San Gorgonio Pass Special Fault Study Area

Transcription

1 San Gorgonio Pass Special Fault Study Area Michele Cooke, David Oglesby and Doug Yule Whitney Behr, Kim Blisniuk, Jim Brune, Sarah Carena, Judi Chester, Gary Fuis, Thomas Goebel, Peter Gold, Egill Hauksson, Dick Heermance, Katherine Kendrick, Vicki Langenheim, at Lifton, Jon Matti, Sally McGill, Craig icholson, Mike Oskin, Kate Scharer, Warren Sharp, Zheqiang Shi, Kerry Sieh, Josh Spinler and Mike Rymer.

2 Small region within a large system The San Gorgonio Pass comprises the southern Big Bend of the San Andreas fault Field et al., 2014, UCERF3 2

3 The San Gorgonio Pass (modified from McGill et al., GSA Bull. 2013) 3

4 Guiding questions What is the subsurface geometry of active faulting through the San Gorgonio Pass? What is the earthquake potential in the San Gorgonio Pass? What is the probability of a through-going San Andreas rupture? 4

5 Time line of activity Start of SCEC4 2 day Workshop 33 participants Established as one of two SFSA 1 day workshop 54 participants 2015 annual meeting end of SCEC CFM v.4 icholson et al. McGill et al., GSA Bull Herbert, Cooke & Marshall, JGR CFM v.5 Plesch et al Kendrick et al., JGR McGill et al., JGR Goebel et al., JGI Gold et al. JGR Many future papers 5

6 Data and tools icholson Hauksson Goebel Carena seismicity Spinler McGill Geodesy Cooke geophysics Crustal deformation models geology Behr, Blisniuk, Brune, Chester, Gold, Heermance, Kendrick, Lifton, Matti, McGill, Oskin, Scharer, Sharp, Sieh, Rymer, Yule Fuis Langenheim Oglesby Dynamic rupture models Shi, Day, Oglesby, Tarnowski 6

7 What is the active geometry of faults? SSIP seismic line 6 crosses within the SGP SFSA Reveals multiple E dipping strands of the San Andreas Fuis, Bauer, Goldman, Ryberg, Langenheim, Scheirer, Rymer, Stock, Hole and Catchings, submitted 7

8 What is the active geometry of faults? CFM 5.0 Updated 3D Fault Set Mill Creek strand orth Frontal Thrust fault orth Frontal Deep detachment Crafton Hills complex San Gorgonio Pass thrust San Andreas-Banning strand Mission Creek strand SGP detachments Increasing fault complexity in SGP & adjacent areas defined by seismicity blind, oblique en echelon faults blind Palm Springs fault Garnet Hill strand San Jacinto fault zone orth Palm Springs fault Pinto Mountain fault icholson, Plesch and Hauksson 8

9 What is the active geometry of faults? 9

10 How is slip partitioned? Mc Gill et al., 2015 McGill, Spinler et al., unpublished 2015 ] 10

11 How is slip partitioned? Variable slip rates along the San Andreas through the San Gorgonio Pass. Mechanical models match this variability data gap (McGill et al., GSA Bull. 2013) (modified from Herbert & Cooke, BSSA 2012) (McGill et al., in prep) 11

12 How is slip partitioned? Active strands Mill Creek-Mission Creek Banning-Garnet Hill (Gold, Behr et al., JGR 2015) 12

13 How is slip partitioned? Mill Creek strand o offset of Holocene/Latest Pleistocene alluvial deposits at Upper Raywood Flats (Kendrick et al., JGR. 2015) The Pinto Mountain fault offsets the Mill Creek strand (Kendrick et al., JGR. 2015) 13

14 Mission Creek fault How is slip partitioned? The Mill Creek strand Lidar scarp analysis suggests that slip may by-pass upper Raywood flats via the Galena Peak fault. Fault kinematics consistent with slip transfer West of Mill Creek Fault Bend Mill Creek fault n=107 Poles to Faults Slip Vectors n=79 Subsidiary Fault Kinematics East of Mill Creek Fault Bend n=458 Poles to Faults Slip Vectors n=265 Raywood Flat Area n=90 n=43 Poles to Faults Slip Vectors Galena Peak fault B-Axes Transfer Fault n=79 Fault Bend Transfer fault B-Axes n=265 MCJO Raywood Flat n=43 B-Axes MFJO Mission Creek Fault n=290 n=199 Poles to Faults Slip Vectors Mission Creek fault n=15 n=5 Poles to Faults Slip Vectors n=199 B-Axes B-Axes n=5 (Morelan, Oskin, Chester and Elizondo, in prep) 14

15 How is slip partitioned? The Banning Strand Offset alluvial fan reveals relatively slow slip rates ~(4-5 mm/yr) along the Banning fault Slip rate at SE end of Indio Hills (Scharer) is also 2-6 mm/yr Holocene rate: /-1.6 to /-0.9 mm/yr Gold, Behr et al., JGR

16 How is slip partitioned?: Banning strand & San Gorgonio Pass thrust Heermance and Yule, in prep 16

17 How is slip partitioned?: The Mission Creek strand Mission Creek strand: mm/yr (~90 ka, ~70 ka, & ~25 ka) Banning strand: 4-6 mm/yr since ~6ka 2.1 to 2.4 km offset since / -7 ka: / -3 mm/yr 1.3 to 1.7 km offset since / -2 ka: / -3 mm/yr 0.6 to 0.9 km offset since / -4 ka: / -6 mm/yr Blisniuk, Scharer, Sharp, Burgmann in prep 17

