Interferometric synthetic aperture radar study of recent eruptive activity at Shrub mud volcano, Alaska
Shrub mud volcano is one of three large mud volcanoes that comprise the Klawasi Group in the Copper River Basin of southcentral Alaska. Except for minor discharges in the mid-1950s when the group was first described, Shrub was dormant prior to its reactivation in summer 1996. From 1997 to 1999, Shrub vigorously erupted more than 5 x 105 cubic meters of saline mud and carbon dioxide-rich gas at temperatures as high as 54 degrees C. Thereafter, activity waned but continued at least through 2015. We analyzed 192 interferograms derived from 106 synthetic aperture radar (SAR) images acquired by the JERS-1 (L-band), ERS-1/2 (C-band), RADARSAT-1 (C-band), and ALOS PALSAR (L-band) satellites to characterize ground deformation at Shrub before, during, and after its reactivation. Collectively, the interferograms span 1992–2000 and 2006–2011. We fit the observations with two deformation sources: a deflating, steeply dipping, pipe-like body under the summit area and an inflating, shallow-dipping, sill-like body under the southwest flank. Both sources are shallow, with centroids less than 1 km beneath the summit. Prior to reactivation, the flank source inflated ~0.35 x 105 cubic meters per year from July 1992 to May 1996. During eruptive activity, the summit source deflated at higher rates that peaked at ~8.71 x 105 cubic meters per year during May–November 1997 and continued at ~0.95 x 105 cubic meters per year during the 2006–2011 observation window. Cumulative source-volume loss is comparable to the volume of mud erupted. We interpret the summit source as the volcano’s feeder conduit that pressurized prior to the first SAR observation in 1992. Also before 1992, the conduit ruptured to feed a lateral intrusion of mud under the southwest flank, perhaps along a bedding plane in underlying glaciolacustrine deposits. The growing sill caused the southwest flank to inflate while it accommodated the mud supply from depth, which explains why we observed pre-eruptive inflation of the flank but not the summit. The summit began deflating when the conduit ruptured to the surface at the onset of eruptive activity. The flank source did not deflate concurrently because the weight of the thin overburden was insufficient to collapse the sill. There is a suggestion in the modern topography that lateral intrusions under Shrub’s southwest flank are a common feature of activity there.