1

II. Technical Report

A. Introduction

The seismic catalog comprises the foundational dataset for assessing seismic hazards. The spatial dis-tribution of seismicity reveals subsurface active fault structures [e.g. Madden & Pollard, 2012; Nicholson et al., 2014]. Earthquake magnitude distributions describe Gutenberg-Richter relationships with a- and b-values that are critical for assessing seismic hazards [Field et al., 2014]. The assumption built into seis-mic hazard assessments, such as Unified California Earthquake Rupture Forecast [Field et al., 2014], is that the seismic catalog collected over the past several decades provides a representative sample of the state of stress within southern California. This assumption is challenged by the limited duration of the seismic catalog compared to the several hundred-year recurrence intervals along most faults within southern California. For example, during the recording of the seismic catalog the San Andreas Fault (SAF) south of Cajon Pass has had fewer earthquakes than nearby faults [e.g. Yang et al., 2012]. Alt-hough this fault has the greatest seismic potential in southern California, it may be under-sampled within the seismic catalog. Consequently, the utility of our interpretations from the seismic catalog, such as stress state within the crust, may be compromised if the seismicity within the past few decades does not adequately represent the loading of the major faults capable of producing devastating earthquakes.

During the recording of our seismic catalog in southern California, the southern SAF has been loading and presumably accumulating shear stresses without producing any large earthquakes that would signifi-cantly reduce the fault stress. Small earthquakes during this loading are revealed by microseismicity. While we might expect the focal mechanisms from microseismicity along the southern SAF to reveal strike-slip, the focal mechanism analysis of Yang et al. [2012] show that some portions of the fault system produce predominantly normal-slip events (Fig. 1). These focal mechanisms contrast long-term slip vec-tors along these portions of the SAF, such as the northern San Bernardino strand of the San Andreas fault, that show strike-slip [Mc Gill et al., 2013]. The focal mechanisms also disagree with crustal defor-mation models of the region that show dextral loading of the SAF system [SCEC CSM, 2017]. This dis-crepancy suggests that the seismicity in this region may not reflect the loading of the southern SAF sys-tem. A discrepancy between focal mechanism information and the expected loading may arise if the stresses associated with geometric irregularities along the fault produce failure of surfaces off of the pri-mary faults that slip differently from the overall dextral strike-slip loading.

Non-planar faults or slip gradients along planar faults, produce stresses off of the primary faults that may differ from the slip sense of the primary fault. For example, a strike-slip earthquake through a releasing

Figure1:A)MapoftheSanGorgonioPassregionshowsslipsenseofhighquality(nodalplaneuncertainty<45˚)focalmechanismsfromYangetal.(2012)indicatedincolors.B)ExpectedslipsenseatpositionsoffocalmechanismsinA)frominterseismicmodelwithdeeplockingdepth.HistogramsCandDshowthedistributionofslipstyle.Althoughweexpectdextralloadingofthisregion,muchoftheseismicityshowsnormalslip.

2

stopover may produce local normal slip aftershocks within the stepover. In southern California, this was observed with normal-slip aftershocks within stepovers of the strike-slip 1992 Landers rupture [e.g. Maerten et al., 2016]. In addition to aftershocks of moderate events (M>5.5) that can produce microseis-micity inconsistent with long-term fault loading, aseismic creep during the interseismic period can drive off-fault deformation that differs from slip on the primary fault. The microseismicity in our multi-decal cata-log may record both strike-slip loading of the fault between earthquakes as well as off-fault deformation that results from either aftershocks or aseismic creep. We use crustal deformation models of the San Gorgonio Pass region to show the potential for moderate magnitude earthquakes and creep to produce off-fault microseismicity that obfuscates our interpretation of crustal loading from the seismic catalog.

