Faults move on time scales from centimeters per year to meters per second. How faults respond to stress, through creep or earthquakes, depends on a range of factors including geometry, stress, mineralogy, pressure, temperature, and fluids. As a result, patterns of failure vary along the length of large tectonic faults, and with depth in the crust (e.g., Peng & Gomberg, 2010;Zoback et al., 1987). It is generally accepted that frictional faulting and large earthquakes nucleate in the seismogenic zone, but rupture may extend into the overlying diagenetic zone (Sibson, 1983). Surface rupture is seen for many large earthquakes, but not all. Blind faults can cause significant damages with little previous evidence of their existences (e.g., the Northridge earthquake: Shaw & Shearer, 1999). For earthquakes that do not show clear rupture on the surface (e.g., the Haiti earthquake: Hayes et al., 2010), there is a question of how deformation extends into the shallow crust and can we observe in-situ evidence of this strain?