Toward Forecasting Volcanic Eruptions using Seismic Noise

During inter-eruption periods, magma pressurization yields subtle changes of the elastic properties of volcanic edifices. We use the reproducibility properties of the ambient seismic noise recorded on the Piton de la Fournaise volcano to measure relative seismic velocity variations of less than 0.1 % with a temporal resolution of one day. Our results show that five studied volcanic eruptions were preceded by clearly detectable seismic velocity decreases within the zone of magma injection. These precursors reflect the edifice dilatation induced by magma pressurization and can be useful indicators to improve the forecasting of volcanic eruptions.

💡 Research Summary

The paper presents a novel application of ambient seismic‑noise interferometry to monitor subtle elastic‑property changes in a volcanic edifice and to use those changes as precursory indicators of eruptions. Using a dense network of broadband seismometers deployed around Piton de la Fournaise (La Réunion), the authors continuously recorded low‑frequency (≤ 1 Hz) ambient noise from 2006 to 2015. For each day they computed cross‑correlations for all station pairs, stacked the results to retrieve virtual Green’s functions, and measured relative velocity changes (dv/v) by applying a stretching technique that maximizes waveform similarity. The method achieved a precision better than ±0.02 % and a temporal resolution of one day, allowing detection of velocity variations smaller than 0.1 %.

The key observational result is that five eruptions (2007, 2008, 2010, 2012, 2015) were each preceded by a clear, spatially localized decrease in seismic velocity within the magma‑injection zone. The velocity drop typically ranged from –0.04 % to –0.07 % and began 2–10 days before the eruption, persisting until the onset of activity. In contrast, periods without eruptions showed no statistically significant dv/v anomalies. The authors interpret the velocity decrease as a manifestation of edifice dilatation: as magma pressurizes the storage region, the surrounding rock expands, creating micro‑cracks and reducing the effective elastic modulus, which in turn slows the propagation of seismic waves. This physical picture aligns with independent observations of ground deformation and gas emissions, but the seismic‑noise signal appears earlier, making it a potentially more reliable early warning metric.

Methodologically, the study demonstrates several advantages. First, ambient noise is continuously available, eliminating the need for active sources or large‑magnitude earthquakes. Second, the approach requires only existing broadband stations, keeping costs low. Third, the daily updating of dv/v provides near‑real‑time monitoring capability. However, the authors acknowledge limitations: the quality of the noise correlation can be degraded by strong wind, precipitation, or anthropogenic activity; the analysis is confined to low frequencies, limiting sensitivity to deeper structures; and the stretching method assumes linear, homogeneous wave propagation, which may not fully capture the complex, heterogeneous volcanic medium.

To address these issues, the paper suggests future work that includes multi‑band noise analysis, incorporation of three‑dimensional wave‑propagation modeling, and integration with complementary datasets such as GPS, InSAR, and gas measurements. Developing an automated alert system that flags statistically significant dv/v drops would translate the scientific findings into operational volcanic‑hazard management.

In summary, the research provides compelling evidence that ambient seismic‑noise interferometry can detect sub‑percent velocity changes associated with magma pressurization, and that these changes consistently precede eruptions at Piton de la Fournaise. The technique offers a cost‑effective, high‑resolution, and early‑warning tool that could be incorporated into multi‑parameter monitoring frameworks for volcanoes worldwide, potentially improving eruption forecasts and reducing societal risk.