Lost in the Midnight Sun: Misreporting The Aurora and Science (Il)Literacy in Journalism

Attention all journalists! Especially science reporters. We need to talk.

On July 10, a post on Reddit caught my eye. It was a link to an article in USA Today with the title, Someone at USA Today doesn’t know how summer works here. Here’s a screenshot of the article:

If you want to read the article, there’s a link in the Reddit post linked above. I’ve decided not to post links to any of the articles here. I implore you to not visit them, or at least you make sure you have a working adblocker so as to not give revenue to these sites for posting what is essentially click-bait from major news organizations, most likely AI written and not given a second thought before publishing.

So, what is the issue?

Well, for much of the state, it doesn’t get dark now. Fairbanks is situated as close to the middle of the state as you get for a large town or small city (if you exclude the panhandle). The next sunset in Fairbanks after this article publishes on July 13 will be at 12:05 am tomorrow morning. Anchorage is a little better, experiencing about 2 hours of nautical twilight tonight, where the brighter stars will be visible, but for the most part, you can still see your surroundings. It would technically be possible, but highly unlikely to be able to see aurora. The USA Today article even claims that it will be visible in Utqiaġvik! The Sun does not set in Utqiaġvik (formerly known as Barrow) until August 2. I assure you, you will not see the aurora when the Sun is in the sky. If you do, we are in a lot of trouble.

Usually, the aurora season starts when an area starts having at least astronomical twilight again, give or take a few days. You’d really have to go down near Sitka or Ketchikan for it to be dark enough to have a solid chance of seeing the aurora. Here’s a map I spent 2 seconds preparing for reference:

Of course, I’m ignoring something. It might be dark enough at those lowest latitudes of the state, but the aurora needs to be pretty active to be visible there. And it isn’t actually forecasted to be. But I’ll come back to this because this is another huge issue with this and other reporting from the last few days.

Let’s rewind to the start. I’m pretty confident that USA Today was not the origin of this fiasco, having published their article on July 9, 2023. Rather a day earlier, the Associated Press ran a story, “Solar storm on Thursday expected to make northern lights visible in limited US states.” However, most of the hullabaloo seemed to surround the USA Today article. But they really ran with it, posting individual articles for states that “may see” it, copy-pasted articles with only the names of the states changed. Tim Ballisty posted an insightful Twitter thread that helped trace the wildfire spread of this misinformation.

Hold up! 17 US states seeing the northern lights this week? That’s the kind of sensational headline that rakes in the clicks and ad revenue! Tim astutely noted that this story was promptly parroted by NPR, NBC, CBS, ABC, World News Tonight. Each one jumped on the bandwagon, missing the glaring issue at hand.

Here’s a gem from Fox News . . .

I really enjoyed the “according to the institute.” They never mentioned any institute previously, so I have absolutely no idea what they are referring to. Maybe it’s the Geophysical Institute (GI at UAF) that they mention in the last paragraph. That last mention, GI forecast of Kp 6. There was a brief forecast at Kp 6 for July 13 on the 10th (wayback machine), the day these reports really started making headlines. That would potentially spark auroras in the lower 48. However, that forecast was soon downgraded to Kp 4 or lower.

So, here’s what I suspect happened.

A reporter caught wind that there was a Kp 6 prediction from the GI, looked up a map for the aurora statistical oval at Kp 6, didn’t bother confirming anything, used AI to write a sensationalized article and headline, other news organizations saw it and didn’t want to be left in the dust, had their own AI using reporters re-write the original article without fact checking anything, here we are today. Of course, that’s just my speculation.

USA Today is still going at it with a video posted today titled “Solar storm could cause the northern lights to reach a third of U.S. states.” No. No, it won’t. There is a possibility of Kp 5 (GI forecast) on the 15th due to a coronal mass ejection (CME) that may land a glancing blow on Earth’s magnetosphere. That could spark some auroras over much of Canada and the northern U.S., the standard Minnesota, North Dakota, and Montana, but not much more. Maaaaybe Washington, Idaho, or northern New England if lucky. One-third of 50 is 16 (truncating or rounding down for their benefit). But fractions are hard.

NOAA is predicting a G1 (minor) geomagnetic storm, Kp 4.67 for the 15th. Even less likely that the lower 48 would see much of anything.

Do you know who else isn’t going to see it? About 95% of Alaska. And I’ll bet that 5% will be cloudy.

Bigger Beef

Maybe one of the biggest issues I have with this cascade of ill-reporting is that no one bothered to check with any experts. Call up the GI, guys! Get a statement, and find out more info you can pass on to your viewers or readers.

In all the articles, I did see that when USA Today doing a follow-up article, relaying that possibly the aurora wouldn’t be as good (note, not retracting or correcting the original six articles saying otherwise), they asked a meteorologist with Accuweather.

I’m aware that not everyone knows what causes the aurora (hey, I wrote an article on that; check the link if you want). If you do happen to have a basic understanding or have the ability to use Google or Wikipedia, you will quickly find that the leading experts on aurora and spaceweather are not meteorologists with a digital media company. How on earth does a reporter not know what or who a reliable and reputable source is? Like, if they have one job, shouldn’t that be it?

To summarize, I have three huge issues with this whole mess:

1. A troubling lack of scientific literacy in reporting.
2. The amplification of sensationalist headlines.
3. The domino effect of news organizations bandwagoning unverified, likely AI-generated, content.

