Northern Lights Forecast
A northern lights forecast is really two forecasts stitched together: a sharp, near-certain nowcast for the next half hour, and a fuzzier outlook that reaches days ahead as the Sun rotates. Knowing which part you're reading is the difference between a wasted drive and a night under a dancing sky.
Open the live tracker →The two time horizons of a forecast
Every northern lights forecast blends two very different kinds of prediction, and they don't share the same confidence. The short-term nowcast covers roughly the next 30–60 minutes and is the most trustworthy figure you'll see. The multi-day outlook stretches one to three days ahead and is a probability, not a promise. Treat them as separate tools: the nowcast tells you whether to grab your coat right now; the outlook tells you which nights this week are worth keeping free.
Now → 60 min
Driven by spacecraft parked at the L1 point, ~1.5 million km sunward. They sample the solar wind before it reaches Earth, giving a reliable head start. Space weather →
1–3 days
Built from the Sun's ~27-day rotation, known coronal holes and any CMEs already launched. Useful for planning, but the arrival time and strength can shift by hours. Geomagnetic storms →
Tonight
The practical middle ground — a read on whether the coming dark hours are likely to deliver. Northern lights tonight →
Why the short-term nowcast is so reliable
The near-certainty of the 30–60 minute forecast isn't luck — it's geometry. Spacecraft at the L1 Lagrange point sit directly upstream between the Sun and Earth and physically measure the solar wind before it arrives. When they detect the interplanetary magnetic field turning sharply southward (a negative Bz, ideally below −10 nT) inside a fast stream of 500–800 km/s wind, that same parcel of plasma is minutes from slamming into Earth's magnetosphere. There's little left to guess. This is why a good nowcast feels almost like watching weather roll in on radar: the trigger has already happened, and you're simply waiting for it to reach the ground.
Why the multi-day outlook stays uncertain
The days-ahead half of a northern lights forecast is harder because the drivers haven't left the Sun's neighbourhood yet — or have only just departed. Two mechanisms dominate the outlook. Coronal holes are cooler, open regions that fling out high-speed solar wind; because the Sun rotates roughly every 27 days, a hole that lit up the sky last month often swings back around to face Earth again, making the outlook partly a calendar exercise. Coronal mass ejections (CMEs) are the big-ticket events — billions of tonnes of plasma hurled outward by a solar eruption. Forecasters can see a CME leave the Sun, but predicting its exact travel time (typically 1–3 days) and whether its embedded magnetic field will arrive pointing south is genuinely difficult. A CME that looked storm-worthy can fizzle on arrival, and a modest one can over-deliver if its Bz swings hard negative. That irreducible uncertainty is why the outlook is always framed as odds, and why the KP index forecast for tomorrow carries error bars the nowcast never needs.
Turning the forecast into a go / no-go call
A forecast is only useful once it becomes a decision. Work from the outlook down to the nowcast. Start days ahead: if the multi-day outlook flags an elevated KP or a coronal-hole stream, pencil in the night and check the sky-cloud forecast — clear, dark skies away from light pollution matter as much as the space weather. On the night itself, watch the live numbers converge. The green-light combination is three things agreeing at once: Bz sitting firmly southward, solar wind speed lifting toward the upper end of its ~300–800 km/s range, and the KP index climbing to 5 or above (a G1+ geomagnetic storm) so the auroral oval expands toward your latitude. When those align in the short-term nowcast, that's your go signal — head out. If Bz is stubbornly northward or the wind is slack, it's a no-go no matter how promising the day-ahead outlook looked. Let aurora alerts do the waiting for you, and check the aurora map to see whether the oval is actually reaching down to where you are.
Live tracker
KP, Bz, solar wind and a single 0–9 read on the odds, updated from satellite data. Open it →
Aurora tracker
Follow the drivers as they evolve through the night. Aurora tracker →
Bz & solar wind
The two switches that decide whether energy actually gets in. Bz & solar wind →
Learn aurora
How the whole chain works, from Sun to sky. Learn aurora →
Northern lights forecast FAQ
How far ahead can a northern lights forecast actually see?
Reliably, only about 30–60 minutes — the time it takes solar wind measured at the L1 point to reach Earth. Beyond that you're into an outlook, not a forecast: a 1–3 day probability based on the Sun's rotation, known coronal holes and any CMEs already in transit. The further out it reaches, the wider the uncertainty on both timing and strength. More on space weather →
Why does the forecast keep changing through the evening?
Because the most important driver — the direction of the interplanetary magnetic field (Bz) — flips on its own, sometimes every few minutes. A night can look quiet, then a southward turn arrives inside a solar-wind stream and activity jumps within the hour. That's normal, and it's exactly why the short-term nowcast is worth rechecking rather than trusting a single day-ahead number. Why Bz matters →
The outlook promised a storm and nothing happened — why?
Almost always it's the CME's magnetic field. Forecasters can predict that a coronal mass ejection will arrive, but not reliably whether its embedded field will point north or south. If it arrives northward, it barely couples to Earth's magnetosphere and the aurora stays weak even at high solar-wind speeds. The reverse also happens: a modest-looking event over-delivers when its Bz swings hard negative. About geomagnetic storms →
Watch the forecast turn real
See the nowcast and the outlook side by side — live KP index, Bz, solar wind and a single 0–9 score, refreshed every minute from satellite data.
Open the live tracker →