Forecasting eruptions using pre-eruptive seismic patterns at Sinabung Volcano, Indonesia

Abstract:

Forecasting the size, timing and style of volcanic eruptions is of primary interest to observatories and civil authorities world-wide, yet most observatories only have access to long-term data at a very limited number of volcanoes under their jurisdiction. When extensive long-term data sets are available to responsible agencies, volcanic eruptive size, timing and style can usually be successfully forecast using current monitoring data and knowledge of precursory eruptive patterns, enabling the communication of timely forecasts to civil authorities. Experienced agencies, such as Indonesia’s Center for Volcanology and Geologic Hazards Mitigation and the USAID-USGS Volcano Disaster Assistance Program, utilize extensive collective experiences with multiple monitoring streams over multiple eruption cycles and across volcano types to successfully forecast eruption size, style and onset, as well as changes in eruptive style and size within ongoing eruptions.

The longest-term real-time monitoring parameter commonly available at volcanoes worldwide is seismic data. Seismic data is a direct measure of rate-dependent strain changes in the magmatic system from the deep magmatic input to shallow eruptive processes. Patterns of pre-eruptive earthquakes coupled with other available monitoring data and conceptual models of magma ascent enable short-term forecasting of eruption size, style, and onset. First order event locations, characterization of background seismicity, and changes in earthquake types and energy release are most important to successful eruption forecasting. This study demonstrates how this approach has been used to successfully forecast eruption onsets, changes in eruptive style, and to change alert levels and extend or contract evacuation zones during the ongoing eruption of Sinabung Volcano, Indonesia.