EARTHQUAKE-INDUCED FLUID PRESSURE CHANGES AT LONG VALLEY, CALIFORNIA: RESULTS FROM THE LONG VALLEY EXPLORATION WELL

At Long Valley caldera, California, both local and distant earthquakes have produced persistent water level drops in wells on and north of the caldera’s resurgent dome, accompanied by earthquake-induced fluid level increases in the caldera’s hydrothermally active south moat.

Earthquake-induced water-level drops on and north of the resurgent dome may represent diffusion of localized fluid pressure drops induced abruptly by seismic waves. We evaluate this hypothesis using recent data from the Long Valley Exploration Well (LVEW), a 3 km deep borehole penetrating the center of the resurgent dome. The uncased interval of the LVEW penetrates relatively impermeable metapelite, and fluid communication is primarily through a conductive fracture 2.6 km below the surface. It has proven useful to collect data using pressure-temperature logging tools built to operate for extended periods at the depth of the fracture zone (temperature 102° C, pressure 22.3 MPa). Data from a pumping test in September, 2001, are consistent with a high-conductance planar fracture 10-100 m long in country rock with hydraulic diffusivity of 1-10 m²/s. Tidal amplitudes at the fracture zone depth are 1.93 kPa peak-to-peak, and the ratio of pressure change to strain based on the M2 earth tide constituent is 30.3 kPa/microstrain.

Both the Hector Mine, California (Oct. 16, 1999, M7.1) and Denali Fault, Alaska (Nov. 4, 2002, M7. 9) earthquakes produced steplike pressure drops in the LVEW which reached their minimum values (1.24 kPa and1.17 kPa, respectively) within 3 hours. In contrast, water level drops induced by the same two earthquakes at the shallower LKT well, 4 km NW of the center of the dome, required 10 to 15 days to reach their minima. The Denali Fault earthquake response at the LVEW was not only faster than at LKT, but also about twice as large, consistent with fluid pressure changes occurring abruptly with the passage of seismic waves at a location closer to the LVEW than to LKT. The pressure drops caused by the Hector Mine earthquake are less consistent with this conceptual model, because the fluid pressure dropped slightly more in LKT than in the LVEW. Continued fluid-pressure monitoring should yield more information about the time-space distribution of earthquake-induced fluid-pressure changes at Long Valley.