GPS Studies of Crustal Deformation

Regional deformation and kinematics from GPS data

Jessica Murray, Jerry Svarc, Elizabeth Hearn, and Wayne Thatcher U. S. Geological Survey

Acknowledgements: • Rob McCaffrey, Portland State University • UCERF3 Working Group • Robert Simpson, USGS

November 7, 2012

U.S. Department of the Interior U.S. Geological Survey Questions to address

• What conclusions can be drawn regarding the location and uncertainty in the Sierra Nevada-Great Valley microplate – Pacific plate pole of rotation?

• Can bounds be put on the total slip rate budget in the DCPP vicinity (i.e., faults west of the West Huasna fault), and slip rates on individual faults, especially the Hosgri fault?

• What are the patterns and rates of deformation west of the San Andreas fault in central California (as informed by block modeling)? GPS data used in this study

• Additional campaign and semi- permanent data collection to complement continuous GPS

• Data contain coseismic offsets and strong postseismic signals due to San Simeon and Parkfield earthquakes

• Many CGPS sites installed after earthquakes

• Earthquake corrections based on slip models can introduce bias into secular rate estimates

• Instead, use pre-earthquake data where possible or increase velocity uncertainties to account for contamination GPS velocity field

• Central California GPS data collected and processed for this project augmented by PBO solution for Sierra Nevada block and stable North America

• Additional sites included to define Pacific plate motion Block modeling of regional crustal deformation

• Features: • Implicitly satisfies relative plate motion • Provides some kinematic constraint on covariance among fault slip rates • Software: tdefnode (McCaffrey, GRL, ‘09) • Model parameters: • poles and rates of block rotation • fault coupling distribution • Block geometry: simplified version of UCERF3 block geometry with alternative boundaries west of SAF Questions to address

• What conclusions can be drawn regarding the location and uncertainty in the Sierra Nevada-Great Valley microplate – Pacific plate pole of rotation?

• Can bounds be put on the total slip rate budget in the DCPP vicinity (i.e., faults west of the West Huasna fault), and slip rates on individual faults, especially the Hosgri fault?

• What are the patterns and rates of deformation west of the San Andreas fault in central California (as informed by block modeling)? Location of estimated Pacific – North America pole of rotation Range of locations for SNGV-Pacific estimated poles from full block models (this study)

• SNGV-Pacific pole shows more similarity with other published results.

• SNGV-North America pole of SNGV-North rotation is poorly constrained, America both in this and other studies. Estimated block poles of rotation relative to Pacific, simple block geometry and simultaneous estimation of fault locking

CW rotation of named block relative to Pacific CCW rotation of named block relative to Pacific Questions to address

• What conclusions can be drawn regarding the location and uncertainty in the Sierra Nevada-Great Valley microplate – Pacific plate pole of rotation?

• Can bounds be put on the total slip rate budget in the DCPP vicinity (i.e., faults west of the West Huasna fault), and slip rates on individual faults, especially the Hosgri fault?

• What are the patterns and rates of deformation west of the San Andreas fault in central California (as informed by block modeling)? Velocity signal due to faults west of the San Andreas: Los Osos • Predicted velocity signal per mm/yr of slip on the Los Osos fault below 15 km

• 1 mm/yr of slip would generate ~0.13 mm/yr max. surface velocity Velocity signal due to faults west of the San Andreas: Shoreline • Predicted velocity signal per mm/yr of slip on the Shoreline fault below 15 km

• 1 mm/yr of slip would generate ~0.05 mm/yr surface velocity Velocity signal due to faults west of the San Andreas: Hosgri-San Simeon • Predicted velocity signal per mm/yr of slip on the Hosgri-San Simeon fault • Velocity signals due to locking on small, below 15 km low slip rate faults <= uncertainties • Slip rate estimates for closely-spaced • 3 mm/yr of slip would generate ~0.6 faults using elastic block model are mm/yr surface velocity highly correlated • Slip rates on major faults bounding regions that are large relative to the locking depth are better constrained • Kinematic block models can provide slip rate estimates on major faults and a measure of distributed strain rates • Other approaches allowing for different boundary conditions could shed more light on small faults Questions to address

• What conclusions can be drawn regarding the location and uncertainty in the Sierra Nevada-Great Valley microplate – Pacific plate pole of rotation?

• Can bounds be put on the total slip rate budget in the DCPP vicinity (i.e., faults west of the West Huasna fault), and slip rates on individual faults, especially the Hosgri fault?