B. Methods Faults surfaces based on the SCEC Community Fault Model v. 4 are discretized into triangles and loaded by tectonic velocities at the edges of the model. With the 3D Boundary Element Method (BEM) models, faults of the CFM are extended to 35 km depth where they merge with a horizontal crack. This crack pro-vides distributed deformation below the seismogenic crust. Velocities consistent with the plate motions are applied to the outer edges of the deep horizontal crack. The shear traction-free faults in the model and the center of the deep horizontal crack slip in response to this loading and interaction with each other [e.g. Fattaruso et al., 2015]. The faults are discretized into triangles so that the meshed surfaces can suf-ficiently replicate the complex fault topology of the San Gorgonio Pass region. To investigate stressing off of faults, we have developed stressing rate models that simulate interseismic period using a two-step approach. Within the first step, multiple earthquake cycles are simulated within a steady-state model where all portions of the fault surfaces slip. The second step of the approach uses a back slip approach to simulate interseismic loading of the faults. The slip distribution from the steady-state model is applied to faults below the locking depth [e.g. Marshall et al., 2009]. If our catalog of micro-seismicity within the San Gorgonio Pass spanning the past few decades represents the interseismic stressing due to deep slip on the San Andreas fault, then the stresses from the microseismicity focal mechanisms should resemble the stressing rate pattern from the interseismic model. Deviations of the focal mechanisms from the interseismic model predictions may reveal local processes with the crust that produce seismicity that is inconsistent with fault loading.

C. Results We use the focal mechanisms from Yang et al. [2010] along with subsequent updates to the catalog and investigate focal mechanisms with nodal plane uncertainty < 45˚ between 1981 and 2015. The interseis-mic loading of the region shows predominantly strike-slip faulting that disagrees with the focal mechanism distributions. Areas of discrepancy include 1) region of the 1986 M6.0 North Palm Springs earthquake, 2) region of the 1992 M6.1 Joshua Tree and M7.3 Landers earthquakes, 3) adjacent to the northwestern potion of the San Jacinto fault and 4) northeast of the juncture of the Coachella segment of the SAF and the Mission Creek and Banning strands of the SAF (Fig 1).

1. Impact of recent moderate earthquakes Yang et al. [2012] report short temporary changes in slip style after large magnitude earthquakes in southern California. While these changes may not greatly impact the overall dis-tribution of slip sense of the entire southern California catalog [Yang et al., 2012], the occur-

Figure2:Slipsense(0-1:normal,1-2:strike-slip,2-3:reverse)throughtimeinthestudyarea.The1986NorthPalmSpringsand1992JoshuaTreeandLanderseventsaltertheaverageslipsense(blackline).Completenessdecreaseswithtime.Catalogexcludingevents<M1.5

3

rence of many aftershock focal mechanisms near the 1986 North Palm Springs, 1992 Joshua Tree and 1992 Landers earthquakes may skew the inferred stress state within our study area. In the time periods of one year after the 1986 North Palm Springs earthquake and 5 years after the 1992 Joshua Tree and Landers earthquakes, the average focal mechanism slip sense deviates from that before the events (Fig 2). A representative catalog for this study area should exclude events from these time periods. In order to reliably use the seismic catalog to assess off-fault deformation we should also exclude earth-quakes smaller than the completeness limit. Including these events would overly weight focal mecha-nisms that occur close to seismic instruments. The completeness of the study area catalog improves with with time as the seismic network is upgraded (Yang et al, 2012, Fig. 2). The period from 2001-2015 did not have any earthquakes M>5.5 in the study area and has completeness limit of M1.5. Excluding events smaller than M 1.5 from the catalog of high quality (nodal plane uncertainty < 45˚) events from 2001-2015 leaves a reliable catalog of ~6500 events. From this 14-year catalog of reliable events, we average the slip sense within 5 km by 5km moving win-dows in order to compare the bulk slip sense of the catalog to the interseismic model using a 25 km lock-ing depth. This locking depth is chosen to be deeper than than earthquake events. The primary region of discrepancy between the catalog and the interseismic loading is along the northern San Jacinto fault where most of the normal slip focal mechanisms occur on the northeast side of the fault (Fig. 3).

