Sunday, February 25, 2024

On The Big X-class Solar Flares & the AT&T Blackout

Last week saw the strongest solar flares of solar cycle 25 and the strongest since the great solar storms of Sept. 2017.  The X6.3 solar flare erupted from sunspot AR3590 on Feb. 22 @ 2234 UT or 5:34 PM EST.  What was unique about this strong flare was that there was no Coronal Mass Ejection (CME) with it, and the charged particles from a CME are most likely to cause a geomagnetic storm when they reach Earth.  No such storm has occurred.  The flare itself produced some ionization in the upper reaches of the ionosphere and caused some brief radio propagation issues. included this plot of signal strengths at 20 MHz, submitted by Jim Tegerdine from Roseburg, Oregon:

Image credit to Jim Tegerdine.

"I was monitoring radio transmissions at 20 MHz when all the signals started to fade," says Tegerdine. "At the peak of the flare (2234 UT), the signal strength was near zero."

Tegerdine's chart recorder plot shows that the fading began about 10 minutes before the flare's peak, giving him an early warning of what was about to happen. After reaching the nadir of signal strength, the blackout persisted as a "deep fade" for more than 20 minutes. In total, there was loss of signal for about a half an hour.

The rule of thumb with disturbances like this is that lower HF frequencies suffer more attenuation for longer periods than frequencies like 20 MHz.  I'd be completely unsurprised to hear that the 3.5 or 7.0 MHz were blacked out for longer than a half hour.

Sunspot AR3590 is close to the center of the sun's Earth-facing side, while moderately high in latitude and among the first, if not THE first, naked-eye sunspot of cycle 25.  It's wider than 10 Earth diameters and relatively easy to see, although I feel obligated to say not to look at the sun without something like Eclipse glasses - dark enough film to reduce the brightness down to safe levels.  If you have a welding helmet, you're there.  In previous sunspot cycles I would catch a glimpse of big spots on the way into work this time of year, when the rising sun might be behind clouds dense enough to make looking at the sun painless.  

Image credits as shown in the image.

AR3509 has the magnetic configuration that makes more flares possible as it goes through its second week of rotation with the sun. No one can predict the strength of potential flares, at least with much accuracy, but NOAA is putting the chances of at least an X1 flare at 30% through the next 48 hours.  I view that as just a bit more trustworthy than the local fortune teller.  (You know - the gypsy with the gold-capped tooth.)

There was a lot of speculation that the blackout some cellular providers experienced on the 22nd, particularly AT&T, was caused by that X6.3 flare.  People who study these things were quick to point out that without a CME to bring charged particles to Earth, that was pretty unlikely.  By the next day, AT&T was saying, "sorry - our bad" and taking responsibility for the whole mess.  "Anonymous sources" said it was a software update that went wrong.  Today, I've gotten an email and text messages from AT&T (they're my phone service provider) saying they're giving everyone a credit on our bills to compensate for no service while it was down.

"Based on our initial review, we believe that today's outage was caused by the application and execution of an incorrect process used as we were expanding our network, not a cyber attack," AT&T said on its website last night. "We are continuing our assessment of today's outage to ensure we keep delivering the service that our customers deserve."

I've seen some messages that say other providers had issues, but with the way all those networks are interconnected I can't say they couldn't have affected other providers.

To err is human but to really screw things up requires software.


  1. Definitely surprising that there was no CME. Will certainly keep watching what the sun is doing as we approach peak cycle.

  2. I understand that the most popular site that tracks cell outages counts "out" as when a connection cannot be made. That means that both ends get dinged, which means that anyone from AT&T calling anyone on Verizon will show a failed connection for each provider. Thus the outage plots all look the same due to cross-provider calls.

    I would take AT&T at their word that they caused it. It's like pulling teeth to get them to admit fault, so when they do it's likely the correct answer.

  3. Oddly enough, I have AT&T and had no issues that day... But out here, there were times other companies did, I assume due to their interconnectedness.