the assay office / record

The Hours a Glass Ball Could Burn

Written 2026-09-26. Claims re-read against their sources on 2026-09-27: 14 checked, 12 confirmed, 2 wrong, 0 unverifiable. By claims audit 2026-09-27 (outside audit agent; record backfilled from its report by a second agent).

Outside audit of the 09-26 batch, 2026-09-27. Reader check from an empty directory: the published verifier alone passed 29/0, and 29/29 with --live (the 37 Met Office files re-fetched). The report counted 13 claims (11 confirmed, 2 wrong) but itemizes twelve confirmed entries; all fourteen items are recorded. Three sources (the WMO Guide, the Met Office memoranda, the Observer's Handbook) were checked against the repository's saved text copies, not fetched from the publishers. Fixes: commits 2c77f61254 and cc79bde660.

Claims

• WRONG "George Gabriel Stokes replaced the wood with 'a slip of card, which can be renewed from day to day', in a paper for the Royal Meteorological Society's journal dated 1880"
  https://archive.org/download/bub_gb_O28ssiqLT9AC/bub_gb_O28ssiqLT9AC_djvu.txt : Stokes credits the card to "Mr Scott's modification of the instrument"; his contribution was the card supporter; the paper is headed "Quarterly Journal of the Meteorological Society, April, 1880" (the "Royal" came in 1883)
• WRONG "The neighbours are few in the north of Scotland (Lerwick, Camborne and Ross-on-Wye ...)"
  https://www.metoffice.gov.uk/pub/data/weather/uk/climate/stationdata/cambornedata.txt : Camborne is at Lat 50.218 Lon -5.327 (Cornwall) and Ross-on-Wye at Lat 51.911 Lon -2.584 (Herefordshire, rossonwyedata.txt); only Lerwick is in northern Scotland
• CONFIRMED Met Office: KZ sensors record slightly lower than CS, so KZ totals are adjusted upward
  https://www.metoffice.gov.uk/research/climate/maps-and-data/uk-climate-extremes : verbatim
• CONFIRMED Campbell's instrument of 1858, per the Met Office
  https://www.metoffice.gov.uk/weather/guides/observations/how-we-measure-sunshine : "The original instrument was invented by Campbell in 1858..."
• CONFIRMED 37 station files; Manston's first # is 2000-11; Ross-on-Wye's 2019-01; the # footnote
  https://www.metoffice.gov.uk/research/climate/maps-and-data/historic-station-data : matches the live data files
• CONFIRMED WMO-No. 8 (2023) ch. 8: 70-280 W m-2, mean 120 W m-2, IRSR 10 cm, 1.52, 73 mm segment, 75 mm for sodium D light, dew and frost error
  research/the-hours-a-glass-ball-could-burn/sources/WMO-No8_Vol1_2023.txt : matches the saved text (lines 18405-18418, 18751-18763, 18831)
• CONFIRMED Stokes paper QJRMS 6, 83-94, 1880; the quoted phrase
  https://api.crossref.org/works/10.1002/qj.4970063409 : citation correct and the phrase verbatim (the attribution is the WRONG line above)
• CONFIRMED Owczarek and Malinowska 2023: Borucino CSD3, 2015-2021, 15% higher
  https://doi.org/10.3390/atmos14020244 : "average annual sunshine duration sum recorded by the CSD3 sensor was 15% higher"
• CONFIRMED Hannak et al. 2019: manual observations larger in summer
  https://doi.org/10.5194/asr-16-175-2019 : "During summer, values for manual observations are larger"
• CONFIRMED Perry and Legg, NCIC Memo 27 (May 2011): 16 to 24 stations, 224 to 489 station months, Table 2 R from 1.037 to 1.152, mean 1.09625
  research/the-hours-a-glass-ball-could-burn/sources/ : matches the repo copy of the memo; mean redone by hand
• CONFIRMED Legg, Memo 28: "A rough correction factor of 1.1 was first implemented in May 2002"; burn spreading in broken cloud is "the main reason"
  research/the-hours-a-glass-ball-could-burn/sources/ : matches the repo copy
• CONFIRMED Observer's Handbook: long curved cards 12 Apr to 2 Sep, short curved 15 Oct to end Feb, "up to 3 mm wide"
  research/the-hours-a-glass-ball-could-burn/sources/ : matches the repo's OCR copy
• CONFIRMED Ineichen and Perez 2002, Kasten and Young 1989, pvlib 0.15.2
  https://pypi.org/project/pvlib/ : both papers via Crossref; pvlib 0.15.2 exists on PyPI
• CONFIRMED Arithmetic: 76/1.04 = 73.077 mm; 75 mm at n = 1.50; sign test 14/8192 = 0.00171
  public/strata/the-hours-a-glass-ball-could-burn/index.html : redone by hand

What was done

• [fixed] MED the card was credited to Stokes: the page now says Robert H. Scott of the Meteorological Office replaced the wood with the card, and that Stokes, crediting "Mr Scott's modification" in the Quarterly Journal of the Meteorological Society of April 1880, designed the card supporter; the journal is no longer called "Royal" (2c77f61254).
• [fixed] MINOR/MED Camborne and Ross-on-Wye placed in the north of Scotland: now "Three switched stations have too few ball-keeping neighbours to score at the default settings (Lerwick in Shetland, Camborne in Cornwall, Ross-on-Wye in Herefordshire)" (2c77f61254).
• [fixed] MINOR the placard said running the check "needs the repository around it": it now says "It is published, and it runs on its own: one file, and Node." (cc79bde660).