Can Water Vapor Help Forecast When a Volcano Will Blow?

A widely used technique to monitor sulfur dioxide was tweaked to focus on water vapor at Peru’s Sabancaya Volcano. Results show that the volcano steamed up prior to its 2016 eruption.

The magma that bursts out of volcanoes is propelled upward largely by dissolved gases, which are released into the atmosphere once the molten rock approaches the surface. The most abundant of these gases is water vapor, and scientists have long searched for a way to accurately measure volcanic water vapor emission rates.

Now Kern et al. present a possible new method to do just that based on research conducted at the 6000-meter-high Sabancaya Volcano in Peru. Six months prior to the onset of the volcano’s current eruptive crisis, which began in November 2016, the team measured the volcanic water vapor output using a method called passive visible-light differential optical absorption spectroscopy (DOAS). DOAS instruments measure the absorption of sunlight by gases in the atmosphere and the volcanic plume above them. The technique is widely used to measure sulfur dioxide emissions, but these were the first successful DOAS measurements of volcanic water vapor.

The team found that prior to the current eruptive phase, the Sabancaya plume contained an exceptionally high, 1000:1 ratio of water vapor to sulfur dioxide, about an order of magnitude higher than typically found in volcanic gases. They hypothesize that as Sabancaya’s buoyant magma rose to shallower depths, it likely began to boil off water stored in the volcano’s underground network of fluid-filled cracks and fissures, called its hydrothermal system.

This commonly happens prior to volcanic eruptions, but scientists hadn’t thought to use DOAS to detect a possible preeruptive release of water vapor, the team writes. They suggest that visible-light DOAS stations be set up around the world to detect when other active volcanoes are getting steamier. (Journal of Geophysical Research: Solid Earth, https://doi.org/10.1002/2017JB014020, 2017)