S31D-01 INVITED
Two-Stage Farallon Subduction History Under North America Inferred From Multi- Frequency P-Wave Tomography
We apply a new tomographic method, multi-frequency body-wave tomography, to teleseismic P-wave measurements in North America. The goal is to systematically exploit frequency-dependent information contained in broadband waveforms of body waves for imaging three-dimensional mantle structure. 420,000 traveltime and 160,000 amplitude measurements of teleseismic P-waves are jointly inverted for P-velocity and attenuation to ~1800 km depth. Resolution is crucially improved by the new USArray stations in the western U.S.; their dense spacing plays to the strengths of finite-frequency tomography. We discover several previously unknown or unrecognized fragments of the subducted Farallon plate in the lower mantle beneath the western U.S., as well as tears or breaks in the plate that are thousands of kilometers long. We also resolve large amounts of older plate material foundering in the transition zone beneath the eastern half of the continent (this material must be older since the trench has been moving westward continuously for ~150 million years). The Farallon fragments hence do not follow a simple, proportional relationship between depth and time since subduction, with interesting implications for the state of the mantle. We attribute the complex geometry to a frontal, trench-parallel break at the end of the Laramide era (~50 Myr), a time of unusual tectonic and volcanic surface activity credited to extremely flat subduction. We propose that the newly discovered tears solve the geodynamical requirement of mass exchange (mantle flow) that enabled the flattened slab to begin its descent while remaining largely sheet-like. Yellowstone sits atop the junction of two such tears; this would have allowed for an unimpeded ascent of the presumed mantle plume head that generated the Columbia River Basalts at ~12 Myr.
S31D-02 INVITED
Complex geological interactions in the mantle beneath western USA
Using finite-frequency sensitivity kernels we construct tomography images of mantle structure beneath the
western USA. Our Dynamic North America Models of P- and S-velocity structure (DNA08-P and DNA08-S)
use teleseismic body-waves recorded at over 800 seismic stations; half provided by the Earthscope
Transportable Array and the other half from regional networks. DNA08-P and -S benefit from the
unprecedented aperture of the network while maintaining a dense station distribution providing high-
resolution body-wave imaging of features through the transition zone and into the lower mantle. North of the
Mendocino Triple Junction (MTJ) the mantle is characterized by broad high velocity features, while south of
the MTJ the mantle is dominated by broad low velocity anomalies. There is a complex bottom edge to the
high velocity Juan de Fuca slab which extends to ~800 km depth beneath northern California and only
~400 km beneath Washington. The narrow low velocity conduit beneath Yellowstone National Park dips
to the northwest in the upper mantle and connects to a much broader low-velocity anomaly in the uppermost
lower mantle. The slab and the plume have likely been interacting with one another for at least 20 Ma
resulting in a tear in the slab propagating from the north to the south. The broadly low-velocity mantle south
of the MTJ is punctuated by high-velocity drips in several locations beneath southern California and Nevada.
The DNA08 models reveal extremely heterogeneous mantle structure beneath the western USA and suggest
we are only just beginning to image the complex interactions between geologic objects. The transportable
array allows for analysis of the relationships between these anomalies in an internally consistent single
tomographic model.
http://dna.berkeley.edu
S31D-03
Spot checks on North American mantle structure using Earthscope's Transportable Array and preceding PASSCAL arrays
We have measured relative arrival times of teleseismic P and S waves at seismic arrays in North America. These arrays are primarily IRIS-PASSCAL arrays in tectonically stable North America that were not utilized for teleseismic tomography. Teleseismic S-wave delay times averaged for all events measured at a given PASSCAL station illustrate intracratonic structural variations of up to 1 s over several hundreds of km, while the delay-time difference between on-craton stations and station on the Paleozoic margin of North America is more than 2.5 s over less than 500 km. Such measurements can provide integral checks on numerous tomographic models that have been developed since the recent onset of Earthscope. The ratio of S delays to P delays across the various arrays in stable North America is not significantly different from 3, but the best fitting value is 2.5. This is consistent with the relatively cool thermal state of the cratonic lithosphere of stable North America implying a smaller contrast between the temperature sensitivity of S and P velocities than for hotter mantle, such as that beneath tectonically active western North America. The tectonically active part of North America has been sampled spectacularly by Earthscope's Transportable Array (TA). Across this western configuration of the TA teleseismic P-wave arrival time delays differ by up to 4 s to date. A pattern of delays from incidence angles from the southeast shows a region of increased delays beneath northern Arizona, as well as near the Snake River Plain. A thin north-south oriented line of stations along the Cascades shows distinctly early arrivals. Expectedly, but never as beautifully illustrated with smaller arrays, re-measuring the P-wave arrival times after low-pass filtering produces a much smoother and damped pattern of delays, not unlike that found for the S-wave delays, which is consistent with increased averaging of structure within larger Fresnel zones at lower frequencies.
