Stromboli Lab Tests Show Volcanic Tsunami Models Overshoot
Laboratory experiments built to copy Stromboli's flank overestimate the size of the tsunamis the volcano actually makes, according to the first study to compare lab results directly against field observations. The new experiments, run in a 7-meter water channel, reproduced Stromboli's three-wave pattern and broadly matched how a volcanic flow hands its energy to the sea.
A tank of water and a slope of beads can't be scaled straight up to a volcano, and now that's measured rather than assumed. The correction matters because hazard maps are built from these models.
Tsunamis made by volcanoes don't start with an earthquake, and that makes them hard to forecast. When a flow of hot ash and rock pours off a volcano's flank and into the sea, the waves it raises depend on how the flow and the water trade energy in the first seconds. A study published in the Journal of Geophysical Research: Oceans has now compared the laboratory version of that process against the real thing for the first time.
The test case is Stromboli, the Italian island volcano whose eruptions regularly send pyroclastic flows down its Sciara del Fuoco slope and into the Tyrrhenian Sea. Bougouin and colleagues worked from three Stromboli eruptions that triggered tsunamis between 2019 and 2021, including the July 3, 2019 event, and set them against both existing laboratory work and new experiments of their own.
The lab setup is deliberately simple. The team used a channel 7 meters long and 20 centimeters wide, filled it with water, and released a mass of polystyrene beads down a slope shaped to replicate Stromboli's flank. The beads stand in for the volcanic flow, which at Stromboli is expected to be mostly less dense than water, and they tumble into the channel the way the real flow enters the sea.
The headline finding is a caution. Lab experiments can't be scaled directly up to predict real-world events, and most models based on those experiments overestimate the size of tsunamis in the field. A hazard assessment built straight from tank results would plan for bigger waves than Stromboli actually produced in the three events studied.
The new experiments still earned their place. They reproduced key features seen at Stromboli, including the characteristic three-wave pattern the real eruptions generated, and they broadly captured the energy transfer from the flow into the resulting wave train. The gap between tank and volcano narrowed, even though the researchers are clear that some limitations remain.
That distinction is the practical value of the work. Volcanic tsunamis are rare, fast and local, and the coasts at risk, from the Aeolian Islands to Calabria and Sicily, get minutes of warning rather than hours. Models are what hazard planners have, and this study tells them which way their models lean: toward overshoot, by a margin that's now documented against field data instead of guessed.
The paper, "Forecasting of Volcanic-Flow Tsunamis Through Integrated Data From Laboratory Experiments and the 2019-2021 Stromboli Events, Italy," is described by its authors as an initial step toward improving volcanic tsunami hazard assessment. The next step is the unglamorous one: more paired comparisons, so the correction between the tank and the sea becomes a number planners can use rather than a warning they have to remember.
Stromboli is the right volcano for the question because it repeats the experiment on its own. The island is in near-continuous eruption, its flows reach the sea from the same unstable flank, and the 2019 to 2021 events gave researchers several measured cases of the same process instead of a single anomaly. That repetition is what makes a lab-to-field comparison possible at all: one eruption could be an outlier, while three start to describe how the volcano actually behaves when its flank meets the water.
Sources
The conversation.
Loading comments…

A little more perspective.
Reading is free. Verified members can comment and reply for $2 a month.