Aurora Forecast Methodology

How Aurora Norway combines NOAA, Kp, darkness and Norwegian cloud forecasts to estimate where the northern lights are most visible.

Forecast Inputs

Our forecast combines NOAA's OVATION aurora model, the planetary Kp index, solar-elevation darkness calculated per location, and cloud forecasts from MET Norway (Yr).

OVATION gives the probability of aurora being active directly overhead at a point. That is not the same as your probability of seeing one. Aurora sits 100–300 km up, so a display centred well north of you is still visible on your northern horizon — often from several hundred kilometres outside the area OVATION colours in.

How We Use Kp — And Why It Matters Less Than You Think

Kp measures how far the auroral oval has been pushed towards the equator. It is a good indicator in Oslo or Bergen, where aurora only appears when the oval stretches unusually far south.

Inside the oval — Troms, Finnmark, Svalbard — it is a poor one. Those places sit under the aurora on an ordinary night, and Kp 1 or 2 regularly produces a strong display there. Very high Kp can even work against you: it drags the brightest band south, so a Kp 7 sky over Alta can look emptier than a Kp 3 sky did.

The comparison list uses current OVATION activity weighted by evening cloud cover, with clearer skies first when scores tie. This is a relative sorting score, not a measured sighting probability. Kp also informs the plain-language outlook when local model activity is low or unavailable.

Update Cadence

The open page refreshes every five minutes. NOAA responses are cached for up to five minutes and MET Norway weather for up to fifteen minutes. The selected place loads first; comparisons cover the 24 northernmost places and Oslo. Selecting another place loads its weather on demand. The displayed timestamps belong to the original data sources, not the page refresh.

Cloud Cover, Darkness and Activity

Percentages on the map show forecast cloud cover. The evening window runs from 18:00 to 06:00 in Norwegian time, with the date carried across midnight. Dark viewing hours use an approximate solar elevation below −12 degrees. The suggested window is the longest continuous stretch of dark hours with less than 40% forecast cloud cover.

OVATION describes short-term aurora activity overhead, not a forecast for the entire coming night. The tomorrow view therefore uses cloud and darkness forecasts without copying current activity into tomorrow. Conditions can change between checking the site and going out.

Missing or stale data remains unavailable: cloud forecasts older than eight hours, Kp observations older than six hours and OVATION forecasts older than ninety minutes are not treated as current readings. Places without usable ranking data are not assigned invented scores.

Limitations

Worth stating plainly, because these bound what any aurora forecast can do:

  • Aurora activity is not usefully predictable more than a few days ahead. Anything presenting odds for a date months away — including our own trip planner — is modelling typical conditions, not forecasting your night.
  • Cloud forecasts are point forecasts. Coastal Norway has sharp microclimates; a fjord ten kilometres away can be clear while the town is not. A local guide who can drive to a gap will beat any model on this.
  • Substorms are the reason to stay out. Aurora can go from nothing to overhead in minutes and fade just as fast, and no public model predicts the timing. A quiet sky at 21:00 says little about 23:00.
  • Our trip-planner figure assumes typical local weather and average solar activity, and discounts consecutive nights because weather systems and solar activity both persist for days. It is for comparing trip lengths, not for setting expectations about a specific night.

Editorial Safeguards

Automated research may assist topic discovery, but the publishing workflow checks source presence, structure and internal links before release. These automated checks cannot verify every real-world claim or replace a named human review.