Forecasting steam-burst eruptions

When fluid in a shallow aquifer near a volcano becomes heated and pressurized, the result can be an explosion of steam, water, ash, and rock called a phreatic eruption. Such eruptions, which are difficult to predict, are extremely hazardous: Over the past century in Japan, 10 eruptions each have killed at least 10 people, including the 2014 eruption of Mount Ontake, which claimed the lives of 58 hikers. But the eruptions’ unpredictability might soon change. By analyzing satellite ground deformation data, Tomokazu Kobayashi, Yu Morishita, and Hiroshi Munekane at the Geospatial Information Authority of Japan have flagged a precursory signal for a recent phreatic eruption.

One difficulty in monitoring for phreatic eruptions is that potential precursors, including deformation of Earth’s crust and seismicity, are generally too small to detect and cannot help in inferring the location of a pending event. To counter that problem, Kobayashi and colleagues developed an algorithm that tracks temporal changes in high-spatial-resolution ground position by using interferometric synthetic aperture radar (InSAR) data (see Physics Today, August 2015, page 76 ) from Japan’s Advanced Land Observing Satellite-2. The researchers applied their tool to crustal deformation data at Japan’s Hakone volcano, which is shown in the photo, before a small phreatic eruption in 2015.

Kobayashi’s team found that in the six months before the eruption, the surface rose at 5 mm/month over a 200 m radius. The rate of ground inflation accelerated two months before the eruption and was accompanied by swarms of several hundred earthquakes per day. The synchronization between seismicity and ground deformation indicates that heat from deep within Earth drove shallow hydrothermal activity, which in turn might be linked to magma movements.

The authors are applying their analysis technique to InSAR satellite data from areas where historical phreatic eruptions have taken place. But data have not been readily available in the years preceding more recent phreatic eruptions, and a forecasting system will require scientists to have access to a more steady stream of images. The work suggests that satellite-based techniques could greatly improve scientists’ ability to predict dangerous phreatic events. (T. Kobayashi, Y. Morishita, H. Munekane, Earth Planet. Sci. Lett. 491, 244, 2018 .)