Argon-argon dating of Vesuvius sanidine lands eight years from Pliny's eruption date

Pliny the Younger saw the eruption that buried Pompeii and killed his uncle, Pliny the Elder. He also wrote down enough about its timing to fix the date ^[2]. Graduate student Caroline Hasler went back through the historical record and confirmed Aug. 24 to within two months ^[10].

The method measures a ratio. Potassium-40 in volcanic rock decays to argon-40, and argon-40 is not normally present in minerals before an eruption ^[13]. The lab irradiates the crystals with neutrons to turn stable potassium-39 into argon-39, then compares the two argon isotopes. More argon-40 relative to argon-39 means an older sample ^[13]. The newer samples also came from a different part of the eruption. Magma chambers under stratovolcanoes like Vesuvius stratify, so the potassium-rich magma near the top comes out first ^[15]. In 1998 co-author Andrea Marzoli of the University of Padua collected pumice from that earliest stage at Oplontis, another Roman town buried in ash. It held more potassium than the earlier samples ^[14].

The eight-year gap between the crystals and Pliny sits inside the 13-year error bar ^[1]. The team reports a precision of 0.7 percent and an accuracy of 0.4 percent ^[9]. Those figures are 13 years out of 1,938 and 8 years out of 1,946 ^[2]. Accuracy here means closeness to the true age, and precision means how reproducible the result is ^[18]. The denominator is eight crystals from one eruption ^[7].

The known date also went into the half-life. The new value's uncertainty is about 0.73 percent of the value ^[3]. According to the phys.org report, the tighter constant will help ground-truth carbon-14 dates of organic material and uranium-lead dates of billion-year-old rocks ^[4]. The report does not say which half-life produced the 1,938-year figure. If the historical date informed both, the eight-year agreement tests the calibration's internal consistency more than it tests the method blind.

The hazard benefit is a forecast by the authors. According to the phys.org report, the gain will let geologists date past eruptions of volcanoes that threaten Mexico City, Naples and Yogyakarta more accurately ^[3]. What was tested is one eruption, sampled from its most potassium-rich first phase ^[14]. "If you want to put together the eruptive history of a volcano in relatively recent time, precision and accuracy really count," said study leader Paul Renne, director of the Berkeley Geochronology Center ^[5]. "The study shows that you can achieve that kind of highly useful precision and accuracy into the historical realm," he said ^[6].

Renne has used tight dates to argue about sequence before. A decade ago his argon-argon dates put a meteor impact, volcanic eruptions in India and the dinosaur extinction within a few tens of thousands of years of one another, 66 million years ago. Those dates supported the idea that the impact intensified the volcanism ^[17]. Tighter dates settle which event came first, while the claim that one caused another still rests on other evidence. "This lets us more precisely infer causality between events in the geologic record, for example a meteor impact structure and a mass extinction," Renne said ^[16].