S31D-04
Pn Tomography of the Western United States Using USArray
USArray has now provided several years of high-quality seismic data and improved ray coverage for much of the western United States, and will enable improved resolution for studies of the lithospheric and deeper structure of the North American continent. Here we analyze Pn arrival times from the transportable stations of USArray to resolve crust and uppermost-mantle structure. We begin with 59,500 Pn picks from January 2005 to September 2007 as measured by the Array Network Facility (ANF) at epicentral distances from 150 to 1800 km. These picks are derived from 430 individual stations at ~70 km spacing and 7630 earthquakes and quarry blasts. Applying the classic time-term method, we use a regularized least-squares inversion to estimate crustal thickness variations and image velocity perturbations in the uppermost mantle just below the Moho. Our crustal thickness map generally agrees with receiver function results from other researchers but differs in some details. Our preliminary Pn velocity inversion indicates an average upper- mantle velocity of 7.9 km/s, with higher velocities beneath central Nevada and eastern Washington and lower velocities near the California-Mexico border, the Sierra Nevada, the coastal Oregon region, and the greater Yellowstone area. We are now extending our modeling method to include possible azimuthal anisotropy in Pn velocities for those regions with adequate ray coverage. These results should complement other seismic studies (e.g., body and surface wave tomography, shear-wave splitting, etc.) to provide information about composition, temperature, and tectonic processes in the western United States.
S31D-05
The Moho and the Lithosphere-Asthenosphere Boundary under the western U.S. from USArray PdS Receiver Functions
We have made PdS receiver functions from USArray Transportable Array data to examine lithosphere and
upper mantle structure beneath most of the tectonic western US. We present a CCP stacked volume of
~10,000 receiver functions from 61 events recorded at 520 stations. Receiver functions were formed
both by spectral division with water level stabilization, and by using a time domain iterative deconvolution
method. The data were depth converted, laterally repositioned, and CCP stacked using linear tomography
corrections for a 3D velocity model constructed by combining the mantle model NA04 (van der Lee and
Frederiksen, 2005) with the Crust2.0 model (Bassin et al., 2000).
The PdS stack shows a large number of lithospheric features: The Moho is well imaged and forms a nearly
continuous surface under the most of the western U.S. The strength of the Moho conversion is large under
most of the Basin and Range, the Mojave block, parts of the Snake River plain, and the southern Colorado
Plateau. The Moho signal is weak in the drip region of the southern Sierra Nevada, around the Wallowa
Mountains where another delamination event has been identified, in north central Nevada, and in the arc and
backarc region of Cascadia. The Moho of the oceanic crust in the subducting Juan de Fuca plate can be
traced to ~60 km depth in parts of Cascadia.
A series of events that we interpret as the lithosphere-asthenosphere boundary (LAB) are visible under much
of the west. We identify the LAB as one or more negative polarity events that appear at a range of depths
(~35-100 km), but above the zone of contamination from crustal multiples. The LAB is not a continuous
interface under the western U.S., but is rather a series of discontinuous surfaces. The Sierra Nevada has
complications in the lower crust and upper mantle, with the LAB rising to relatively shallow levels in the
western drip region. In the northernmost Basin and Range, the LAB appears to shallow almost to Moho
depths. Beneath the rest of the Basin and Range the LAB is at about 60km depth. To the north beneath the
Snake River plain and Columbia flood basalts, the LAB is deeper, broader, and centered at 60-75 km depth.
LAB depth increases rapidly from the Basin and Range to its greatest depth, ~100 km, under the central
Colorado Plateau. It also reaches this depth in places under the Southern Rocky Mountains, but overall, the
LAB depth is highly variable under the Rockies.