2. Interseismic off-fault deformation may indicate aseismic slip below 10 km on the San Jacinto fault

Some of the normal slip events that occur with-in the San Bernardino basin just to the north-east of the northern San Jacinto fault, have been associated with secondary faults [Ander-son et al., 2004]. Geophysical data reveal nor-mal faults that are sub-parallel to the San Jacinto fault and bound the edges of basement lows [Anderson et al., 2004]. The loading of these normal events between 2001 and 2015 remains enigmatic since the region experienc-es dextral interseismic loading (Fig. 3D). The majority of these normal slip events occur be-tween 15 and 20 km depth where rocks are expected to experience dextral loading from flow beneath the seismogenic crust. Such in-terseismic loading cannot account for why the secondary faults are slipping with normal slip. The location of the normal slip events is con-spicuously consistent with the dextral slip gra-dient observed along the northern San Jacinto fault within steady-state crustal deformation models of the region that simulate deformation over multiple earthquake cycles [Resor et al., 2016]. Increasing dextral slip rate from the northern tip of the San Jacinto fault southward produces a region of positive dilation within the San Bernardino basin, between the San Jacin-to and San Andreas faults. The deformation in the steady-state model results matches availa-ble slip rates along the San Jacinto and San Andreas faults [Fattaruso and Cooke, 2014;

Figure3:Slipsensefromfocalmechanisms(A&B),aninter-seismicmodelwithuniform25kmlockingdepth(B&C)andamodelwith10kmlockingdepthonthenorthernSanJacintofaultwith20kmlockingdepthelsewhere.Becausethenormalslipeventsoccuraround15kmdepth,interseismiccreepalongthenorthernSanJacintofaultbelow10km,whichhassouth-wardincreasingslipgradient,couldaccountforloadingofsec-ondaryfaultsandoff-faultdeformationwithnormalslip.

4

Resor et al., 2016] as well as subsidence of the San Bernardino basin [Cooke and Dair, 2011].

Asiesmic deep creep along the San Jacinto fault could account for the occurrence of off-fault normal slip events. If the northern San Jacinto fault were creeping below 10 km with slip rates that increase to the south, the crust northeast of the fault and below~ 10 km would experience dilation associated with that deep creep. This dilation would promote normal slip off-fault microseismicity. An interseismic model using 10 km locking depth on the San Jacinto fault and 20 km locking depth on all other faults, produces normal slip events that are spatially con-sistent with the previously enigmatic normal slip focal mecha-nisms in the catalog (Fig. 3E). While aseismic creep below 10-13 km has been proposed along the southern San Jacinto fault near the Salton Sea where heat flux is high [Fialko, 2006; Wdinowski, 2009; Smith-Konter & Sandwell, 2011], geodetic inversions for the northern San Jacinto show deep (~20 km) locking depth [Smith-Konter & Sandwell, 2011]. Because the San Jacinto and San Andreas faults approach within 10 km of each other at the San Bernar-dino basin, the inversions of geodetic data for locking depth in this region cannot distinguish the locking depths of the San

Jacinto from the San Andreas faults. We investigate the interseismic GPS velocities of one model with 15 km locking depth on all faults with another with 10 km locking depth on the San Jacinto and 20 km locking depth on the San Andreas fault. The station velocities cannot be distinguished (Fig. 4). Consequently, geodetic data cannot rule out deep aseismic creep on the northern San Jacinto fault that can account for the normal slip events within the San Bernardino basin. 3. Could off-fault deformation account

for anomalously high stress drops in the SGP?

The multi-earthquake cycle models with re-sidual off-fault stressing show spatially varia-ble mean normal stress that correlate with spatial variations in stress drop analyzed by Goebel et al. [2015] (Fig. 5). The SGP hosts a region of >18 MPa stress drop, larger than observed elsewhere in southern California [Goebel et al., 2015]. These anomalously high stress drops occur at 15-25 km depth within the foot wall of the San Gorgonio Pass thrust. The location and depth of these high stress drops closely correlates to locations of negative first stress invariant, high mean compression (blue on Fig. 5), in the off-fault stressing rates. This suggests that localized compression around the restraining bend may account for the anomalously high stress drops. The small micro-faults off of the prima-ry fault structures are clamped by this high compression so that they behave as stronger faults and produce larger stress drops. Local-ly high stress drops could be attributed to local increases in fault static frictional

Figure4:ComparisonofmodeledvelocityatGPSstationsformodelwith15kmlock-ingdepthonallfaultsandmodelwithlockingdepthsof10and20kmontheSanJacintoandSanAndreasfaultsrespective-ly.TheGPSdatacannottestthepresenceofdeepcreepalongtheSanJacintofault.

