Alert track record

Every aurora alert email we have a record of, and whether the night turned out the way the forecast said. Misses stay on the list.

Nights judged so far

Activity reached the level

5 of 14

Reached it, and the sky was not overcast

3 of 14

Each alert is sent when NOAA's forecast says Kp will reach the level a place needs. Kp is a 0 to 9 measure of how stirred up Earth's magnetic field is: the higher it goes, the further from the poles the aurora reaches. After the night we check the Kp that was actually measured over the hours the email was about. If it reached the level, the forecast came true. A measured Kp just under the level, like 3.7 against 4, counts as short.

Cloud is judged separately. Above 80% average cloud, a good night can show nothing at all. Neither check means anyone saw the aurora: it is a record of conditions, not sightings.

Homer · 25 Oct
Forecast not recorded (sent before 26 Oct)
Night not over yet
Jasper · 25 Oct
Forecast not recorded (sent before 26 Oct)
Night not over yet
Fort William · 25 Oct
Forecast not recorded (sent before 26 Oct)
Fell short: measured Kp 1.3, needed 461% cloud
Dundee · 25 Oct
Forecast not recorded (sent before 26 Oct)
Fell short: measured Kp 1.3, needed 438% cloud
Egilsstaðir · 23 Oct
Forecast not recorded (sent before 26 Oct)
Fell short: measured Kp 2, needed 487% cloud, too cloudy to see
Banff · 23 Oct
Forecast not recorded (sent before 26 Oct)
Reached it: measured Kp 2, needed 249% cloud
Sudbury · 21 Oct
Forecast not recorded (sent before 26 Oct)
Reached it: measured Kp 4.7, needed 432% cloud
Vík · 21 Oct
Forecast not recorded (sent before 26 Oct)
Reached it: measured Kp 4.7, needed 430% cloud
Londonderry · 19 Oct
Forecast not recorded (sent before 26 Oct)
Fell short: measured Kp 1.3, needed 523% cloud
Rochester · 19 Oct
Forecast not recorded (sent before 26 Oct)
Fell short: measured Kp 2.3, needed 537% cloud
Storm alert, any place · 19 Oct · 3 emails
Forecast not recorded (sent before 26 Oct)
Fell short: measured Kp 2.3, needed 5
Copenhagen · 19 Oct
Forecast not recorded (sent before 26 Oct)
Fell short: measured Kp 1.3, needed 527% cloud
Thurso · 16 Oct
Forecast not recorded (sent before 26 Oct)
Fell short: measured Kp 2, needed 378% cloud
Edinburgh · 11 Oct · 2 emails
Forecast not recorded (sent before 26 Oct)
Fell short: measured Kp 2, needed 538% cloud
Dublin · 10 Oct
Forecast not recorded (sent before 26 Oct)
Reached it: measured Kp 7, needed 698% cloud, too cloudy to see
Storm alert, any place · 10 Oct
Forecast not recorded (sent before 26 Oct)
Reached it: measured Kp 7.3, needed 5

Before 26 Oct 2026 we kept only the latest alert sent to each person, so some earlier emails are missing from this list, and for those we have no record of what the forecast said. Earlier alerts are judged against each place's level as the model sets it today. From 26 Oct every email is kept with the forecast it quoted.

Forecast from NOAA SWPC. Measured Kp from GFZ Potsdam (CC BY 4.0), recent values provisional. Cloud from Open-Meteo.