Data Services Products: EMC-WUS-CAMH-2015-Supplemental 3D shear-wave velocity model of the western United States - Supplemental

Summary

The WUS-CAMH-2015 shear velocity model combines spatially interpolated/smoothed receiver functions, surface-wave dispersion and gravity observations through a 3D simultaneous inversion to image the subsurface S-wave velocity structure of the western U.S. region.

Description

The following contains images comparing single-station averaged receiver functions with spatially interpolated/smoothed receiver functions, clustering analysis results of the WUS-CAMH-2015 3D shear wave velocity model, synthetic tests of receiver function interpolation, applications of receiver function interpolation for smaller seismic arrays, and an animation of different time slices of receiver functions.

Images

Time slices at time 3.8 s from the receiver function wavefield in the western U.S. and adjacent Canada.
Each circle represents a seismic station at which we have computed a stacked (single-station averaged) receiver function. The color indicates the amplitude of (left) the receiver function stacked at each station or (right) the interpolated value of the receiver function wavefield. Inset shows the stack of all receiver functions, and the red line shows the location of the time slice, which was chosen to shown the structure near the time of the Ps converted phase arrival from the crust-mantle transition region. Note that spatially coherent part of the wavefield is extracted using receiver-function spatial smoothing.

Stacked and interpolated receiver functions for TA stations in central Nevada
The interpolated receiver functions were computed with the binned receiver functions within 250 km distance of the station using weights according to the distance from each station. Midrange ray parameter bins with Gaussian 1.0 are shown for the interpolated receiver functions. The simple stack uses all receiver functions with Gaussian 1.0. The amplitude scale is the same for all waveforms. Scattering effects that dominate several single-station stacked waveforms (e.g. TA-M08A, TA-M10A, TA-P08A, TA-P10A) are reduced in spatially smoothed receiver functions, but subtle variations are evident.

Stacked and interpolated receiver functions for TA stations in central Nevada

Automated model clusters generated using a simple Euclidean distance metric
Dark lines show the physiographic boundaries [Fenneman, 1917]. The large cell size prohibits resolution of smaller provinces, but the rough correspondence of the clusters to geologic regions are (1) Oceanic; (2) Gulf of California region; (3) NW Pacific Coast-Range System; (4) Northern Basin and Range; (5) Southern BR Region; (6) Colorado Plateaus; (7) Columbia River Plateau Region; (8) Northern Rocky Mountains; (9) Middle and Southern Rocky Mountains; (10) Great Plains; (11) Williston and Denver Basins; (12) Canadian Craton; (13) Gulf Coast Basin. The velocity profile within each cluster is summarized in the following image.

Automated model clusters generated using a simple Euclidean distance metric

Earth model clusters constructed using a Euclidean distance metric
The clusters correspond well to the known geomorphic/geologic provinces in the western United States. Westernmost regions are shown on the left, regions from the interior of North America toward the right, and oceanic regions are on the first left (sorted by ocean age). Velocity profiles within the cluster are sorted from north to south by row (like lines in a book). The large cell size prohibits resolution of small provinces, but the rough correspondence of the clusters to geologic regions are (1) oceanic, (2) Gulf of California region, (3) NW pacific Coast Range System, (4) Northern Basin and Range, (5) Southern Basin and Range region, (6) Colorado Plateaus, (7) Columbia River Plateau region, (8) Northern Rocky Mountains, (9) Middle and Southern Rocky Mountains, (10) Great Plains, (11) Williston and Denver Basins, (12) Canadian Craton, and (13) Gulf Coast Basin.

Earth model clusters constructed using a Euclidean distance metric

Shear wave speed model dendrogram detailing the structure of the clusters
Significant differences in terms of the Euclidean distance are found between major geologic provinces. CP means Colorado Plateau, and BR represents Basin and Range. The colors separate the clusters with a cutoff value of 0.65 km/s. We can clearly see some clusters are over-divided. Note the clusters shown in the above image are a simplified version of those color-coded in this figure. Some major clusters are labeled according to the distribution of the clusters with different cutoff values, which is not shown.