• What are the patterns and rates of deformation west of the San Andreas fault in central California (as informed by block modeling)? Block model approach

• Initial block geometry: Salinian block is not divided (neither the green nor red boundary is used)

• Estimate block rotation and degree of fault locking

• Estimate uniform strain rate within Ventura and Mojave blocks

• Inspect uniform block strain rate calculated from velocity residuals to infer additional deformation within blocks Mojave Ventura Model 1 (Salinian block not divided) • San Andreas fault slip rate is lower than geologically-inferred rate • Hosgri fault slip rate: 3 mm/yr • Minor contraction west of SAF but extension on northern edge of Ventura block • Systematic residuals around Big Bend and in western versus eastern part of Salinian block

residuals

Freely-slipping boundary Red numbers: slip rates (mm/yr) Green strain rate tensors: estimated Boundary on which locking is estimated Black strain rate tensors: residual Large residuals in this area have been seen in other studies

Meade and Hager (JGR, 2005)

• Inclusion of “Big Pine” block reduces residuals

Becker et al. (GJI, 2005) • No corresponding mapped fault here Long-term postseismic effect?

Due to the viscoelastic postseismic effects of the 1857 Ft. Tejon earthquake, the present-day velocity estimates may be lower than their average values over the full earthquake cycle.

Predicted velocity perturbation based on one model of 1857 postseismic (Hearn et al., in review) Model 2: Constrain the SAF slip rate to 34 +/- 2 mm/yr

• Hosgri fault slip rates decrease • Residual strain rates west of San Andreas change little

No constraints With constraints

• Substantially larger left-lateral residuals along SAF when constraints are used • NE-SW pattern of residuals in Los Osos domain persists

No constraints With constraints

residuals residuals

Freely-slipping boundary

Boundary on which locking is estimated Model 3: Oceanic/West Huasna is a block boundary • Slip rate decreases along much of Hosgri fault in favor of contraction along Oceanic/W.H. • SAF slip rates still low • Residuals west of the Oceanic/W.H. are decreased; small NW/SE residual extension

residuals

Freely-slipping boundary Red numbers: slip rates (mm/yr) Green strain rate tensors: estimated Boundary on which locking is estimated Black strain rate tensors: residual Model 4: Constrain Oceanic/W. Huasna slip rate to < 1 mm/yr • Slip rate increases along Hosgri and San Andreas faults • Little residual strain rate west of the Oceanic/W.H. despite small increase in residuals

No constraints With constraints

Freely-slipping boundary Red numbers: slip rates (mm/yr) Green strain rate tensors: estimated Boundary on which locking is estimated Black strain rate tensors: residual Model 4: Constrain Oceanic/W. Huasna slip rate to < 1 mm/yr • Residuals west of Oceanic/West Huasna faults increase

No constraints With constraints

residuals residuals

Freely-slipping boundary

Boundary on which locking is estimated • The Sierra Nevada – Pacific pole of rotation inferred a simple three-block model is similar to that from published results, but results using a more complex block model are substantially different.

• Hosgri fault slip rates range from 0.5 – 3 mm/yr across the models tested.

• Velocity residuals show spatial patterns within subregions of the Salinian block, for example in the Los Osos domain.

• Model results show more evidence of NE-SW contraction between the Los Osos domain and the region between it and the San Andreas than within the Los Osos domain itself.

• While the pattern of residuals in the Los Osos domain is systematic, most of them are not significant given the standard errors on the data.

• Slip rates on the smaller faults in the region will be difficult to constrain given the uncertainties in the data and the limitations of kinematic models. Finite element models that allow for more physically-realistic boundary conditions may be able to say more. References

Time series analysis methodology: Langbein, J. (2004), Noise in two-color electronic distance meter measurements revisited, J. Geophys. Res., 109, doi:10.1029/2003JB002819.

Block modeling: McCaffrey, R. (2005), Block kinematics of the Pacific–North America plate boundary in the southwestern United States from inversion of GPS, seismological, and geologic data, J. Geophys. Res., 110,doi:10.1029/2004JB003307. McCaffrey, R. (2009), Time-dependent inversion of three-component continuous GPS for steady and transient sources in northern Cascadia, Geophys. Res. Lett., 36, doi:10.1029/2008GL036784.

Data and block geometry • Campaign GPS RINEX data: http://www.ncedc.org/survey-gps/ • PBO velocity field: http://pbo.unavco.org/data/gps • Processed results: http://earthquake.usgs.gov/monitoring/gps/ • UCERF3 block geometry and geologic slip rates: http://wgcep.org/UCERF3-TR8

Velocity field and initial block model results Murray-Moraleda, J., W. Thatcher, C. Onishi, J. Svarc (2011), Crustal deformation in the central California coast region inferred from Global Positioning System data, Abstract G13B-02 presented at 2011 Fall Meeting, AGU, San Francisco, Calif., 5-9 Dec.