A Pyroclastic Current That Swallowed Air, and Grew Taller the Farther It Ran

Pick up a fist-sized stone from the ash country east of Naples, and you are holding a thermometer. It is not a precise one, and it stopped working around 40,000 years ago. But the magnetic minerals locked inside it recorded the temperature at which the stone came to rest, and they have been holding that reading ever since.

The stone arrived by air. About 40,000 years ago, the Campanian Ignimbrite eruption emptied a magma system beneath the Campi Flegrei caldera, west of Naples, and laid a dilute, ash-choked current across the landscape. It ran roughly 75 kilometers. It did not stop at the Apennines; it went over them, and the awkward question ever since has been how anything at the end of a 75-kilometer run could still have the height and the energy to climb a mountain range.

A five-author team has now put a number on the part of the answer nobody could quantify: how much ordinary air the current had to swallow along the way. Michael H. Ort of Northern Arizona University, Guido Giordano and Aurora Silleni of Roma Tre University, Elena Zanella of the University of Turin, and Roberto Isaia of Italy's National Institute of Geophysics and Volcanology read the temperature history out of stones the current picked up and carried, then worked backward through a heat budget. Their paper went online on July 30 in the Bulletin of Volcanology.

The thermometer works like this. Heat a rock past the blocking temperature of the magnetic minerals inside it, and their existing magnetization is wiped; as the rock cools, it locks in the direction of whatever magnetic field surrounds it at that moment. A lithic clast swept up by a pyroclastic current and dropped in the deposit, therefore, carries a record of how hot it was when it stopped, provided you sample it in place and note which way it was pointing. The team did that from proximal sites near the caldera out to distal ones, and found single magnetization components up to more than 580 degrees Celsius, matching the magnetization of the ignimbrite around them. That is the floor of the emplacement temperature rather than the value: above 580 C, everywhere they looked.

There is a ceiling too, and it comes from the deposit's texture rather than its magnetism. Ash welds when it is hot enough for the glass shards in it to stick and deform. In the Campanian deposit, welding is incipient close to the vent and absent elsewhere, which puts emplacement at or below the glass transition of about 690 C throughout.

Now the arithmetic that matters. The magma feeding the eruption was around 900 C, a value the authors adopt from two-feldspar thermometry rather than determine themselves. So a current that began near 900 C was somewhere between 580 and 690 C by the time it settled, and something had to absorb the difference. Expanding magmatic gas carries off only a small share of it. Cold rock torn from the ground and swallowed water are the other two candidates, and the field evidence, the authors say, does not support either in the quantities required. That leaves the atmosphere. To cool the system as much as the stones say it cooled, the current had to take in nine to 40 times more air, by mass, than its own magmatic gas supplied.

Nine to 40 is a factor-of-four range, and it should be read as a requirement rather than an event. This is what a heat budget demands once the other ways of losing heat are ruled out, not a measurement of air taken at the time. The shape of the argument is the interesting part. The number falls out of an elimination, and it holds exactly as firmly as the exclusion of rock and water does.

Swallowed air has a second consequence besides cooling. It adds volume, and a current with more volume in it gets taller. Instead of thinning as it spread, the reconstruction has the current thickening with distance, from about 500 meters near the caldera to more than 800 meters at its first encounter with the Apennine Mountains, 40 kilometers from the vent. That 800-meter figure is a place, not a maximum. It describes one point at 40 kilometers on a run of about 75, and earlier work on the same currents by several of the same authors, published in the GSA Bulletin in 2024, argues from ground-layer deposits stranded high on ridges that the current was at least 1.5 kilometers thick. The two readings are probably compatible on a flow that kept growing past the 40-kilometer mark.

Growing taller did not mean speeding up. The paper is explicit that the current decelerated as it thickened, and it places a change of flow regime inside that slowdown. Volcanologists describe currents like this with a Richardson number, which sets the stability of a density-stratified flow against the shear trying to stir it. Below 1, the flow is supercritical: thin, fast, mixing hard along its upper surface. Above 1, it is subcritical, thicker and slower and less easily stirred. The Campanian current crossed from one regime to the other somewhere between 25 and 37 kilometers from the vent, four to six minutes after the eruption began.

Air ingestion, the authors add, probably became less efficient once that transition happened. The two statements are easy to collapse into a contradiction, and they are not one. Entrainment slowing after 25 kilometers is a claim about rate. The nine-to-40 figure is the integrated total over the whole run.

On the mountain-climbing question, what the reconstruction offers is an explanation rather than a demonstration. The data and model, the authors write, "provide explanations for how a large dilute pyroclastic density current can erupt and flow long distances while maintaining enough energy and thickness to overtop topographic obstacles." A current that gets taller as it goes has more of itself above any given ridge line and less of its mass sitting where the ridge can strip it away. Whether that is the whole reason this one cleared the Apennines is a question the paper frames rather than closes.

Three quantitative claims here are new and so far untested by anyone outside the paper: the air requirement, the thickening profile, and the located regime change. For work one day old, that is expected rather than a gap. The method underneath them is not new at all. Emplacement temperatures have been read from thermal remanence at Taupo, Santorini, Fogo, and Colli Albani, which is why the 580 C floor is the least contestable thing in the study. What the paper adds is a number where the field had a description. Anyone modeling a dilute current of this size now has a range to hit, and a way to fail: entrain much less than nine times the magmatic gas mass, and the deposit comes out hotter than the stones say it was.