In August 79, Pliny the Younger was seventeen. Years later he wrote two letters about the death of his uncle, Pliny the Elder, killed by the eruption of Vesuvius. In one of them the eruption begins on the ninth day before the Kalends of September: 24 August.
That date has become a laboratory tool. Paul Renne, of the Berkeley Geochronology Center and the University of California, Berkeley, and six colleagues, among them Andrea Marzoli of the Università di Padova (University of Padua), used it to calibrate argon-argon dating, one of the two great clocks geologists use to put an age on rocks. The work was published in Science Advances on 25 September 2026.

The method relies on potassium-40, a radioactive isotope of potassium that slowly produces argon-40, a gas. While magma is molten the argon escapes. Once lava cools and crystals form, the argon produced from then on stays locked inside. In the laboratory the samples are irradiated with neutrons, which turn part of the stable potassium into argon-39. The ratio between the two argon isotopes gives the age: the more argon-40, the older the crystal.
The samples come from Oplontis, the Roman settlement buried along with Pompeii. Marzoli collected them in 1998. They are pumice from the first phase of the eruption, richer in potassium than the material the same group used in a 1997 study, and they sat in storage for almost thirty years. Eight samples of sanidine, a potassium-rich feldspar, gave an age of 1,938 years before the 2025 measurement, with a standard deviation of 13 years. Counted from 24 August 79, the true age is 1,946 years. The measurement falls eight years short of the historical value, inside its own margin. Precision is 0.7 per cent; the deviation from the expected value is 0.4 per cent.

Pliny’s date is not undisputed. Some historians move the eruption to the autumn, partly on the strength of a coin found at Pompeii that, in their view, could not have been minted before September. Caroline Hasler, a graduate student in Renne’s group, compared that coin with other issues of the same period and judged it minted earlier. In the paper the authors conclude there is no solid reason to doubt the accounts of Pliny and Cassius Dio. As a precaution they still allowed the date a two-month margin.
The second result concerns potassium-40 itself. The group derives a half-life of 12.044 billion years, plus or minus 0.088, for its decay to argon-40, and calls it twice as precise as the value nuclear physics has given so far. The figure needs careful reading. Potassium-40 decays along two paths, mostly to calcium and to a lesser extent to argon, and its overall half-life is about 1.25 billion years. The 12 billion refers only to the branch that leads to argon, the one the clock depends on. Here the two-month margin on the historical date matters: it barely affects the calibration of the method, but it does affect the half-life.

What is an eight-year gap over two thousand worth? Renne’s answer is that greater precision makes it possible to establish more firmly the order of events in geological history, for instance between a meteorite impact and a mass extinction. About a decade ago his group used the same method to date the end-Cretaceous impact, the great eruptions in India and the disappearance of the non-avian dinosaurs: three events packed into a few tens of thousands of years, around 66 million years ago.
There is a closer consequence too. The university press release lists Naples, together with Mexico City and Yogyakarta, among the areas where finer dating of recent eruptions helps reconstruct the history of volcanoes that are still active. The authors now speak of decadal accuracy for eruptions of historical age.
The limitation is stated. The calibration rests on a single eruption and on a date that comes from a text, not an instrument. The work feeds into a Bayesian statistical scheme, published by the same group in 2025, that tries to reconcile argon-argon with uranium-lead dating, the other great clock of the rocks. The two methods have not always agreed, and Vesuvius now enters that calculation with more weight than before.
Pliny wrote to tell how a man had died. He left a date that now measures deep time.

References
Renne P.R., Cassata W.S., Hasler C.E.J., Carter J.N., Fuentes A.J., Tholt A.J., Marzoli A., «Pliny the Younger advances geochronology», Science Advances, 12, 39, eaeg2624, 25 September 2026. doi: 10.1126/sciadv.aeg2624
University of California, Berkeley, «The eruption that destroyed Pompeii is helping scientists date Earth’s history», Berkeley News, 25 September 2026 (press release; republished by ScienceDaily on 30 September 2026).