What can seismicity tell us about subterranean volcanic activity?

A zoned orthopyroxene. This example is NOT
from the paper being discussed. It is from here.

In a Science paper*** last week, Saunders et al. present a tantalizing relationship between episodes of seismicity at active volcanoes, the growth of rims on small crystals (of orthopyroxene), and intrusion of new batches of magma into shallow levels of the system. The underlying science is that the composition of crystals growing in a magma responds to any changes in the conditions of the magma--composition, volatile content, temperature, pressure, and oxidation state. Saunders et al. examined iron (Fe)-magnesium (Mg) zoning in 579 orthopyroxene crystals takenfrom Mount St. Helens eruptions between 1980 and 1986. They found four categories of zoned crystals: some with Fe-rich cores and Mg-rich rims, so-called 'reverse-zoned' crystals; some with Mg-rich cores and Fe-rich rims, so-called 'normal-zoned', multiply zoned, and patchy zoned. By modelling diffusion of elements they calculated that the crystals formed in <12 months. The authors attribute the zonation to changing oxygen fugacity and H2O concentrations.

The cores of the opx crystals appear to have been entrained from an old, partially crystallized magma. Normally zoned crystals may have resulted from rapid cooling of crystals suspended in a new magma pulse as they intrude into the reservoir, or of rapid crystallization induced by fluxing with CO2-rich gas. Reversely zoned crystals may have been produced as crystals already in the chamber are subjected to heating by the intruding magma. These crystals are first partially resorbed, and then overgrown with a Mg-rich zone that is in equilibrium with the new hotter, adjacent mel. Unzoned crystals are ambiguous and the authors offer several interpretations.

The authors then took the date that the crystal was erupted and subtracted the rim growth time to obtain the month in which the opx rim began, presumeably due to a magmatic perturbation. They plotted these dates on a graph of the seismicity and SO2 emission flux and found that the peaks in the growth dates corresponded to episodes of deep seismicity, especially in 1980 and 1982. The implication is that at least in some instances, the seismicity is related to intrusion of magma.

***Saunders, K., Blundy, J., Dohmen, R., and Cashman, K., Linking petrology and seismology at an active volcano, Science, 336, 1023-1027, 2012.