stations in North America, which mostly comprise the GSN and US regional networks as well as the permanent and transportable USArray stations distributed by the IRIS DMC. Preliminary multi-mode phase-speed models show large-scale patterns of isotropic heterogeneity, such as a strong velocity contrast between the western and central/eastern United States, which are consistent with the recent global and regional models (e.g., Marone, et al. 2007; Nettles & Dziewonski, 2008). We will also discuss radial anisotropy of shear wave speed beneath North America from multi-mode dispersion measurements of Love and Rayleigh waves.
S33B-1950
Results from Ambient Noise Tomography in the Western USA using the USArray Transportable Array
We present current results from ambient noise tomography (ANT) applied to USArray Transportable Array
(TA) data in the western United States (USA). We have processed TA and regional network data since
October 2004 to produce a cumulative data set that now includes over 150,000 inter-station paths between
more than 500 stations within the continental US. The high spatial density of these stations provides short- to
intermediate-period (6 to 40 sec period) Rayleigh and Love wave dispersion maps with a resolution of about
the average inter-station distance (70 km) within the f
ootprint of the TA. As the TA rolls to the east, emerging data provides for new, high resolution group and
phase velocity estimates across large regions of the western and central USA that, heretofore, had poor
station coverage. The
ray path coverage and station density provided in this data set provide an unprecedented opportunity to
examine regional velocity structure, anisotropy and wave propagation which improve understanding of the
structure and dynamics of the crust and upper mantle beneath the western USA.
In addition to the updated Rayleigh and Love wave dispersion maps, we present results from three research
efforts based on ANT: (1) a 3-D shear-velocity model of the crust and uppermost mantle from a joint
inversion of ambient noise and earthquake data, (2) azimuthal and (3) radial anisotropy structure beneath
the western USA. The recent development and application of a new method of surface wave tomography that
models wavefront complexity and off great-circle propagation, which we call Eikonal tomography, allows for
the robust estimation of Rayleigh wave phase velocity azimuthal anisotropy. We invert for a 3-D shear
velocity model with azimuthal anisotropy in the middle/lower crust and the uppermost mantle. Overall, the fast
directions and the strengths of anisotropy in the uppermost mantle are similar to those derived from SKS
splitting, thus suggesting a common structural cause. The radial anisotropy model derives from the
simultaneous inversion of Rayleigh and Love wave ANT dispersion measurements. No simple isotropic 3-D
crustal and uppermost mantle shear-velocity model will fit the Rayleigh and Love wave dispersion
measurements together across large regions of the western USA. This Rayleigh-Love discrepancy cannot be
resolved by introducing radial anisotropy in the mantle alone, but can be resolved by adding positive radial
anisotropy (Vsh > Vsv) in the middle and lower crust in several regions. Large amplitude crustal radial
anisotropy is found in the predominant extensional regions of the western US, particularly in the Basin and
Range Province. The observed crustal radial anisotropy presumably results from the alignment of anisotropic
minerals during extension and crustal flow.
http://ciei.colorado.edu/ambient_noise
S33B-1951
Surface-Wave Phase-Velocity Maps of the Western US From USArray Data
Taking advantage of the exceptional quality and spatial sampling of the data from USArray, we present phase-velocity maps of the western US obtained by inversion of a large data set of inter-station surface-wave phase delays. Single-station phase anomalies are measured using the method of Ekström, Tromp and Larson (1997) for Love and Rayleigh waves in the period range 25-100 seconds. Two-station phase delays are then derived by differencing phase anomalies for those station pairs that lie within 2.5° azimuth of the great-circle path. The initial data set, consisting of up to 385,000 two-station phase delays at a given period, is reduced by assigning a single median measurement with an associated uncertainty to each unique station pair. The phase-velocity maps are parameterized in terms of 0.5° by 0.5° pixels, and a smoothness constraint is applied to regularize the inversions. Results in the 25-75 s period range for Rayleigh waves and 25-60 s period range for Love waves are robust, based on comparisons of multiple inversions derived using distinct subsets of the data, grouped on the basis of inter-station distance or azimuth. The phase-velocity maps correlate well with variations in crustal thickness, elevation, and known tectonic features, with the Yellowstone hotspot being notable for its prominence in the intermediate-period Rayleigh wave maps. We address the potential limitations of our analysis caused by the application of simple ray theory to interpret the phase delays, and investigate our ability to resolve azimuthal anisotropy from these data.