Figure5:leftcolumn)Meannormalstressingratefromdifferentdepthsofthemulti-cycleoff-faultdeformationmodel.Rightcol-umn)FocalmechanismstressdropatdifferentdepthsfromGoe-beletal.(2015).Thelocationofhighstressdropcoincideswithhighpotentialoff-faultclamping(negativemeannormalstress).

5

strength [Goebel et al., 2015]. However, heterogeneous fault strength is not required if the normal com-pressive stress is high within the region of high stress drop. D. Conclusions The correlation of stress patterns from focal mechanism inversions with patterns of off-fault stressing rate over multiple earthquake cycles suggests that the microseismicity of the San Gorgonio Pass region may be recording permanent off-fault deformation that is distinctly different from the long term loading of the San Andreas fault. The stressing rate models within the SCEC Community Fault Model produce a strike-slip stress regime within the San Gorgonio Pass [Hardebeck, 2014] in disagreement with focal mecha-nism stress inversions. This suggests that our current seismic catalog under-samples large strike-slip events along the SAF. Instead, the stress inversions from the seismic catalog may provide information about permanent distributed off-fault deformation than loading of the faults. The prominent signal of nor-mal faulting within the San Bernardino basin may indicate creep below 10 km along the San Jacinto fault. Such creep that has southward increasing dextral slip rate along the San Andreas fault could produce dilation within the San Bernardino basin that would unclamp secondary structures and lead to normal slip events during the interseismic period. The spatial correlation of mean normal compressive stress with regions of high stress drop from the focal mechanism inversions suggests that small faults here effectively strengthened due to localized clamping. High fault strength is consistent with long recurrence times of ruptures though the San Gorgonio Pass [Yule et al., 2014]. The multi-cycle model assumes uniform fault strength so the variation in mean nor-mal stress is entirely due to the influence of fault geometry. This study suggests that spatial variation in stress drop does not require higher explicit fault strength but the heterogeneous stress state can alter ap-parent fault strength. E. References Anderson, M., Matti, J. and Jachens, R., 2004. Structural model of the San Bernardino basin, California,

from analysis of gravity, aeromagnetic, and seismicity data. Journal of Geophysical Research: Solid Earth, 109(B4).

Bird, P., 2009, Long‐term fault slip rates, distributed deformation rates, and forecast of seismicity in the western United States from joint fitting of community geologic, geodetic, and stress direction data sets: Journal of Geophysical Research Solid Earth, v. 114.

Blisniuk, K., K. M. Scharer, W. D. Sharp, R. Burgmann, M. J. Rymer, and P. L. Williams (2013), New slip rate estimates for the Mission Creek strand of the San Andreas fault zone, paper presented at AGU Fall Meeting, American Geophysical Union, San Francisco, CA

Cooke, M.L., and Marshall, S.T., 2006, Fault slip rates from three‐dimensional models of the Los Angeles metropolitan area, California: Geophysical Research Letters 33.21.

Cooke, Michele L., Mariel T. Schottenfeld and Steven W. Buchanan, 2013. Evolution of Fault Efficiency at Restraining Bends within Wet Kaolin, Journal of Structural Geology, doi:10.1016/j.jsg.2013.01.010.

Cooke, M.L., and Dair, L.C., 2011, Simulating the recent evolution of the southern big bend of the San Andreas fault, Southern California: Journal of Geophysical Research 116, B04405, doi:10.1029/2010JB007835.

Dair, L., and Cooke, M.L., 2009, San Andreas Fault geometry through the San Gorgonio Pass, California. Geology [Boulder] 37, p. 119–122.

Fialko, Y., 2006. Interseismic strain accumulation and the earthquake potential on the southern San An-dreas fault system. Nature, 441(7096), pp.968-971.

