3-D seismic attenuation structure of Long Valley caldera: looking for melt bodies in the shallow crust

Abstract

SUMMARY Unrest at Long Valley caldera (California) during the past few decades has been attributed to the ascent of hydrothermal fluids or magma recharge. The difference is critical for assessing volcanic hazard. To better constrain subsurface structures in the upper crust and to help distinguish between these two competing hypotheses for the origin of unrest, we model the 3-D seismic attenuation structure because attenuation is particularly sensitive to the presence of melt. We analyse more than 47 000 vertical component waveforms recorded from January 2000 through November 2016 obtained from the Northern California Earthquake Data Center. We then inverted the S-to-coda energy ratios using the coda normalization method and obtained an average Q of 250. Low attenuation anomalies are imaged in the fluid-rich western and eastern areas of the caldera, one of which corresponds to the location of an earthquake swarm that occurred in 2014. From a comparison with other geophysical images (magnetotellurics, seismic tomography) we attribute the high attenuation anomalies to hydrothermal systems. Average to high attenuation values are also observed at Mammoth Mountain (southwest of the caldera), and may also have a hydrothermal origin. A large high attenuation anomaly within the caldera extends from the surface to the depths we can resolve at 9 km. Shallow rocks here are cold and this is where earthquakes occur. Together, these observations imply that the high attenuation region is not imaging a large magma body at shallow depths nor do we image any isolated high attenuation bodies in the upper ≈8 km that would be clear-cut evidence for partially molten bodies such as sills or other magma bodies.