Photography

Aurora photography in Alaska is less about settings than about equipment surviving the night. Every item below has ended somebody's evening, and only one of them is expensive to solve.

The condensation one is the failure people do not see coming, because it happens after the photography is over. A camera that has been at minus 30 for three hours will fog inside the lens the moment it meets a heated van, and there is nothing to be done once it has. A freezer bag costs nothing and prevents it entirely.

Three of these carry more than 500 reviews each, which for a category of five tours is remarkable and says something about how Fairbanks competes: not on price, but on what you leave with.

Two structural things are worth noticing. First, every one of the five includes photographs at no extra cost, which is not true in Rovaniemi or Reykjavík, where the picture is usually the upsell. In a market where a seat costs 163 EUR rather than 65, operators compete on deliverables instead of price. Second, the half-night format exists nowhere else in this catalogue, and it solves a genuine problem: a flight that lands at nine in the evening, or a schedule with exactly one usable window and no appetite for a six-hour commitment.

The aurora forecast that half the world reads comes from the Geophysical Institute at the University of Alaska Fairbanks, established here in 1946. Thirty miles up the Steese Highway is Poker Flat, opened in 1969, the only university-owned rocket range in the world, which fires instrumented rockets directly into active aurora to measure it from inside.

It is a strange thing about this town that the people driving you to a viewpoint and the people launching rockets into the same sky are part of the same small community. It also means the language on Fairbanks listings tends to be more precise than elsewhere: when a guide here talks about substorm onset or the Kp index, they are usually using the terms correctly.

This page previously recommended the shoulder months, on the reasoning that works everywhere in Scandinavia: equinox activity, temperatures a camera survives, avoid midwinter. Having run the numbers for Fairbanks specifically, that advice was wrong, and it is worth showing the working rather than quietly changing it.

Two things decide how much aurora time a month actually contains: how often the sky is usable, and how many hours of real darkness there are to use. Multiply them and you get the expected usable dark hours per night.

September, which we had recommended, is the worst month of the Fairbanks season by this measure and by a wide margin: a mediocre sky over barely three hours of darkness. January is nearly three times better. The Scandinavian rule does not cross the Atlantic, because in Fairbanks the cold and the clear sky are the same dry continental air.

For photography specifically, February is the month we would book. It gives 3.02 expected hours against January's 4.01, which is a real loss, and it hands back something the table cannot show: fifteen to twenty degrees. At minus 35 the failures in the table above stop being inconveniences and start ending the night, and the difference between a battery lasting twenty minutes and an hour decides how much of the substorm you actually photograph. January if you have Arctic experience and spare bodies; February if this is your first time.

Two limits on the cloud figure, stated because they cut against us. Our window covers September to March, so the April end of the operators' season is outside it and we say nothing about it. And below roughly minus 35 the Tanana valley fills with ice fog, a shallow local effect that a reanalysis on a 28 km grid will under-resolve: the 58.1% describes the regional sky, which is what a tour driving up onto Cleary Summit or Murphy Dome actually gets, and it overstates what somebody standing downtown sees in a deep cold snap.