Goebel, T. H. W., Becker, T. W., Sammis, C. G., Dresen, G., & Schorlemmer, D., 2014. Off-fault damage and acoustic emission distributions during the evolution of structurally complex faults over series of stick-slip events. Geophysical Journal International, 197(3), 1705-1718.

6

Goebel, T. H. W., et al., 2015, Stress-drop heterogeneity within tectonically complex regions: a case study of San Gorgonio Pass, southern California: Geophysical Journal International, v. 202.1, p. 514-528.

Gold, P.O., et al., 2015, Holocene geologic slip rate for the Banning strand of the southern San Andreas Fault, southern California: Journal of Geophysical Research Solid Earth, v. 120.8, p. 5639-5663.

Hatem, Alex, Michele L. Cooke and Elizabeth H. Madden, 2015. Evolving efficiency of restraining bends within wet kaolin analog experiments, Journal of Geophysical Research, 119, doi:10.1002/2014JB011735

Hardebeck, J., 2014. Current status of the CSM, at SCEC Community Stress Model Workshop October 27, 2014.

Hauksson, E. 2014. Average Stress Drops of Southern California Earthquakes in the Context of Crustal Geophysics: Implications for Fault Zone Healing. Pure and Applied Geophysics, 1-12.

Herbert, J.W., and Cooke, M.L., 2012, Sensitivity of the Southern San Andreas Fault System to Tectonic Boundary Conditions and Fault Configurations: Bulletin of the Seismological Society of America 102, p. 2046–2062.

Herbert, J.W., Cooke, M.L., Oskin, M., and Difo, O., 2014a, How much can off-fault deformation contrib-ute to the slip rate discepency within the Eastern California shear zone?: Geology [Boulder] 42(1). P. 71-75. doi:10.1130/G34738.1.

Herbert, Justin W., Michele L. Cooke and Scott T. Marshall, 2014b. Influence of fault connectivity on slip rates in southern California: Potential impact on discrepancies between geodetic derived and geolog-ic slip rates, Journal of Geophysical Research Solid Earth, doi:10.1002/2013JB010472K.J.

Kendrick, K.J., Matti, J.C., Mahan, S.A., (2015), Late Quaternary slip history of the Mill Creek strand of the San Andreas fault in San Gorgonio Pass, southern California: The role of subsidiary left-lateral fault in strand switching: GSA Bulletin, v. 127, no. 5/6, p. 825-849

Johnson, K. M., 2013, Slip rates and off‐fault deformation in Southern California inferred from GPS data and models: Journal of Geophysical Research Solid Earth, v. 118.10, p. 5643-5664.

Loveless, J.P., and Meade, B.J. , 2011, Stress modulation on the San Andreas fault by interseismic fault system interactions: Geology, 39.

Maerten, F., Madden, E.H., Pollard, D.D. and Maerten, L., 2016. Incorporating fault mechanics into inver-sions of aftershock data for the regional remote stress, with application to the 1992 Landers, Califor-nia earthquake. Tectonophysics, 674, pp.52-64.

Marshall, S.T., Cooke, M.L., and Owen, S.E., 2008, Effects of Nonplanar Fault Topology and Mechanical Interaction on Fault-Slip Distributions in the Ventura Basin, California: Bulletin of the Seismological Society of America 98, p. 1113–1127.

Marshall, Scott T, Michele L. Cooke and Susan E. Owen, 2009. Interseismic Deformation Associated with Three-Dimensional Faults in the Greater Los Angeles Region, California, Journal of Geophysical Re-search, v, 114, B12403, doi:10.1029/2009JB006439.

Matti, J.C., Morton, D.M., and Cox, B.F., 1992, The San Andreas fault system in the vicinity of the central Transverse Ranges province, Southern California (No. 0196-1497). U. S. Geological Survey : Reston, VA, United States, United States.

McGill, S. F., Owen, L. A., Weldon, R. J., & Kendrick, K. J. (2013). Latest Pleistocene and Holocene slip rate for the San Bernardino strand of the San Andreas fault, Plunge Creek, Southern California: Im-plications for strain partitioning within the southern San Andreas fault system for the last∼ 35 ky.Geological Society of America Bulletin, 125(1-2), 48-72.