S33B-1952
Progress in Deriving Upper-Mantle Structure beneath Western U.S.
Recent upper-mantle triplication data recorded by USArray display sharp variations in both travel time and differential times between branches similar to that reported by Song and Helmberger (2006). Jumps between branches by up to 8 second are common in S-data. Generally, the recent tomography images of P- travel time data by the EarthScope community, Burdick et al (2008), predicts the horizontal geometry quite well, but does not predict the waveform triplications because of the choice of reference model. The present western U.S. reference triplication P-model, GCA, has the 410 discontinuity at 395 km, which is incompatible with recent S-models. Hence, we propose a modified P-model that can be used with S to construct a travel-time-delay map of paths sampling above the 410 and in the transition zone similar to that derived by Chu and Zhu (2008) for Tibet. Such data can then be added to conventional datasets using a hybrid tomographic method correcting for realistic path corrections in the upper mantle. Models with low- velocity zones and high-velocity zones should greatly help in producing accurate synthetic modeling. Deep events provide the best sources, such as the November 26, 2007 Guerrero, Mexico earthquake. We first inverted the focal mechanism and earthquake depth from teleseismic seismograms using a grid search algorithm. The best-fit solution shows a normal faulting at the depth of 52 km. This event provided a complete map of triplications from 10 to 30 degree. The AB and CD branches of P waves cross at the distance of 17 degree, which is 0.5 degree larger than that predicted by the GCA velocity model. The separation between CD and EF branches is about 0.5 second smaller at 20 degree compared with the GCA prediction, while the separation agrees with the GCA prediction at larger distances. Based on this observation, a 1D P-wave velocity model has been constructed for western U.S.. In this model, the 410 velocity discontinuity is at 420 km, which is 25 km deeper than the GCA. The 660 velocity discontinuity is at 648 km, 12 km shallower than the GCA model. Velocity jump at the 660 discontinuity is 4.20%, compared to 5.78% for the GCA model.
S33B-1953
Imaging Lithospheric Discontinuities Beneath North America with Teleseismic Ps and Sp Receiver Functions
Locating and characterizing upper mantle discontinuities such as the lithosphere-asthenosphere boundary (LAB) can provide important constraints on tectonic and geologic processes, for example the stability and evolution of continental lithosphere and the composition and rheology of the asthenosphere. Teleseismic receiver functions offer an effective means of estimating the depth and velocity contrast of these discontinuities. We have employed simultaneous frequency-domain deconvolution to obtain both Ps and Sp receiver functions at nearly 100 permanent broadband seismic stations across North America. At this stage, the scattered phases are migrated to depth using one-dimensional reference Earth models, but we also plan to migrate the data using a three-dimensional velocity model for North America obtained by inverting long- period surface waveforms in order to account for the effects of lateral heterogeneity in mantle structure. Preliminary results include strong Moho phases at many stations that are consistent between Ps and Sp. Negative Ps phases are observed at some stations in the tectonically active western U.S. and along the eastern continental margin at depths (70-110 km) that lie within the broad lithosphere-asthenosphere velocity gradient imaged by long period surface waves. These arrivals could represent scattering from a relatively sharp LAB. Imaging at LAB depths beneath the thicker cratonic lithosphere is more challenging due to interference from crustal reverberations in Ps data and lower signal quality in Sp. However, the data from a few stations hint at direct arrivals that could originate from the LAB. Our eventual goal is to integrate lithospheric discontinuities revealed through receiver function migration with surface wave inversions for three-dimensional upper mantle structure.
