A line of ash from Naples runs under the fields of the Lower Danube

#A 23-hour eruption next to Naples, by the model

The Campanian Ignimbrite, from the Campi Flegrei caldera west of Naples, is the most powerful known eruption in Europe in the past 200,000 years. Argon dating puts it 39,850 ± 140 years ago. Older work, which many Romanian studies follow, used 39,280 ± 110 years.

A model published in 2016 rebuilt the eruption from the measured ash: first a plinian column 44 km high for 4 hours, then, after the caldera collapsed, a 37 km cloud for another 19 hours. In this reconstruction about 208 cubic kilometres of loose volcanic material fell; spread evenly, it would cover all of Romania 87 cm deep.

#The wind took it east

Stratospheric winds pushed the cloud north-east over the Adriatic, the Balkans and the Black Sea, as far as the Russian Plain. The pale field on the map is the 2016 model: the thickness the ash would have left, not a place where it was measured.

The model puts 10–20 cm of ash over parts of North Macedonia, Bulgaria and Romania, and at least half a centimetre over about 3 million square kilometres.

#Eight places where it was found and dated

To count as a find, a place needed three things: volcanic glass analysed chemically and matched to Naples, an exact location, and ages measured at the same site, above or below the ash. Eight places in Romania and Bulgaria meet all three: four loess sections in Romania (Urluia, Rasova, Căciulătești, Balta Albă) and four caves in northern Bulgaria (Magura, Kozarnika, Toplitsa, Temnata).

All eight lie on the same bearing from the crater, between 61 and 69 degrees (east-north-east), 765–1,192 km away. Another 4 Romanian places are reported with the same ash but miss one thing: a date at the spot or a chemical analysis. Tap a point.

#The wall at Urluia

At Urluia, in Constanța county in Dobrogea, the strip of Romania between the Danube and the Black Sea, about 15 km south of the Danube, a quarry cuts through more than 20 m of loess, the dust blown in during the ice ages. On its wall the ash is a white bed lying on an old slope and thickening into a hollow: from nothing to 100 cm.

Geologists cleaned the wall at two points. Towards the north-west the bed is about 50 cm, with another 25 cm of ash mixed with loess above it; towards the south-east it is about 15 cm and breaks into rusty blocks. Luminescence ages below the bed (38.7, 41.1 and 48.8 thousand years) and above it (36.2 and 23.9 thousand years) bracket it.

#The plants under the ash

At the base of the bed, where the ash rests directly on the loess, imprints of vegetation survive: the plants it fell on. The authors of the Urluia study write that the ash blanketed the land “both thickly and rapidly”.

Above the ash and its mixed zone lies first a carbonate-rich loess, which the authors link to the cold Heinrich 4 interval. Then, in a slightly milder spell, a weak soil about a metre thick formed. Above it the wind brought loess again, up to today’s black soil. The field on top of the drawn block is the one ploughed now.

#The test that failed: thickness

Before this page was built, a test was set down in writing: the ash should thin with distance from the crater. Across 20 places with a published thickness, from Italy to the Don, the expected decrease does not appear: the Spearman correlation comes out between −0.13 and +0.03 (where a paper gives only an upper limit, both half and the full value were tried), against a threshold of −0.5. Without Drăgănești-Olt, whose “under 60 cm” could not be checked, it is between −0.21 and −0.04. In this sample, thickness does not fall with distance.

The departure comes mostly from Romania. At Urluia (about 50 cm at the sampled profile) and Căciulătești (50–60 cm) the bed is thicker than in lakes and caves much closer to Naples. The Urluia study called the bed there “up to twenty times thicker” than the 2012 model predicted; the 2016 model gives 5–20 cm here.

The explanations are not exclusive: the models may underestimate the ash that fell far away, and on the ground the thickness also depends on where it came to rest. At Urluia it runs from 0 to 100 cm along one wall, although the authors read it mostly as a rapid deposit with little reworking; at Balta Albă it is 4 cm and at Săgeata, 32 km away, more than 50; near Kostenki on the Don, two places at the same distance from the crater hold 1.4 and 15 cm.

#A line that marks the same day

The ash comes from a single eruption that lasted about a day. Wherever they find it, geologists and archaeologists know its layer has the same age, however different the ground around it. That lets them line up layers from places that have nothing else in common.

In the Bulgarian caves the level with volcanic glass, visible or only under the microscope, lies over layers with early Upper Palaeolithic tools. At Toplitsa there is a hearth beneath it and, lower down, a layer dated to 49,770–45,050 calibrated years before 1950; at Kozarnika the layer below has, for example, 37,000 ± 800 uncalibrated radiocarbon years; at Magura the layer above has 32,750 ± 500 uncalibrated radiocarbon years. On that same clock the eruption is 34,290 ± 90 years old.

#Who lived here

155 km from Magura, in Banat, south-western Romania, Peștera cu Oase (the Cave with Bones) kept the jaw of a modern human who lived 37,000–42,000 years ago. His genome is 6–9% Neanderthal, from an ancestor four to six generations back. Neanderthals disappeared from Europe around 39,000–41,000 years ago.

The Oase genome shows that modern humans and Neanderthals had interbred in this period, and the man’s age range overlaps the eruption. The studies that traced the ash through caves do not find that the eruption was the main cause of Neanderthal extinction in northern and eastern Europe: modern humans had spread across Europe before it, and in several caves people went on living there after it fell.