Observations and Interpretation of Strain Changes Caused by Volcanic Activity: Significance for Earthscope

Sacks-Evertson borehole strainmeters installed in a number of tectonically active regions have recorded strain changes due to activity in volcanic systems in Japan, Iceland and California. The eruption of Izu-Oshima, Japan, in 1986 produced observed strain changes at distances up to ~55 km from the volcano. Data for the first stage of the eruption show that, while the ultimate reservoir was relatively shallow (5-10 km deep), during and immediately following the eruption that shallow reservoir was replenished by magma flowing up from a reservoir about 30 km deep. The second stage of the eruption (a fissure eruption) was preceded (by almost 2 hours) by strain changes due to dike initiation and propagation. Large strain changes continued for about 10 days after the eruption. These strain changes are well satisfied by a model comprised of a reservoir and 2 dikes. One dike (to the north-west) resulted in fissures in the caldera and on the north-west flank; formation of this dike initiated about 2 hours before the fissuring. The second dike, which did not break the surface, propagated more slowly to the south-west and caused strain changes that continued for many days after the eruptive activity. The 1991 and 2000 eruptions of Hekla, Iceland, were well recorded by borehole strainmeters. Only one site was close enough (15 km) to record pre-eruptive (by about 30 min) changes due to dike propagation but the strain data were sufficient to allow modeling of the eruption in terms of a deflating Mogi reservoir together with a vertically propagating dike. The 2000 eruption was somewhat smaller than the 1991 event but the strain records were very similar. That, together with the increase in local seismicity, allowed a specific prediction and warning to be issued 20 minutes before the eruption. These examples (and data from Long Valley showing significant deformation triggered by the Landers earthquake) have allowed new insight into the physics of volcanic activity despite the fact that the strainmeter arrays (Japan, Iceland) were not designed for monitoring the volcanoes. The Earthscope plan includes several arrays of borehole strainmeters for monitoring selected volcanoes. This clearly presents an opportunity for significantly improving our understanding of volcanoes.