Assuming Airy isostasy and using 100 km as the compensation depth, we calculate the average density of
the asthensophere under the western U.S. to be 3161 kg/m3. We are currently comparing USArray
PdS images to SdP images to further constrain the LAB (Miller et al., 2008, this meeting).
http://terra.rice.edu/web/people/levander/USArray.html
S31D-06
High Resolution Transition Zone Structures Beneath Western USA
The western USA is a complex tectonic setting resulting from long lived ocean-continent convergence followed by transform motion along part of the margin. Apparently coupled downgoing slabs and mantle upwellings have been identified in a number of parts of the western U.S. The dense USArray deployment with a uniform grid of ~ 0.7 degree provides us an unprecedented opportunity to explore more detailed regional mantle transition zone structures: such as regional topography of the 410 and 660 discontinuities, and regional variations of thickness of the transition zone. We calculated 8117 receiver functions from 44 events (Mw>6.0) recorded by USArray in the period of 2005 to 2008. The data were lowpass filtered at 1 Hz, and then repositioned in space using a 3D hybrid velocity model based on the NA04 shear wave model (van der Lee and Frederiksen, 2005), and Crust2.0 (Bassin et al., 2000) . We constructed a Common Conversion Point (CCP) stacked volume for all of our receiver functions which provides a very clear image of the transition zone discontinuities that appears consistent with recent high resolution tomography models of the western U.S. (e.g. Sigloch et al., 2008; Burdick et al., 2008). The average thickness of transition zone in our region of study is ~6 km greater than the global average, 248 km vs 242 km, (Lawrence and Shearer, 2006), with a large standard deviation σ=±17.4km, which in part reflects the complicated local geodynamic processes in the region. Generally speaking from the latitude of the Mendocino Triple junction and the northern Basin and Range to the north, localized uplift of the 410 and downwarping of the 660 can be correlated to subduction features, either slabs traceable to the surface of the tomography models, or what appear to be slab fragments. To the south in the modern transform regime through the Salton Trough, the transition zone discontinuities are largely flat, with the transition zone being ~240-245 km thick. One local feature of interest is a substantial uplift (~30 km) of the 410 beneath the southern Sierra Nevada drip and northern Transverse Ranges, which is not associated with a positive tomography anomaly. We are currently examining the sensitivity of the discontinuities depths to the velocity model, by reposition the receiver functions using more recently developed 3D velocity models.
S31D-07
P-wave tomography of the Mendocino Triple Junction, northern California
An 80 station broadband seismic array (FAME-Flex Array MEnodocino) across the greater northern California area has been deployed. We present results of P-wave tomography using finite-frequency sensitivity kernels. Data analyzed includes 50 Transportable Array stations and 80 stations from FAME, spanning a time period from July 2005 to September 2008. Preliminary results show fast velocity anomalies beneath the northern Great Valley area, likely owing to the southern edge of the Juan de Fuca slab and the Great Valley Ophiolite. Slow anomalies occur in the Clear Lake volcanic area, the northernmost volcanic center associated with the northward-propagating Mendocino Triple Junction (MTJ). We target a precise location of the southern Juan de Fuca slab, mantle structure beneath the southernmost Cascades volcanic arc, and unique features adjacent to the MTJ.
S31D-08 INVITED
CAFE: a seismic investigation of water percolation in the Cascadia subduction zone
Subduction zones transport water into the Earth's interior. The subsequent release of this water through dehydration reactions may trigger intraslab earthquakes and arc volcanism, regulate slip on the plate interface, control plate buoyancy, and regulate the long-term budget of water on the planet's surface. As part of Earthscope, we have undertaken an experiment named CAFE (Cascadia Arrays for Earthscope) seeking to better constrain these effects in the Cascadia subduction zone. The basic experiment has four components: (1) a 47-element broadband imaging array of Flexible Array instruments integrated with Bigfoot; (2) three small-aperture seismic arrays with 15 additional short-period instruments near known sources of Episodic Tremor and Slip (ETS) events; (3) analysis of the PBO and PANGA GPS data sets to define the details of episodic slip events; and (4) integrative modeling with complementary constraints from petrology and geodynamics. Here, we present a summary of the results that have been obtained to date by CAFE, with a focus on high-resolution seismic imaging. A 250 km-long by 120 km-deep seismic profile extending eastward from the Washington coast was generated by 2-D Generalized Radon Transform Inversion of the broadband data. It images the subducted crust as a shallow-dipping, low-velocity layer from 20km depth beneath the coast to 40km depth beneath the forearc. The termination of the low-velocity layer is consistent with the depth at which hydrated metabasalts of the subducted crust are expected to undergo eclogitization, a reaction that is accompanied by the release of water and an increase in seismic velocities. Slab earthquakes are located in both the oceanic crust and mantle at depths <40 km, and exclusively in the oceanic mantle at greater depth, as would be expected if they are related to slab dehydration. Two ETS events have occurred during the course of the deployment. They were precisely located and are confined to the region above which the crust exhibits low-velocities and is believed to undergo progressive dehydration, further supporting the proposition that water plays a role in ETS.