Shear wave speed model dendrogram detailing the structure of the clusters

Comparison of synthetic test receiver function stacks (left) and interpolation (right) for a Gaussian width parameter of 1.0
The red lines show the response from the reference model. The time shifts occurred every 10 waveforms on the right panel were caused by edge effects.

Comparison of synthetic test receiver function stacks and interpolation for a Gaussian width parameter of 1.0

Comparison of synthetic test receiver function stacks (left) and interpolation (right) for a Gaussian width parameter of 2.5

Comparison of synthetic test receiver function stacks and interpolation for a Gaussian width parameter of 2.5

Station map (a) and interpolated receiver functions (b) for XY network (Batholith Broadband Network) using different interpolation parameters
From left to right, the first panel shows the stacked receiver functions. The second panel shows the interpolated receive functions with d1 = 5 km and d2 = 10 km. The third panel uses d1 = 10 km and d2 = 20 km. The 4th panel uses d1 = 20 km and d2 = 40 km. The 5th panel uses d1 = 40 km and d2 = 80 km. The 6th panel uses d1 = 80 km and d2 = 150 km. The last panel uses d1 = 150 km and d2 = 250 km, which is used in the paper.

Station map,a, and interpolated receiver functions, b, for XY network  using different interpolation parameters

Station map (a) and interpolated receiver functions (b) for Y5 network (Canadian Rockies and Alberta Network) using different interpolation parameters
From left to right, the first panel shows the stacked receiver functions. The second panel shows the interpolated receive functions with d1 = 5 km and d2 = 10 km. The third panel uses d1 = 10 km and d2 = 20 km. The 4th panel uses d1 = 20 km and d2 = 40 km. The 5th panel uses d1 = 40 km and d2 = 80 km. The 6th panel uses d1 = 80 km and d2 = 150 km. The last panel uses d1 = 150 km and d2 = 250 km, which is used in the paper.

Station map, a, and interpolated receiver functions, b,for Y5 network using different interpolation parameters

Animation

Animation compares the interpolated receiver function wavefield with the simple-stacked one.

Citations and DOIs

To cite IRIS DMC Data Products effort:

  • Trabant, C., A. R. Hutko, M. Bahavar, R. Karstens, T. Ahern, and R. Aster (2012), Data Products at the IRIS DMC: Stepping Stones for Research and Other Applications, Seismological Research Letters, 83(5), 846–854, http://doi:10.1785/0220120032.

To cite the source of this Earth model:

  • Chai, C., C. J. Ammon, M. Maceira, and R. B. Herrmann (2015), Inverting interpolated receiver functions with surface wave dispersion and gravity: Application to the western U.S. and adjacent Canada and Mexico, Geophysical Research Letters, 42(11), 4359-4366, http://doi:10.1002/2015GL063733.

To reference the use of this Earth model hosted by EMC:

To cite IRIS Earth Model Collaboration (EMC) data product or reference use of its repository:

References:

Balmino, G., N. Vales, S. Bonvalot, and A. Briais (2012), Spherical harmonic modelling to ultra-high degree of Bouguer and isostatic anomalies, J. Geod., 86(7), 499–520, http://doi:10.1007/s00190-011-0533-4.

Ekström, G. (2011), A global model of Love and Rayleigh surface wave dispersion and anisotropy, 25–250 s, Geophys. J. Int., 187(3), 1668–1686, http://doi:10.1111/j.1365-246X.2011.05225.x.

Herrmann, R. B., C. J. Ammon, and H. Benz (2013), Group velocity dispersion for North America. [Available at http://www.eas.slu.edu/eqc/eqc_research/NATOMO.]

Laske, G., G. Masters, Z. Ma, and M. Pasyanos (2013), Update on CRUST1.0—A 1-degree global model of Earth’s crust, Abstracts EGU2013-2658 presented at 2013 EGU General Assembly Conference, Vienna, Austria, 7–12 April.

Maceira, M., and C. J. Ammon (2009), Joint inversion of surface wave velocity and gravity observations and its application to central Asian basins shear velocity structure, J. Geophys. Res., 114, B02314, http://doi:10.1029/2007JB005157.

Credits

Chengping Chai, Charles J. Ammon, Monica Maceira and Robert B. Herrmann

Categories