18 How is slip partitioned? effect of active Mill creek τ τ slip o slip Mill Creek not in model Mill Creek slips Strike slip is transferred to the Mill Creek strand. San Jacinto and Banning have slower slip rates Cooke, in prep. 18

19 How is slip partitioned? effect of active Mill Creek Slip partitioning is sensitive to active fault geometry through the pass dextral slip (mm/yr) San Bernardino Mill Creek Mission Creek Banning Coachella??? distance from Cajon pass (km) dextral slip (mm/yr) ? ?? distance from Cajon pass (km) Cooke, in prep. 19

20 What is the stress state? Insights from microseismicity Large stress drops within the San Gorgonio Pass Goebel et al., JGR

21 What is the stress state? Insights from crustal deformation models Off-fault deformation matches better the stress inversions from focal mechanisms than interseismic stressing rates 35 Interseismic stressing rate Focal Mechanisms Yang et al 2012 Off-fault stressing rate Mean ormal Stress Compressive <->Tensile Cooke, in prep

22 Correlation to stress drops in SGP Regions of large stress drops correlate with compressive mean stress of off-fault deformation Stress drop may relate to fault geometry rather than material contrast at step in base of seismicity Off-fault stressing Depth =5km Depth =11km Goebel et al., JGR 2015 Mean ormal Stress Compressive <->Tensile Depth =17km Goebel et al., JGR

23 Can earthquakes rupture through the Pass? Paleoseismology Only 4 earthquakes in 5500 years Complex slip patterns: m uplift in single event Most recent event was ~1400 A.D. The Cabazon MEGA trench ~0.5 m uplift; ~1 m slip Yule, Scharer in prep 23

24 full stress tensor, mean of van Mises stress: σ vm = 8.5e-01 AU sticks: horizontal projection of most compressive eigenvector, e 3, length scaled background shading: stress, A φ of Simpson (JGR, 102, 17909, 1997) full stress tensor, mean of van Mises stress: σ vm = 1.5e+00 AU sticks: horizontal projection of most compressive eigenvector, e 3, length scaled background shading: stress, A φ of Simpson (JGR, 102, 17909, 1997) SCEC CSM, v made with GMT - twb-at-usc-dot-edu SCEC CSM, v made with GMT - twb-at-usc-dot-edu full stress tensor, mean of van Mises stress: σ vm = 1.3e+02 AU sticks: horizontal projection of most compressive eigenvector, e 3, length scaled background shading: stress, A φ of Simpson (JGR, 102, 17909, 1997) full stress tensor, mean of van Mises stress: σ vm = 1.1e+02 AU sticks: horizontal projection of most compressive eigenvector, e 3, length scaled background shading: stress, A φ of Simpson (JGR, 102, 17909, 1997) SCEC CSM, v made with GMT - twb-at-usc-dot-edu SCEC CSM, v made with GMT - twb-at-usc-dot-edu Can earthquakes rupture through the Pass? Dynamic rupture San Andreas Fault x 2 x 3 x 1 San Bernadio Segment (km) planar free surface ( x 2 = 0 ) Shi and Day San Bernadino Section Coachella Valley Segment Elsinore Fault San Jacinto Fault 0 10 Banning Strand 20 (km) Coachella Section SCEC CFM-v5 rupture nucleation centers north and south of SGP Candidate SCEC CSMs as Reference of Stress Input Artificial band-limited roughness superposed on fault 3 km km / 0 [%] / 0 [%] additionally with σ vm e 3 azimuth, α = 7.2 [ o ] normal strike-slip reverse A φ [AU] A φ [AU] additionally with σ vm e 3 azimuth, α = 7.2 [ o ] normal strike-slip reverse km km / 0 [%] additionally with σ vm e 3 azimuth, α = 9.4 [ o ] normal strike-slip reverse A φ [AU] / 0 [%] additionally with σ vm e 3 azimuth, α = 9.7 [ o ] normal strike-slip reverse A φ [AU] The initial stress field dominates rupture behavior, compared other factors including small-scale fault geometric complexities. Different stress models in their present forms will lead to vastly different rupture scenarios regarding the likelihood of throughgoing rupture along SGP. 24

25 Can earthquakes rupture through the Pass? Ruptures starting on the Banning strand can pass to the San Bernardino strand Ruptures from the San Bernardino strand are less likely to pass to the Banning. Tarnowski, Kyriakopoulos, and Oglesby Tarnowski and Ogelsby 25

26 Can earthquakes rupture through the Pass? Ruptures starting on the Banning strand can pass to the San Bernardino strand Ruptures from the San Bernardino strand are less likely to pass to the Banning. Tarnowski, Kyriakopoulos, and Oglesby 26

27 San Gorgonio Pass SFSA outcomes Some but not all ruptures can pass through the SGP as large events. The region hosts slow slip rates, low strain rates and unusually high stress drops, which owe to fault geometry. Activity distributed among multiple strands rather than along one dominate structure. Cross-disciplinary discussions and collaborations Leveraging for projects funded by USGS and SF. 27

28 Thank you! Photo along the Mill Creek strand of the San Andreas fault 28