S33B-1954
A Three – Dimensional Receiver Function Study of the Western United States
The western United States has a complex geologic history and has been the focus of many regional scale PASSCAL seismic studies that investigate depth variations to the Moho, the 410 km discontinuity, and the 660 km discontinuities. Analysis of depth variations to the Moho in relation to topography is important in understanding the isostatic compensation depth, the thermal state of the upper mantle and boundaries between tectonic provinces. Analysis of the 410 and 660 km discontinuities allow us to determine variations in mantle temperature at these depths and facilitates comparison with tectonic boundaries. This abstract summarizes results from stacking Pds phases throughout the western US using data from all available previous PASSCAL studies in the western U.S. together with data from the EarthScope Transportable array. These data sets enable us to produce an image over the entire western US from the Pacific coast to the Rocky mountain front. Common conversion point stacking of Pds phases was performed by back projecting the data through a 3-D seismic velocity model (surface wave tomography model NA04 by Van der Lee). The images produced show large variations in Moho topography with an average depth of 39.6 kilometer over the western US with ± 7.2 km standard deviation in depth. As would be expected the Moho appears to be deepest beneath the Colorado Plateau and central Montana and shallowest throughout the Basin and Raange. The Moho also appears very shallow beneath eastern Washington. There is a band oof thick crust along the Yellowstone hot spot track. The 410 km discontinuity appears to have a mean depth of 427 km with a standard deviation in depth of ± 10.2 km. At this time the images are still very noisy but in a regional sense the 410 appears deepest beneath the southern part of the image and shallower to the north. Depths to the 660 km discontinuity appear to average 675 km with standard deviation of ± 9.8 km. The 660 does not appear to have a north-south change in depth but appears deepest to the Eastern part of the image and shallower to the west. This relationship may indicate that the thermal state of the 410 is controlled by high temperatures to the south associated with the Basin and Range and cooler to the north were subduction is present. The 660 may be controlled by the transition from warm oceanic and transitional lithosphere to the west and cooler continental lithosphere to the east.
S33B-1955
Two is Better Than One: Combining Receiver Functions and Ambient Noise Dispersion Curves
In this study, we jointly invert two previously unpaired seismic data types, receiver functions and ambient noise dispersion, to determine a high-resolution estimate of the earth structure. Jointly inverting receiver functions and surface wave dispersion is a well-tested technique (e.g., Juliá et al. 2000) that provides significantly higher quality estimates of earth structure than inverting each individually. While receiver functions constrain shear velocity of contrasts (like the Moho), surface wave dispersion resolves the shear velocity at different depth ranges. To obtain high resolution images of the shallow earth structure, we replace dispersion curves obtained from traditional teleseismic observations with those from ambient seismic noise data. The cross-correlation (or coherency) of pre-whitened ambient seismic noise measured at two seismic stations gives a waveform that approximates the Green function. This approach satisfactorily addresses some of the limitations imposed by traditional earthquake-based dispersion curves in joint inversion; most notably intrinsic attenuation and scattering are less significant, so the data from higher frequencies is retained and can be used to create more detailed estimates of the crustal structure (Shapiro et al., 2005). Furthermore, ambient noise is advantageous because it does not require earthquakes to occur at particular times or locations, typically providing much better back-azimuthal coverage than earthquake-based techniques. We use all available data from the USArray between 2006 and 2007 for both receiver functions and ambient noise dispersion estimation. Roughly 6000 receiver functions were computed with a new, efficient, semi- automated software package that applies iterative deconvolution in an interactive environment for quality control. High-quality receiver functions and group dispersion curves are obtained and verified with signal-to- noise weight stacking and bootstrap resampling for error estimation.
S33B-1956
Shear wave anisotropy beneath the Sierra Nevada: Implications for lithospheric foundering and mantle flow
Recent work in the western US indicates that the garnet-rich Sierra Nevada batholith root has undergone foundering since the early Cenozoic. The Sierra Nevada EarthScope Project (SNEP), undertaken to gain a better understanding of the foundering process, consists of a network of ~80 broadband seismometers spaced at ~25 km from ~37.0N to 40.5N in the western US. Here we present an SKS shear wave study of seismic anisotropy for teleseismic arrivals recorded at SNEP and US Array Transportable Array stations. We find dt>1s and φ approximately in the E-NE direction over most of the batholith. Splitting of this magnitude cannot be accounted for solely in the crust, and our results to first order are well explained by existing plate motion and mantle flow hypotheses (e.g., Silver and Holt, 2002). However, we observe subtle variations in splitting parameters as a function of backazimuth primarily at stations situated on the western foothills of the Sierra Nevada. These complexities may be indicative of either a two-layer or dipping layer structure beneath the batholith that may be associated with ongoing lithospheric foundering beneath the Sierran range. In the southern part of the study area, we note a reduction in dt for arrivals that sample the high Vp Isabella anomaly - an upper mantle downwelling thought to be a result of recent foundering. In the Lake Tahoe region, short length scale variations in splitting parameters also indicate sensitivity to shallower structures, possibly associated with dike injection beneath the region. Our work complements ongoing SNEP studies that focus on receiver function analyses, dispersion analysis of teleseismic Rayleigh waves, seismic tomography that probe this region of incipient lithospheric foundering.
