WeatherMatrix (Jesse Ferrell)

Why Is the Aurora So Hard to Forecast?

By Jesse Ferrell, Meteorologist/Community Director
11/09/2015, 3:17:57 AM

UPDATE 11/4: The aurora did stay out of the U.S. last night, but those with visibility far enough north were able to see them. The only photo that I saw coming out of Pennsylvania was this one, taken in extreme NW PA:


Here's the satellite shot, showing the Aurora at 0700Z:


UPDATE 5 PM: Current Aurora forecast for tonight is not looking good for skywatchers; the Northern Lights should stay mostly in Canada.


ORIGINAL ENTRY: NOAA, AccuWeather and issued stories yesterday predicting a southern dip in the "auroral oval" (area where you can see the northern lights) last night. Auroras appeared farther south than usual in Nebraska and Iowa, but skywatchers in the Northeast were disappointed with little south of New England. Something similar happened a few months ago. Tonight's success remains to be seen but today's forecast sounds less impressive (Kp of 5 at night) than it did yesterday (Kp of 7 at night). Here's our official news story for tonight:


What is the Kp Index?

First things first, from the NOAA website: "The Kp Index is one of the most common indices used to indicate the severity of the global magnetic disturbances in near-Earth space." Those global magnetic disturbances are what produce the Aurora. From my experience at my latitude, in Pennsylvania, we have little hope to see the Northern Lights with a Kp Index of less than 7 (see map for your area). Here's this afternoon's Kp Index:


Why are the northern lights so hard to forecast?

I'm no astronomy expert by far, and I'm sure those more familiar with the phenomenon can add to this blog in the comments below. To me, as a scientist and casual observer of the assets available to forecast the aurora, it seems like we're in the very early stages of being able to accurately predict this phenomenon. Space weather forecasting history is measured in years, not decades like weather forecasting. As a result, accuracy is not high (but it will continue to improve).


Here are some of the challenges:

FLYING BLIND: One of the major tools used is coronal hole observations from the far side of the sun, two weeks ago. In space physics, as with atmospheric physics, a lot can change in that time period.

PIXIE DUST FORECAST: Even if you could accurately predict the arrival of a solar flare, sometimes the northern lights don't appear over your house. The Earth doesn't get covered in magic aurora potion; the southern extent of the phenomenon is related to the strength of the solar flare, which varies across the globe. Add to this the problem that you can only see the aurora during nighttime hours, and you've got an almost-guaranteed-to-fail prediction.

MORE UNPREDICTABLE THAN THE WEATHER: Think of the weather forecast -- if a snowstorm's arrival is two or three hours late, some people might not even know the difference. If your northern lights forecast is two or three hours late, you're past most peoples' bedtimes, or you're in the morning and the show is hidden by the Sun. This is what happened last night. To add to the problem, the Army says: "Magnetically active times, e.g., Kp > 5, are notoriously difficult to predict." You can literally see this failure in last night's forecast (dots) vs. the actual Kp (blue line):


BETTER SAFE THAN SORRY: The forecasts are what they are, and they are probably scientifically (or at least statistically sound). When it comes to the government, commercial websites and social media that distribute the forecasts, the old adage "better safe than sorry" comes to mind. Given the damage that solar storms can cause, I'll raise the alarm more often, even if it's proven false, rather than miss "the big one." In addition, skywatchers can be a little hopeful (in snow forecasts, we call this "wishcasting") and misread the forecasts as sureties.

Links to Predict the Northern Lights: (sorted by forecast time)

- Current Aurora from NOAA - Updates every hour - Current Aurora* from SSEC VIIRS Day/Night Band - Current Kp Index from NOAA - Updates every 3 hours - 30-Minute Aurora Forecast from SpaceWeatherLive - notes chances of Mid-Latitude Event - 1-Hour and "Today" Aurora Forecast (Maps) from the University of Alaska - AccuWeather Viewing Conditions Map: on Facebook and AccuWeather Astronomy on Facebook (when available - combines the weather and the geographic chances of aurora, as shown above, for one or more nights) - 4-Hour Kp Forecast (Graph) from the U. S. Air Force Weather Agency - 1-Day Aurora Forecast (Map) from the SWPC; Only updates once a day - 3-Day Aurora Forecast (Map) from the SWPC; Only updates once a day - 3-Day Geomagnetic Storm Forecast (Text) by the Space Weather Prediction Center

*This is from the Suomi-NPP Satellite Maps, but they only take three pictures per night. If you can't find the sat pics at that web address, CIMSS usually posts them on Social Media the morning after.

The views expressed are those of the author and not necessarily those of AccuWeather, Inc. or


Comments that don't add to the conversation may be automatically or manually removed by Facebook or AccuWeather. Profanity, personal attacks, and spam will not be tolerated.

WeatherMatrix (Jesse Ferrell)