The 1052 white vapour matches the 1052 radiocarbon candidate

Same year, a northern glow at a mid-latitude site, and a small unconfirmed solar particle event. No published paper connects them as of 24 October 2026.

The carbon-14 side after one day of checking: six annual tree-ring series cover these years. Three (English oak STK-C10, a California full-ring sequoia, the Kujawy oak in Poland) have their first high ring at 1053, exactly what a December 1052 event should produce. Three (Furness Abbey oak, two other California sequoia series, a Finnish pine) have it at 1055, which fits the 1054 supernova instead. Two of the disagreeing series are sequoias from the same mountains, so the split is probably laboratory and ring-dating offsets, not two events. The timing test is a draw until someone measures several trees in one lab.

The record side: Korea logged a turtle-shaped white vapour in July 1052 but nothing in winter. Japan has no catalogued aurora 1033 to 1071. The Annals of Ulster, the British phenomena list and the Anatolian catalogue have nothing for the winter of 1052 to 1053. Ireland has a tower of fire in April 1054, usually called a tornado. Fritz 1873 has no European aurora from 1031 to 1069. Hermann of Reichenau and the Islamic chronicles are still unread.

Still not a finding: one winter record, the weakest aurora class, and a carbon-14 step that one of two English oaks does not show.

Why this record matters more than it looks: the Song history has only one aurora candidate in the entire 1050s decade, and it is this one. The East Asian catalogue has nothing between 1033 and 1067 except Korea in July 1052. Compare the 1260s, when glows were logged most years. A lone glow in an empty stretch is a much less likely coincidence than a glow in a busy one.

Added 25 October 2026 from Miyahara et al. 2022: oaks build spring wood from the previous year's stored sugar, so oak radiocarbon can lag conifers by a year. Two of our three '1053' series are oaks, so they fit an event in 1052 but cannot pin it to December. The conifer series still split: Eastoe's sequoia rises into 1053, Terrasi's sequoia and pine into 1054 to 1055. The timing test stays a draw.

New alternative, 25 October 2026: the Goryeosa records a comet in the southern sky on about 25 February 1053 with a tail over 1 zhang. If that comet was in the northern evening sky in mid December 1052, its tail could be the Chinese 'white vapour in the north'. Worth asking Dr Hayakawa, who will know the Chinese comet records for that winter. Korea's lunar eclipse of 9 December 1052 matches the Moon phase in the Chinese record four days later, a good check that the dates line up.