S33B-1957
Focal Mechanisms for Deep Crustal Earthquakes in the Central Foothills and Near Yosemite National Park in the Sierra Nevada, California
Past studies have observed seismicity occurring to depths near 40 km beneath the central Sierra Nevada in eastern California, but the cause of this unusual activity remains largely unknown. We use seismograms from a recent deployment of the Sierra Nevada EarthScope Project (SNEP) broadband array and interspersed USArray TA stations to study this deep crustal earthquake activity. From June of 2005 to May of 2006, we recorded 126 earthquakes in the central western flank of the Sierra Nevada that relocated in the depth range from 1.0 to 47.6 km. These earthquakes have small magnitudes (M < 3), occur at a rate of ~10 per month, and occasionally display repeating waveforms. The majority of the earthquakes fall into two distinct clusters. One cluster of earthquakes form a diffuse band under the low foothills north of Fresno and have focal depths mostly between 20 and 35 km. The second cluster underlies the higher western slope of the range in a more compact north-south band extending from the southern edge of Yosemite National Park to the San Joaquin River. These events have focal depths from near surface to 30 km, and are located above occasional deep, long-period (LP) events (Pitt, et al., SRL, 2002). We use P- and S-wave polarity picks and P/SH amplitude ratios to construct focal mechanisms for 23 of the larger, well-recorded earthquakes, 14 in the Foothills Cluster and 9 in the Yosemite Cluster. The focal mechanisms show dominantly near vertical and subhorizontal nodal planes, although several events do show clear normal or reverse mechanisms. Although there is some scatter, a majority of the mechanisms from the Foothills Cluster have S-to-SW steeply dipping T-axes. The majority of earthquakes in the Yosemite Cluster have P-axes moderately dipping to the SW and T-axes moderately dipping to the NE, similar to focal mechanisms of earthquakes associated with the recent magma intrusion event under Lake Tahoe (von Seggern, et al., BSSA, 2008). We suggest that the earthquakes in the Foothills Cluster are occurring in response to the downward pull of an attached piece of dense ultramafic batholith residue and the events in the Yosemite Cluster are related to post-delamination crustal magmatic processes.
S33B-1958
Seismic structure of the upper mantle beneath Southern California from teleseismic travel- time tomography
We use Born-theoretical finite-frequency teleseismic travel-time tomography to investigate the seismic structure of the upper mantle beneath Southern California. The presence of strong seismic heterogeneity is established by previous body-wave tomographic studies (Raikes, 1980; Humphreys and Clayton, 1990; Zhao et al., 1996; Kohler et al., 2003; Boyd et al., 2004), but much uncertainty remains regarding the 3-D extent and physical properties of identified structures. Some high amplitude seismic anomalies are suggested to be near vertical in orientation and extend to depths greater than 200 km. Resolution of such seismic structures with teleseismic data is limited by the depth distribution of crossing rays, and modeling of finite-frequency travel-time sensitivity and non-geometrical wave propagation. We use data from 1983-2008 from approximately 500 stations in the Southwestern U.S. to provide unprecedented raypath coverage beneath Southern California. Born-theoretical frequency-dependent travel-time sensitivity kernels (Dahlen et al., 2000) and iterative 3-D shortest-path ray tracing are used to address the challenges of finite-frequency wave propagation in a heterogeneous medium. The SCEC 3-D crustal velocity model, regional crustal tomography models, and regional receiver function studies are used to create crustal corrections for all the travel-times used in our analysis. We currently have results from inversions of P wave data and we hope to have an S wave model by meeting time.
