A Note in the Margin: Roros, February 1947
~ Friday, February 14, 2020 ~
Earlier this year, while requesting digitised historical weather data from the Norwegian Meteorological Institute for a separate project on atmospheric baseline conditions, I received a batch of scanned observation logs from the Røros station. Røros is a mining town approximately fifteen kilometres northwest of the Hessdalen valley. The station has been operating since the nineteenth century. I was looking for temperature and precipitation records. What I found, in a log entry dated February 14, 1947, was something else.
The entry is otherwise unremarkable. Clear sky. Temperature: minus seventeen degrees Celsius. Wind from the northeast. The standard measurements logged at their standard intervals in the hand of a meteorologist who had presumably been doing this for years. At the bottom of the entry, written in a slightly different pen, a marginal note in Norwegian:
Uvanlig lys observert i sørøstlig retning, ca. 22:15. Flimrende. Ikke nordlys.
“Unusual light observed in south-easterly direction, approximately 22:15. Flickering. Not aurora.”
What the exclusion means
I want to draw attention to the last two words: ikke nordlys. Not aurora.
This is a technical exclusion, not a casual remark. A meteorologist stationed in Trøndelag in the 1940s observed aurora regularly. It was part of the normal sky. The observer was ruling out the most obvious natural explanation for an unusual light at high latitude on a clear winter night — not because he had never seen aurora, but because what he saw was different from aurora and he knew it.
He used the correct vocabulary. He was precise about direction (south-easterly), time (approximately 22:15), and character (flickering). He did not file a separate report. He noted the observation in the margin of a routine weather log and moved on.
South-easterly from Røros is the direction of the Hessdalen valley. The note predates the accepted first modern reports of the Hessdalen lights by thirty-four years.
What this does and does not establish
I am not claiming this note proves that the Hessdalen lights were active in 1947. A single marginal observation, without corroborating data, is a data point, not evidence of a pattern.
What it establishes is narrower: that a trained observer at a station with clear sightlines toward the Hessdalen valley noted an unusual, non-aurora light in that direction on a specific date in February 1947, and that this observation exists in the archive of the Norwegian Meteorological Institute.
The lights were first widely reported in 1981. Project Hessdalen has noted, without documentation, that earlier reports exist from the 1930s. The 1947 note is not proof of continuity. It is a single prior observation, logged by someone who knew the difference between what he was seeing and the things he normally saw.
What has not been checked
I have not had time to examine the adjacent station records. The Norwegian Meteorological Institute archive holds manual observation logs from multiple Trøndelag stations going back into the nineteenth century. If a trained observer at Røros noted an unusual light in the south-easterly direction on a clear night in February 1947, there is no particular reason to assume he was the only observer in the region who saw anything that night — or that February 1947 was the only occasion.
Checking those records would be a reasonable archive project. It is not one I am currently undertaking. I am noting the find here because it exists, and because noting finds is part of what this kind of work requires.
The log is held by the Meteorologisk institutt archive in Oslo. The Røros station identifier in their historical database is identifiable from the digitisation project records.
In this series
- Hessdalen: What the Valley Actually Looks Like
- Project Hessdalen: What the Instruments Recorded
- What the Plasma Hypothesis Explains, and What It Doesn’t
- A Note in the Margin: Røros, February 1947
- Earlier Sky Reports from the Scandinavian Record