Meigs, A.J., Cooke, M.L., and Marshall, S.T., 2008, Using vertical rock uplift patterns to constrain the three-dimensional fault configuration in the Los Angeles Basin: Bulletin of the Seismological Society of America 98, p. 106–123.

7

Morelan, Alexander, Michael Oskin, Judith Chester, and Daniel Elizondo, 2014 Active slip on the Mill Creek fault rhough the San Gorognio Pass, paper presented at Southern California Earthquake Center Annual Meeting, Palm Springs, CA.

Oskin, Michael E., Alex E. Morelan, Daniel Elizondo, and Judith S. Chester, 2015. Assessing fault struc-ture and recent activity of the Mill Creek fault across Yucaipa Ridge and Upper Raywood Flat, San Gorgonio Pass Region, paper presented at Southern California Earthquake Center Annual Meeting, Palm Springs, CA.

Scharer, K. M., K. Blisniuk, W. Sharp, P. Williams, and K. Johnson (2014), Vertical deformation along the Indio Hills, San Andreas Fault, California, paper presented at Southern California Earthquake Center Annual Meeting, Palm Springs, CA..

Plesch, A., Shaw, J.H., Benson, C., Bryant, W.A., Carena, S., Cooke, M.L., Dolan, J., Fuis, G.S., Gath, E., Grant, L., Hauksson, E., Jordan, T., Kamerling, M., Legg, M., Lindvall, S., Magistrale, H., Nichol-son, C., Niemi, N., Oskin, M., Perry, S., Planansky, G., Rockwell, T., Shearer, P., Sorlien, C., Suess, M.P., Suppe, J., Treiman, J., and Yeats, R., 2007, Community fault model (CFM) for Southern Cali-fornia. Bulletin of the Seismological Society of America 97, p. 1793–1802.

Resor, P M. Cooke, S. Marshall and B.Madden, 2016 Role of fault geometry on the spatial distribution of the slip budget, SCEC annual meeting.

Smith‐Konter, B.R., Sandwell, D.T. and Shearer, P., 2011. Locking depths estimated from geodesy and seismology along the San Andreas Fault System: Implications for seismic moment release. Journal of Geophysical Research: Solid Earth, 116(B6).

Thomas, A.L., 1993, POLY3D: A three-dimensional, polygonal element, displacement discontinuity boundary element computer program with applications to fractures, faults, and cavities in the Earth’s crust. Stanford University, Stanford, California.

Titus, S.J., Dyson, M., DeMets, C., Tikoff, B., Rolandone, F., and Burgmann, R., 2011, Geologic versus geodetic deformation adjacent to the San Andreas fault, central California: Geological Society of America Bulletin, v. 123, p. 794– 820, doi:10.1130/B30150.1.

Wdowinski, S., 2009. Deep creep as a cause for the excess seismicity along the San Jacinto fault. Nature Geoscience, 2(12), pp.882-885.

Working Group on California Earthquake Probabilities, 2008, The Uniform California Earthquake Rupture Forecast, Version 2 (UCERF 2): U.S. Geological Survey Open-File Report 2007-1437 and California Geological Survey Special Report 203 [http://pubs.usgs.gov/of/2007/1437/].

Yang, W., E. Hauksson and P. M. Shearer, Computing a large refined catalog of focal mechanisms for southern California (1981 - 2010): Temporal Stability of the Style of Faulting, Bull. Seismol. Soc. Am., June 2012, v. 102, p. 1179-1194, doi:10.1785/0120110311, 2012.

Yule, D., and Sieh, K., 2003, Complexities of the San Andreas Fault near San Gorgonio Pass; implica-tions for large earthquakes: Journal of Geophysical Research 108.

Yule, D., R. Heermance, I. Desjarlais, and K. Sieh (2014), Late Pleistocene and Holocene Rates of Slip and Size of Prehistoric Earthquakes on the San Andreas Fault System in San Gorgonio Pass, in Southern California Earthquake Center Annual Meeting, edited, Palm Springs, CA.