Notes on viewing the Aurora Borealis
[Published June 1, 2025, by Mikhail Kats]
Living in Wisconsin gives you the opportunity to occasionally see the Aurora Borealis. After failing to see it many times, and then finally seeing it a few times, I am compiling a short document on what I’ve learned and on how to look for the aurora.
What causes the Aurora Borealis?
Charged particles (electrons, protons) from the sun’s corona known as the solar wind are deflected by the Earth’s magnetic fields in the magnetosphere, and eventually funneled to the polar regions (North and South). The charged particles run into atoms and molecules of the dilute gases in the atmosphere, bringing them to an excited state, and then they emit light.
How to know if the aurora might be visible?
Basically, you want:
To be as far North as possible (or as far South as possible, if in the Southern Hemisphere)
For the geomagnetic activity to be as high as possible
The sky to be as dark and clear as possible
To be patient and get lucky
To know if the conditions will be good, I recommend using an app. I use “My Aurora Forecast” by jRustonApps (no affiliation), but there are others. I have the app set to notify me when the chance of an aurora is high, based on measurements of the solar wind, knowledge of whether there was a recent coronal mass ejection (CME), and other factors.
Where to go
If you think the aurora may be visible in the evening or night, you want to check the weather forecast (or just look up..) to see if the sky is clear. Ideally you should be able to see many stars when it starts to get dark.
Then, you want to get away from as much extraneous light as possible, and find a location with a clear view toward the north. If you want to get fancy, you can look at a map of light pollution (e.g., here) and decide how far you want to drive to a darker location. I found that a 30 min drive away from Madison, WI works quite well, though I’ve even seen auroras from my front porch in the suburbs, so in a strong geomagnetic storm (like a G4), you might not have to go far.
How to look for the aurora
Auroras are fleeting. Sometimes you can look for hours and not see much, and then a great show appears for 15 minutes, and then goes away. So, be patient (and pack warm clothing, if it’s cold).
It is rare to see colorful auroras with your naked eyes unless you get really lucky and/or are really far north. Almost all of the photos taken by other people look better than what you will see with your naked eyes due to a combination of a good camera sensor and long exposure.
Use your phone camera as a viewfinder to spot faint auroras. Even regular photos and videos on modern phones do better than your not-night-adjusted color vision, but it really helps to switch your phone to night mode. If you keep your phone steady for 1-6 seconds on night mode, you will be able to see even faint colorful auroras. For example, here is a photo I took using night mode, where a faint green aurora can be seen, but was not visible at all with my naked eyes:
Once you know what you’re looking for, you can night-adjust your eyes by not looking at any screens or other light sources for a few minutes. It can be helpful to have a red flashlight to help you move around, since red light does not affect your night adaptation. Using your night-adjusted vision, you can see faint auroras as shimmering wisps of white/grey light.
How to photograph the aurora
Night mode on modern phones is surprisingly good, so I strongly recommend that as a starting point for aurora photography. Here is a smartphone photo I took from the front porch of our house during a G4 storm in October 2024:
If you want to use a serious camera, you want to:
Use a tripod
Have a large aperture (low f-number) so as much light gets in as possible
Focus at infinity in advance, and lock or tape down (or at least don’t touch) your focus
Be ready to vary the ISO and exposure as needed
Usually a wider-angle lens is more interesting, maybe something like 20mm. That way, you can get a good view of the sky, possibly including some scenery
Shoot in RAW mode, so you have more flexibility during post-processing
Usually set the shutter speed for 1-15 seconds, where longer exposures introduce more blur
Here is an example aurora shot from Barry van Veen during a G4 storm in May 2024, 30 mins away from Madison:
And here is an aurora over the UW-Madison campus shot from an apartment building by Erick Yin and Sabrina Wu, from April 2023. You can watch the entire timelapse here.
Now for a bit of science
Where do the different colors come from?
Here is a spectrum of that you might expect to see from the Aurora, though keep in mind that the vertical axis is on a log scale. The most prominent lines are marked by the atoms/molecules that are emitting
The most common color is green (specifically at a wavelength of 558 nm), from fluorescence from oxygen at an altitude of around 120-400 km. The red (a number of peaks) is also fluorescence from oxygen, but at higher altitudes above 300 km, where the density is lower. The purples and blues come from molecular nitrogen (N2).
You may have seen a graphic like this floating around the internet (I don’t know the original source):
It's a nice graphic, but it doesn't explain why oxygen at different altitudes emits different colors. Here's why: The ¹D excited state of oxygen has a lifetime of >150 seconds, compared to the ¹S state which has a lifetime of ~1 second. At higher densities which are present at lower altitudes, the oxygen is likely to collide with another gas molecule or atom, knocking it out of the ¹D state.
Now let's talk about how aurora activity is predicted and measured
Geomagnetic activity: every 11 years, the Sun’s magnetic pole flips, so the amount of solar activity leading to strong auroras has a period of 11 years. The current peak is 2024-2025, with the minimum expected in the early 2030s. Coronal mass ejections (CMEs) are big events that result in many charged particles being launched; if a CME is toward the Earth, then we might expect strong aurora activity several days later.
Data and predictions: predictions about aurora strength are done on the basis of several pieces of data, all quite imperfect. These start with observations and measurements of the sun, such as observation of CMEs. Since CMEs take 1-5 days to reach the Earth, that’s the kind of warning you should expect to have for major aurora activity. The NOAA puts out a 3-day forecast.
Planetary metrics: the NOAA created a scale for geomagnetic storms, from G1 (minor) to G5 (extreme). I believe this index is mostly based on the “K index” or “Kp index” or just “Kp”, which generally ranges from 1 to 9 and is a normalized and averaged measurement (on a logarithmic scale) of the horizontal component of Earth's magnetic field at 13 locations on Earth over a three-hour interval. Where we are in Wisconsin (43° North), G3 or higher (Kp = 7 or higher) seems to provide pretty good opportunities to see the aurora. The last two spectacular aurora shows here, during which we took the photos above, were from G4 storms (Kp = 8).
You need more than plenary metrics: a combination of measurements and predictive tools are put together to provide a probability map of where the aurora may be visible. This is in particular informed by measurements of the solar wind by satellites located at the L1 Lagrange point between the Earth and Sun.
Here is what an aurora forecast from the NOAA looks like, which is provided on an hourly basis for the next 15-60 minutes (just a screenshot from when I wrote this, not real time):