S33B-1959
Interpretation of Anisotropic Structures Derived from SKS-SKKS Splitting and Surface Wave Data: A Case Study of Southern California
The two main data sources for determining seismic azimuthal anisotropy in Southern California, SKS splitting data and surface wave data, show inconsistent patterns. Splitting from surface waves have fast axes aligned with faults with, on average, about 0.2 secs splitting time, whereas SKS-SKKS splitting has fast directions WSW and splitting times over 1 second. The primary goal of this project is to understand the source of this discrepancy and to obtain a seismic structure that satisfies both sets of data. The key must be in the depth variations in anisotropy, as the two types of data have different depth sensitivities. We formulated a scheme to invert surface waves, and obtained S-wave velocity anisotropy maps. We have made new shear wave splitting measurements for 126 seismic stations with best SKS-SKKS data (50 earthquakes out of 190 earthquakes) for the period of 1990-2008. For the surface wave anisotropy model we computed predicted splitting times from the mantle-lithosphere. We find that predicted surface wave splitting times obtain their largest values in the mantle lithosphere, but are much less than SKS-SKKS splitting times. The surface wave fast axes directions are also different and are mostly parallel to the relative plate motion direction and major faults. The largest variations occur just south of the big bend where transpression has been greatest. We correct the SKS-SKKS seismograms for anisotropy effects in the mantle lithosphere using the results from the surface wave analysis. Then we invert the corrected data for SKS-SKKS splitting parameters. After correction, fast directions rotate anticlockwise on average about 4 degrees and delay times decrease by on average 0.1 sec. The overall SKS-SKKS pattern is hardly affected. Therefore we conclude anisotropic structure derived from surface waves clearly cannot explain SKS splitting data but is probably related to the finite strain from the tectonics. We suggest that the SKS waves are sensitive to the deeper parts of the upper mantle.
S33B-1960
Relating the Seismic Character of the Crust and Upper Mantle to Late-Cenozoic Extension in Southwestern N.A.
A recent tectonic reconstruction (McQuarrie and Wernicke, 2005) places detailed constraints on the magnitude and scope of late-Cenozoic extension throughout Southwestern North America. This project seeks to better understand the distribution of extension throughout the crust and upper mantle and elucidate the transition from the highly extended Basin and Range to the relatively unextended Colorado Plateau. To this end, we present teleseismic receiver functions generated from 31 broadband seismometers associated with EarthScope's BigFoot Array, TriNet, and PASSCAL stations deployed across Southern California and Arizona. We employ the common-conversion-point stacking method to analyze variations in lithospheric structure. Additionally, in regions with clear converted wave reverberations we analyze the trade-off between crustal thickness and bulk Vp/Vs to improve our view of how crustal thickness and Vp/Vs relate to different tectonic environments and degree of extension. Our preliminary estimates indicate crustal thicknesses of ~25-30 km in eastern California increasing to ~40- 45 km within the southern Colorado Plateau. The transition between thin to thick crust appears to occur over as little as 20 km. Crustal Vp/Vs varies considerably, with Vp/Vs greater than 1.8 near the Transverse Ranges and Colorado Plateau, and less than 1.8 in the southern Basin and Range. We also view a change in the nature of the Moho approaching the Colorado Plateau. Initial calculations indicate the amplitude of the converted wave from the Moho is twice as strong beneath the Mojave and Southern Basin and Range than the Colorado Plateau. Additionally, we observe laminated crust in the western Mojave Desert approaching the Transverse Ranges.
S33B-1961
It's Still Downhill From Tonopah to Las Vegas, but the Crust Doesn't Ride for Free
We investigate the crustal thickness in the central Basin and Range province of the western US. There is a gravity anomaly at 37 degrees N latitude at which the gravity increases ~100mgal from North to South over a distance of ~100 km. The majority of recent publications ascribe the gravity signal to a mantle influence based on observations of near constant crustal thickness in the area. However, Moho depth estimates are sparse in the area, and therefore higher gravity due to a thinner crust in the south is still a possible explanation to date. In order to determine Moho depths, we examined teleseismic receiver functions from broadband and short-period stations from 1993 to 2008 located within the region, including stations from the recent Earthscope Transportable Array deployment. We used a total of 11,751 high-quality receiver functions at 80 stations and picked arrival times of the Moho converted phase from backazimuthal and moveout stacks. Moho depths were determined from these arrival times using a fixed velocity model, as well as from forward modeling of moveout curves of the direct conversion as well as multiples. Our results confirm the presence of thinner crust south of 37N latitude. Assuming an average crustal velocity of 6.3 km/s and a Vp/Vs ratio of 1.732, we found an average crustal thickness between 33 and 34 km north of 37N, and roughly 27 km south of 37N. We also found an interesting pattern of thin crust trending NE from the southern tip of Nevada to approximately 38N, 245E. The findings indicate that a least part of the gravity signal is of crustal origin.
