Aurora Forecast: How to Read Northern Lights Predictions

Aurora forecasting relies on monitoring the sun and the stream of charged particles it continuously hurls toward Earth. The process begins at solar observatories — both ground-based and in space — that watch for coronal mass ejections (CMEs) and solar flares. When a CME erupts from the sun's surface, it launches a cloud of magnetized plasma across 150 million kilometers of space. If that cloud is aimed at Earth, forecasters can estimate its arrival time and intensity days in advance.

The journey from sun to Earth takes roughly 1 to 4 days depending on the CME's speed. Fast CMEs traveling at 1,000+ km/s can arrive in under 24 hours, while slower events may take 3–4 days. During transit, the DSCOVR and ACE satellites at the L1 Lagrange point — about 1.5 million km sunward from Earth — provide critical early warning. When the solar wind reaches these sentinels, scientists gain approximately 15 to 45 minutes of advance notice before it impacts Earth's magnetosphere.

The magnetic orientation of the incoming solar wind is the single most decisive factor. When the interplanetary magnetic field (IMF) tilts southward (negative Bz component), it can reconnect with Earth's northward-pointing field, opening a channel for solar particles to pour into the magnetosphere. A strongly negative Bz of −10 nT or more almost guarantees visible aurora at high latitudes, while a positive Bz tends to keep the magnetosphere closed regardless of solar wind density or speed.

This is why aurora forecasts can shift dramatically at the last minute — a CME might arrive on schedule but with a northward Bz, producing little aurora. Or an unremarkable solar wind stream might suddenly rotate southward, triggering spectacular unexpected displays. Understanding this mechanism is the foundation of everything else in aurora forecasting.

Several government agencies and research institutions provide the raw data that powers aurora forecasts. Understanding where this data originates helps you evaluate forecast reliability and interpret conflicting predictions.

Borealis, our own app, sits on top of these exact sources: it reads NOAA's live feeds and turns them into a Smart Aurora Probability Score for your location, with custom alerts and a live sightings map. It is launching on the App Store soon — see what it does and get notified.

NOAA Space Weather Prediction Center (SWPC) is the gold standard for operational aurora forecasting. Based in Boulder, Colorado, SWPC issues the official U.S. government space weather forecasts: the 3-day aurora forecast, the 30-minute Ovation Prime model, and geomagnetic storm watches and warnings. Their products combine automated model output with human forecaster judgment — much like terrestrial numerical weather prediction.

NASA operates several spacecraft critical to aurora prediction. The Solar Dynamics Observatory (SDO) provides continuous high-resolution images of the sun's surface and atmosphere, allowing forecasters to spot Earth-directed CMEs within minutes of eruption. The ACE satellite at L1 measures real-time solar wind conditions. NASA's Community Coordinated Modeling Center (CCMC) runs the ENLIL model, a 3D simulation of the solar wind that predicts CME arrival times and intensities.

The Finnish Meteorological Institute (FMI) operates auroral cameras and magnetometer networks across Scandinavia. Their real-time magnetograms from stations like Sodankylä and Nurmijärvi confirm whether a predicted storm is producing actual ground-level aurora. The FMI's auroral zone camera network is among the most extensive in the world, providing objective data on auroral activity independent of forecaster models.

The British Geological Survey (BGS) runs AuroraWatch UK, which monitors geomagnetic activity at observatories across the UK and issues color-coded alerts. Norwegian and Icelandic meteorological institutes (met.no and vedur.is) publish their own aurora forecasts combining geomagnetic data with local weather cloud cover — a genuinely practical combination for planning European aurora trips.

An aurora forecast typically presents several key metrics. Learning to read them quickly and together will dramatically improve your viewing success rate.

KP Index (0–9): The planetary geomagnetic activity index, updated every 3 hours. KP 0–1 means quiet conditions with aurora confined to the highest latitudes above 68°N. KP 3–4 indicates moderate activity visible across Scandinavia, Iceland, and northern Canada. KP 5+ signals a geomagnetic storm with aurora potentially reaching Scotland, the northern United States, and southern Scandinavia. KP 7+ is a severe storm that can push displays into central Europe and the southern United States. See our complete KP index guide for precise location-specific visibility thresholds.

Bz Component: Displayed in nanotesla (nT), this measures the north-south orientation of the interplanetary magnetic field. Negative values are favorable — Bz below −5 nT generally enhances aurora, while Bz below −10 nT often triggers strong displays. Many experienced watchers consider Bz more immediately useful than the KP index for real-time decisions, because it updates continuously and responds faster to incoming solar wind changes.