I get it. You need clicks. You need viewers to see your ads to make money. Like this:

---

Affiliate link – I earn a commission if you shop through the link(s) below at no additional cost to you (more info)

---

As soon as you sacrifice your integrity for that ad revenue, you soil your reputation. Now, I don’t expect a ton from USA Today; they’ve been going down this route for a while now. But NPR and Alaska Public Media also jumped on that same bandwagon.

At least NPR caught the mistake and (mostly) issued a correction. I’ll go ahead and post a link to that one, thank you, NPR. I do find it amusing that everyone that issued some form of a correction tried to place the blame on the GI, rather than acknowledge that they ran with articles from a source they didn’t understand or try to understand. From that article:

A previous forecast by the University of Alaska Fairbanks said the aurora would be visible in more than a dozen states.

The university now says the previous forecast derived from a long-term forecast based on the sun’s rotation that went back 54 days.

Ayana Archie – NPR July 12, 2023

The forecast did not say the aurora would be visible in more than a dozen states. The forecast gives a predicted Kp index. News organizations extrapolated that it meant it would be seen in more than a dozen states (17 – see, they rounded up, I truncated earlier). That number isn’t even right, given what the Kp index prediction was. Most impressive (sad that I have to say that) is that NPR actually got statements from representatives at UAF’s Geophysical Institute (from which reporters took the misrepresented data) and NOAA’s Space Weather Prediction Center.

The first rule of scientific literacy is knowing who to ask. I don’t expect a science reporter to know everything they write about. Researchers go to school and research their fields for decades to get the knowledge they possess. Journalists need to know who to ask to properly interpret what they’re seeing so they can convey it to their readers and not misrepresent data, as happened here. It’s a reporter’s job to be able to explain, often complicated, scientific ideas in ways that can be understood by the general population without extensive backgrounds in the fields.

This kind of reporting is blatantly irresponsible, as it passes off scientifically interesting but inaccurate and, frankly, wrong information as headlining news in a way that will have a compounding effect on scientific illiteracy in the general population.

This was more than a case of asking the wrong people; they just didn’t ask. This is obvious since a common thread in many of these articles was stating the cause of the geomagnetic storm was from fast solar wind emanating from a coronal hole. While this is a way of stimulating geomagnetic activity, it is certainly not the case here; there are currently no significant equatorial coronal holes on the Earth-facing side of the Sun. Rather, the recent potential for activity is due to magnetic solar filaments that have released coronal mass ejections that have been partially Earth-directed. It’s as if they were trying to literally get every aspect of their reporting incorrect.

This is what leads me to believe this is just an AI-written dribble. Not only do these “reporters” not fact-check or seek sources, but they’re also bad at AI prompting. Full disclosure: I use AI in my writing, but not to write. I utilize it as a proofreader for revisions (that I make myself) or occasionally rephrase my near-constant run-on sentences. I would never use it to generate content. It’s wrong almost all the time. Tools like chat-GPT are large language models. Not expert scientists. News organizations have no business using it as such.

As stated before, that is just my speculation. The alternative is that all of these organizations are employing really, really bad journalists. When I looked through other publications written by these “Authors,” they had multiple articles a day that were all over the place. They don’t cover any beat; they aren’t science reporters. When I reached out to the author of the USA Today article, Amaris Encinas, the given e-mail address bounced (AEncinas@gannett.com). I’m a little surprised that any major news organization would provide incorrect contact info for journalists. Other articles written by her in the last few days include: “Boy saves 2-year-old sibling from drowning in family pool, “This year’s Gerber Baby is a military kid from Colorado,” Get discounted subs through Sunday in honor of the 15th annual Hoagiefest.” Sigh. I’m tired.

Of course, in the case of the Associated Press, the likely originator of this mess, they don’t include an author, so it’s not known if it is someone who should know better or not. This is actually a major reason that I don’t have a lot of faith in the AP. There doesn’t seem to be accountability for journalists in the organization.

This reporting never should have gotten out or spread as fervently as it did. The author of the Fox article, Bradford Betz, is a “breaking news reporter for Fox News Digital covering crime, political issues, and the economy.” So why . . . I’ll stop there.

These organizations have science reporters. Why doesn’t content go through them? My guess is they are just trying to amalgamate as much sensationalized content as they can to get clicks. Whoever gets the most clicks “wins.” If other organizations are getting clicks on something you don’t have, copy it. Don’t check it, don’t correct it when you find out it’s wrong, write another article explaining it was wrong. Then you get even more clicks. Click, click click. Here’s another ad:

My message to news organizations: Be Better

Leave science reporting to science reporters, and make sure they’re qualified. In my humble opinion, the editors that let this crap go through should be fired. At least, that is, if you want your organization to be credible and not click-bat.

And my message to readers: be aware. They’re all trying to get your clicks. And they won’t let facts stand in their way. I do want to state every major TV network ran with this story, at least on their online editions. A few traditional print outlets like the New York Times and Washington Post did not. While that is just anecdotal, single-point evidence, it already makes me suspect them to be more reputable in their science reporting. To be fair, I already felt that way anyway. I’m a bit surprised that NPR and Alaska Public Media fell into it, but good on them for trying to correct it.

Wow.

tl;dr: You are probably not going to see the aurora tonight in Alaska. But, in about a month, check the skies! Also, besides the NOAA Space Weather Prediction Center, spaceweather.com is a fantastic site to get up-to-date, easy to understand, and very accurate picture of what is going on in our heliosphere. Also, auroranotify and their Facebook or Twitter pages are great places to follow when you’re likely to see it in various places around the world. If only these journalists had looked at any one of these sources, I wouldn’t have written this article. So, I guess I got my clicks. Thanks, guys!