S33B-1962
Using Receiver Functions to Image the Montana Crust and Upper Mantle
We determined receiver functions (RFs) at six permanent Advanced National Seismic System (ANSS) stations to examine crust and upper mantle structure of the Wyoming craton (WC) and Medicine Hat block (MHB). The Deep Probe & SAREX projects (Henstock et al., 1998; Clowes et al., 2002; Gorman et al., 2002) used active source seismics to model a high velocity crustal layer (the so-called 7x layer) beneath the WC. This layer exhibits P wave velocities that are high for lower continental crust (~7+ km/s) and extends from 30-55 km below the surface. Interpretations of the active source data indicate that this layer may represent wide scale crustal underplating of the WC, implying post-Archean craton modification with implications for Laurentia assembly. We used 43 earthquakes from a wide azimuthal distribution recorded at the Montana ANSS stations; high signal-to-noise ratios of 25 of these RFs were acceptable for further analysis. Receiver functions constrain crustal velocity structure beneath a seismometer by using P-to-S wave conversions at sharp velocity contrast boundaries. Preliminary results for seismic stations DGMT, EGMT, and LAO, located to the east of the Deep Probe and SAREX seismic line on the Wyoming craton/Medicine Hat block show the influence of sedimentary cover and a strong Ps phase at approximately four seconds after P. At BOZ and MSO, located in the Rocky mountains, the sedimentary cover signal previously noted is absent, and instead we observe a sharp Ps phase at about four and a half seconds after P. RFs at station RLMT (on the WC) are highly anomalous, probably reflecting complex conversions from two differently oriented dipping layers. We will use the RFs to produce suites of acceptable structural models to test for the presence and lateral extent of the 7x layer and other structural features of the Rocky Mountains-craton transition.
S33B-1963
Detailed Regional Resource Mapping in Nevada Using Natural and Mining Seismic Sources
Our objective is to contribute to geothermal resource assessment in Nevada and the Great Basin by providing uniform, high-resolution, regional maps of seismic velocity in the crust and shallow mantle. We analyze earthquakes and mining explosions of local magnitude greater than 2.5 recorded from 2005 to 2008 by the Earthscope Transportable Array (TA) seismic network, by the Nevada Seismic Network, and by the Southern Great Basin Digital Seismic Network. We use two complementary tomographic methods: 1) P- and S-body-wave arrival times image the mid- to- deep crust and shallow mantle with a block size of 10 km or smaller; 2) surface waves image the shallow to deep crust, with depth sensitivity depending on wave frequency. To identify surface waves, we use improved Rayleigh-wave detection algorithms based on wavelet transforms. Surface-wave detector improvements include solving for Love-Rayleigh interference and the capability to identify and analyze higher-mode Rayleigh waves. Once a Rayleigh phase is identified, group-velocity dispersion curves are determined using a multiple-filter analysis technique. With the advent of the TA station deployment, group-velocity dispersion for the path between two stations in line with the source may be estimated from the dispersion between the two stations. This technique eliminates the need to correct for earthquake source phase. The results of our study will be integrated with previous seismic data compilations and interpreted in terms of parameters of interest to the geothermal community, such as crustal thickness and seismic velocities.
S33B-1964
Mapping Crustal Structures of the Nechako Basin Using Teleseismic Receiver Functions
This poster describes a passive-source seismic mapping project in the Nechako Basin of central British Columbia (BC), Canada, with the goal of assessing the hydrocarbon and mineral potential of the region. An explosion of the mountain pine beetle population in central BC over the last decade has devastated the lodgepole-pine forest industry on which many communities depend. Mineral or energy extraction may provide an alternative economic opportunity for the region. The Nechako Basin has been the focus of limited hydrocarbon exploration since the 1930s. Twelve exploratory wells were drilled; oil stains on drill chip samples and the evidence of gas in drill stem tests indicate the presence of a hydrocarbon system. The basin contains mainly Mesozoic sedimentary and volcanic rocks which are blanketed by Eocene and Neogene volcanic rocks. Because only limited seismic energy is able to penetrate this hard volcanic cover, seismic data collected in the 1980s are generally of poor quality. The present study utilizes recordings of distant earthquakes at an array of seven broadband seismic stations deployed over much of the basin in September, 2006. Receiver functions are calculated from the recorded teleseismic waveforms by deconvolving the radial components with the vertical components, and subsequently inverted for shear-wave velocity structure using the Neighborhood Algorithm. Preliminary results indicate that the crustal thickness of the basin is about 40 km with a low velocity zone at around 20-30 km depth and low-velocities near the surface, which may represent the existence of sediments. This study will complement independent active- source seismic studies planned for the region by providing site-specific images and constraints on the shear- wave structures. This study will also complement magnetotelluric measurements currently underway, providing critical new information on porosity, fractures, and fluids.