Solar Wind Speed: Measured in km/s. Quiet conditions typically show 300–400 km/s. Enhanced speeds of 500–700+ km/s combined with negative Bz create ideal aurora conditions. Very fast solar wind above 800 km/s can produce sustained multi-hour displays as the magnetosphere is continuously energized.

Solar Wind Density: Measured in particles per cubic centimeter (p/cm³). Values above 10 p/cm³ are elevated, and density spikes often correlate with substorm onsets. High density alone does not produce aurora — it needs to combine with a southward Bz to be effective.

The Ovation Model Map: NOAA's Ovation Prime model produces a visual map of the auroral oval showing predicted intensity around the polar region, updated every 30 minutes from real-time DSCOVR satellite data. Color coding runs from green (faint glow) to red (intense, naked-eye display). This map is arguably the single most practical forecasting tool — it shows exactly where aurora is expected, not just a global activity number.

Aurora forecasts operate across several timescales, each with different reliability characteristics. Understanding these distinctions helps you plan both your trip timing and your nightly viewing strategy.

27-day recurrence outlook: Because the sun rotates roughly once every 27 days as seen from Earth, active regions that produced geomagnetic storms can return to an Earth-facing orientation about a month later. Forecasters use this recurrence tendency to issue rough outlooks. Reliability is low — perhaps 30–40% — because solar active regions evolve rapidly. These outlooks are useful for trip planning in a general sense but not for booking specific nights.

3-day forecast: When a CME is observed leaving the sun, forecasters can predict its Earth arrival within a window of roughly ±6–12 hours. NOAA issues 3-day forecasts with predicted KP values for each 3-hour window. If a major CME is definitively Earth-directed, this forecast correctly predicts elevated activity about 60–70% of the time, though exact timing and peak intensity remain uncertain. This is the best range for activating travel plans.

1-day forecast: As CMEs approach and are detected by upstream monitors, the arrival window narrows significantly. One day ahead, forecasters can typically predict KP levels within ±1 point. This is the most actionable lead time — if you have flexibility to drive to a clear-sky location, a strong 24-hour forecast is highly reliable.

30-minute nowcast: The NOAA Ovation Prime model uses real-time DSCOVR data to predict aurora location and intensity 30 minutes ahead. This is the most reliable forecast available, essentially projecting current solar wind conditions forward onto the auroral oval. During active periods, refresh this model frequently as conditions can change dramatically within minutes.

The practical framework: use long-range forecasts to select your travel dates, 3-day forecasts to choose your viewing location, 1-day forecasts to finalize logistics, and the 30-minute nowcast to decide exactly when and where to step outside.

Even experienced aurora watchers fall into forecasting traps that cost them sightings. Here are the most frequent errors and their solutions.

1. Ignoring cloud cover: The most common reason people miss the aurora is not checking the weather forecast alongside the geomagnetic forecast. A KP 7 storm means nothing if your sky is overcast. Always cross-reference aurora predictions with local cloud cover. In Scandinavia use yr.no; in Iceland, vedur.is; in North America, check Environment Canada or the NWS. Our tonight's aurora forecast page covers both dimensions.

2. Giving up before midnight: Auroral substorms most commonly peak between 10 PM and 2 AM local magnetic time. Many people check the sky at 9 PM, see nothing, and go to bed — missing a spectacular display that erupts at midnight. If the forecast shows elevated activity, commit to staying out until at least 1 AM, or set an alarm to go back outside.

3. Fixating on KP alone: The KP index is a 3-hour average that can mask brief but intense substorms. You might see KP 3 on the forecast while a 30-minute burst of KP 5-level activity lights up the sky. Use real-time Bz and magnetometer data alongside KP for a complete picture. A sudden drop in Bz to −10 nT or lower is often the best leading indicator of an imminent substorm.

4. Underestimating light pollution: Urban skyglow can wash out all but the strongest aurora. Even a KP 5 storm can appear disappointing from a city center. Getting just 20–30 km from urban light domes dramatically improves visibility. Use our location-based forecast guide and light pollution maps to find genuinely dark sky sites near you.

5. Trusting a single source: No single forecast model is infallible. Cross-reference NOAA's Ovation map with real-time Bz readings, ACE/DSCOVR solar wind data, and ground-based all-sky camera feeds. When multiple independent indicators align — negative Bz, elevated solar wind speed, rising KP, and active magnetometers — confidence in a display is high. See our aurora forecast tools page for the best combination of sources.