1. Scientific Drilling Into the San Andreas Fault zone San Andreas Fault Observatory at Depth (SAFOD)

2. Objectives Understand the physical and chemical processes that control deformation and earthquake generation within active fault zones Make near-field observations of earthquake nucleation, propagation and arrest to test laboratory-derived concepts of faulting physics. Mechanically weak – Low heat flow – High-angle maximum principal stress S Hmax

3. Why Parkfield? Transition between the locked segment and aseismic creeping segment of the SAF 1857 M 8.2 Fort Tejon event 1906 M 7.8 San Francisco event Repeating earthquakes Shaded relief map of California (Hickman et al., 2004)

4. Repeating earthquakes M 6 EQs have occurred on the Parkfield section at fairly regular intervals - in 1857, 1881, 1901, 1922, 1934, 1966 and 2004 1922, 1934 and 1966 events ruptured the same segment of the fault in a similar manner Significant EQs in the Parkfield since 1850Comparison between 1922 and 1934 events

5. InSAR measurement Linear surface displacement rate between 1993 and 2004 (de Michele et al., 2011) SAFOD The bimodal distribution is consistent with right strike- slip motion The sharp discontinuity in the InSAR signal is a direct consequence of the surface creep The sharpness fades progressively to SE PKF

6. Other measurements The coseismic moment release of the 2004 event is as little as 25% of the total (Johanson et al., 2006) The creep rate from InSAR is consistent with short-range EDM, creepmeter, and alignment array Along-strike surface slip of the SAF from NW to SE (de Michele et al., 2011)

7. Geological & Geophysical Cross-sections Pilot Hole: 2.2-km-deep, drilled through Salinian granite Main Hole: penetrated two actively deforming zones as SDZ and CDZ at ~2.7 km vertical depth Resistivity structure from surface magnetotellurics (Hickman et al., 2004) Geologic cross-section parrallel to the trajectory of SAFOD (Zoback et al., 2011)

8. Geophysical logs from Main Hole Geophysical logs as a function of measured depth (Hickman et al., 2004) Damage Zone – ~200-m-wide – Low P and S velocities, low resistivity – Result of both physical damage and chemical alteration of the rocks due to faulting and fault- related minerals

9. Friction experiment Stepping tests – slip velocity is suddenly increased by an order of magnitude – Friction increases immediately then decay to a new steady- state value Slide-hold-slide tests – Steady-state sliding is followed by a holding for t – followed by a resumption of slip at the former slip velocity Experimental data and frictional-healing determination (Carpenter et al., 2012)

10. Friction behavior: fault gouge Powdered, clay-rich foliated gouge Frictional strength is as low as μ=0.21 in the fault zone SEM image showing shear zones in clay- rich foliated gouge (Carpenter et al., 2011) Frictional strength and healing behavior (Carpenter et al., 2011)

11. Friction behavior: intact cores Intact fabric, saponite and smectite clay Fault zone rock is extraordinary weak (μ=0.09-0.25), with the lowest friction values in the center of the fault Coefficient of sliding friction (Carpenter et al., 2012) Mohr-Coulomb failure envelope (Carpenter et al., 2012)

12. Summary The laboratory data offer a coherent explanation for the weakness of the SAF The intact samples have different friction behavior from the powdered samples due to the mineral fabric.