Plate Nº 28 · recorded November 16, 2026

Earth & ClimateReported finding

Lab Tests Overestimate Volcanic Tsunamis, Stromboli Study Finds

Bougouin and colleagues matched lab waves against three Stromboli tsunamis from 2019–2021 and found most models overpredict real wave heights.

By Marcus Bennett3 min read598 words

In brief

  1. The study compared lab experiments with three Stromboli tsunamis recorded between 2019 and 2021.
  2. Most existing lab-based models overestimate real-world volcanic tsunami size.
  3. The team used a 7-meter-long, 20-centimeter-wide water channel and polystyrene beads to simulate pyroclastic flows.
  4. New experiments reproduced the three-wave pattern observed at Stromboli.
  5. Published in Journal of Geophysical Research: Oceans (2026), DOI: 10.1029/2026jc024008.
The experiments improving our understanding of volcanic tsunamis
Plate Nº 28The experiments improving our understanding of volcanic tsunamis — Alain Bonnardeaux / unsplash

Most laboratory models overestimate the size of real volcanic tsunamis. That is the central finding of a study published in the Journal of Geophysical Research: Oceans (2026), which compared tank experiments directly against three tsunamis generated by eruptions at Stromboli Island, Italy, between 2019 and 2021.

For practitioners who rely on scaled flume experiments to forecast wave hazards, the result matters. A. Bougouin and colleagues showed that lab results cannot simply be scaled up to predict field events — yet the same paper demonstrates how to close part of that gap.

Why volcanic tsunamis resist prediction

Volcanic tsunamis arise when erupting rock and ash plunge into the sea, but the physics of that interaction — how a granular flow transfers energy to water — remains poorly understood. That gap makes it hard to answer two basic questions before an eruption: will the flow trigger a tsunami, and if so, how large will it be?

Laboratory experiments have long served as the main tool for probing the process. Until now, nobody had tested systematically how well those experiments mirror reality. The Bougouin team delivered the first direct comparison between lab data and field observations.

What made Stromboli the natural test site?

Stromboli erupts nearly constantly. Volcanic flows regularly enter the sea and produce tsunamis, and a unique monitoring network — land-based cameras paired with offshore wave gauges — records each event. In effect, the island functions as a field laboratory.

The researchers selected three eruptions from 2019 to 2021 that triggered tsunamis, then set observations from those events against results from existing experiments and from new tests they designed themselves.

How the experiments worked

The setup was deliberately simple:

  • A channel 7 meters long and 20 centimeters wide, filled with water.
  • A slope built to replicate the shape of Stromboli's flank.
  • A mass of polystyrene beads — the small spheres familiar from beanbag filling and packaging — released down the slope to tumble into the water.

The beads stood in for Stromboli's pyroclastic flows, the mixtures of hot gas, ash and rock that sweep down a volcano's sides. The researchers chose buoyant material because they expected these flows to be predominantly less dense than water. Counterintuitively, that lightness does not prevent danger: buoyant flows still generate substantial waves.

For anyone building similar models, the material choice signals a working assumption about flow density that future studies may need to refine.

What did the comparison show?

Two findings stand out. First, the lab cannot be scaled straightforwardly to the field: most models built on earlier experiments overpredict tsunami size at real volcanoes. Second, the new experiments narrowed that gap. They reproduced key features seen at Stromboli, including a three-wave pattern in the tsunami signal, and broadly captured how energy moves from the flow into the resulting wave train.

Limitations remain, and the authors say so plainly. The match is approximate, not exact, and the work covers one volcano over two years of activity. The findings mark an initial step rather than a finished forecasting framework.

What changes for hazard assessment?

The study gives hazard modelers a measured, field-validated benchmark: lab setups that reproduce the three-wave pattern and realistic energy transfer are the ones worth scaling. It also warns against trusting wave-height estimates from older experiments without adjustment, since those estimates likely run high.

The full paper — A. Bougouin et al., "Forecasting of Volcanic-Flow Tsunamis Through Integrated Data From Laboratory Experiments and the 2019–2021 Stromboli Events, Italy," Journal of Geophysical Research: Oceans (2026), DOI: 10.1029/2026jc024008 — sets the baseline every follow-up study will now measure against.

via phys.org/rss-feed/ (Source)

Filed under

  • volcanic-tsunamis
  • stromboli
  • pyroclastic-flows
  • laboratory-modeling
  • hazard-forecasting
Share this article:

More from Marcus Bennett

Marcus Bennett

Show full bio

News editor covering regulation, policy and their impact on general science news at SciBeat.

436 articles

Nearby plates

« Previous article