Korean dim-Sun records track volcanic eruptions (Goryeo period, 918 to 1392)

Tested 25 October 2026 (script: scripts/goryeosa_dimsun.py). Parsed 8,243 dated sentences from the Goryeosa omen and astronomy volumes and found 30 entries of a Sun 'without light', red, blood-coloured or dim, in 21 different years. Compared with the 12 eruptions of over 5 million tonnes of sulfur in the eVolv2k ice-core list.

Result: only 2 of the 21 dim-Sun years fall in the year or two after a big eruption, about what chance gives (p 0.2 to 0.3). Most entries fall in the 2nd and 3rd lunar months, March and April, the season of dust storms from the Gobi. The record also misses most big eruptions: of the nine largest, it caught only 1108 (three red, lightless springs in a row, 1109 to 1111, matching the 1108 to 1110 cluster) and Samalas 1257 ('the Sun was blue, without light', then 'red like blood', in early 1258). It has nothing after 1171, 1182, 1230, 1276, 1286 or 1345, and nothing in 1285 to 1288 for our 1286 puzzle.

Verdict: in the Goryeo period these records are mostly dust and cannot find new eruptions, though they confirm the very largest.