As is the way so often with social media, the message took on a life of its own. Like a game of digital chinese whispers, the more it was retweeted and shared, the more hype it seemed to accrue. “In 38 minutes, it will be Kp 7!” declares a site, accompanied by a tantalizing graphic that shows a huge swathe of green sitting right over the top of the British Isles.

With messages like that, it’s easy to understand why those many people who are desperate to witness the beauty of the Northern Lights themselves are so excited and think that tonight will be the night.


And yet, for so many, the aurora display on Thursday 13th was most certainly a case of unlucky for some! After all the hype, after all the expectation, most came home having seen or captured nothing. What went wrong?

This display serves as a useful case study in managing expectations when it comes to aurora forecasting. I’m going to walk you through the build up to the display and the display itself in the hope that I might be able to help you understand a bit better how to interpret the aurora stats for yourself and to manage your own expectations when you are out next time. I’m going to use my three key questions that I find very useful when explaining to folks how to forecast the aurora for themselves.

1. Is it possible?

Source: http://www.swpc.noaa.gov/

Source: http://www.swpc.noaa.gov/

The auroras we see in the Mid-Latitudes are the result generally of one of two causes: a Coronal Mass Ejection or a Coronal Hole High Speed Stream.

The display on Thursday 13th was due to a Coronal Mass Ejection (an explosion of super-heated plasma from the Sun’s corona) on Monday 9th October. As the image below shows, this was directed towards the Earth (you can just make out a faint halo emanating out from the Sun - when it forms a halo like this, it’s a good indication that the explosion is directed towards our planet). Over the next few days, this stream of energetic plasma travelled the 93 million miles of space towards our planet. Meanwhile, on Earth, expectation was raised amongst us aurora chasers. Something like this meant that a display was certainly possible. We didn’t know precisely when - that depended on the speed that the plasma was travelling at - but we knew to be keeping an eye carefully on the stats in the coming days.

2. Is it probable?

As the solar wind of plasma travels from the Sun to the Earth, it’s a waiting game. No new data is coming in about what it’s like. We have to wait till the plasma arrives at the ACE satellite, about an hour’s journey time from the Earth. As it streams past this satellite, the instruments there measure the key attributes of the solar wind and beam that information back to us.

And, at about 06.00 BST on Thursday 13th, it seemed that it might have arrived. Data started to stream in, and the forecasts about what might happen later began to emerge.

Before we look at those in more detail, let’s step back from a moment and think about what characteristics the solar wind should have if we are to expect a good aurora display.

  • The solar wind needs to pack a decent punch

The aurora is produced by energetic particles from the plasma interacting with molecules in our atmosphere. The more energetic the solar wind is, the greater the kick it gives our atmosphere. And so a couple of the key things we need to look at in the solar wind are its speed and its density. The more tightly packed it is and the faster it is going when it impacts us, the better.

  • The solar wind needs to be orientated or aligned in the right way

The solar wind carries with it its own magnetism. And, for us to see good displays on Earth, it’s best when the magnetism of the solar wind aligns correctly with the magnetic field of our planet. What we are looking for from the ACE satellite for that are the following:

  • The stronger the magnetism of the solar wind, the better. This is measured by the Bt. Generally, values of 10 nT or greater are promising.
  • The orientation of the solar wind needs to be negative to allow the aurora to travel further south. This is where the Bz comes in. It is capped by the Bt value (it cannot have a numerical value greater than that of the Bt) but it can be positive or negative. And negative is good when it comes to the aurora!

So, what did the data from ACE suggest that got everyone so excited? You can see in the graph that the Bt rose up to a very impressive 20 nT -  and stayed elevated (in fact, staying at those very high levels for the next couple of days!). And, at the same time, the Bz dropped negative, for to below -10 nT, and then down to an equally very impressive - 20 nT. And it too stayed there for hours upon hours!

Source: http://www.swpc.noaa.gov/products/real-time-solar-wind%20

These were the stats that were getting folks very excited. A strong magnetism, with a consistently strong negative orientation are both very good signs for possible aurora shows. And, as these stats stayed looking good all throughout the daylight hours, the expectation built and built. Fuelled, of course, by retweets and sharing over the internet.

But, remember. The magnetism of the solar wind is only one of the factors that contribute to the aurora displays. We also need the wind to impact us with enough force to input sufficient energy into the atmosphere. And look at the speed on the graph - it sat at around ambient levels of 350-400 km/s for the entire day. And the density was pretty low too. These figures should have dampened expectations. But, with values of -20 for the Bz (yeah, OMW, negative freakin’ twenty, for goodness sake!!!), hype was building as we headed towards darkness. And lots of photographers were charging their batteries, wiping their lenses, and clearing their memory cards in anticipation of a decent display later.

3. Is it here?

Forecasts aren’t guarantees

When our kids were younger, we had a saying with them: ‘Plans aren’t promises!’ For it seemed that, as soon as we even discussed what we might do on a given day, if we dared change our minds were were bombarded by cries of, ‘But you said…!’

I often think of this when talking to folks about forecasting space weather. Let me adapt the saying a bit: forecasts are not guarantees. That doesn’t mean that they are useless (‘Those forecasters never get anything right!’); they are probabilities based on educated extrapolations from the data and so can give us informed speculation that can help shorten the odds when we're out aurora chasing.

But they can be wrong. Or, at least, over-hyped. Take for example the much used ovation oval. It is a forecast for about the next 45 minutes or so based on the data coming from ACE. It shows where the aurora is likely to be visible overhead. And it can be very useful. Research suggests that it is accurate 86% of the time. That’s not at all bad. But it’s not perfect. It can be misleading on occasion - and Thursday night seems to have been once such time.

What about the forecast Kp index? At one point on Thursday, it was forecasting an amazing 8.67 (out of 9!). Again, this forecast was wrong. The maximum the Kp actually reached as about 6. Respectable, but not anywhere near the dizzying heights forecast!

Source: SWPC http://www.swpc.noaa.gov/news/g1-minor-geomagnetic-storm-watches-issued

Source: SWPC http://www.swpc.noaa.gov/news/g1-minor-geomagnetic-storm-watches-issued

The magnetometers - real time data

Is there an alternative to these forecasts? Thankfully, yes. And it’s more real time data, this time courtesy of the magnetometers. These are Earth-based instruments that measure how our magnetic field is being affected by the solar wind. On these graphs, the bigger the movement on the lines (the greater the amplitude of movement), and the further south the the lines twitch, the better.

As the aurora shows can vary in a matter of minutes, access to this real time data is invaluable. By the time you respond to the ‘aurora will be active in 40 minutes’ kind of alerts, the active period may well have come and gone.

Source: http://flux.phys.uit.no/cgi-bin/mkstackplot.cgi

Back to Thursday 13th. Given the stats, I knew it was decidedly possible. I had a funny feeling that the forecasts were being over-hyped (not the most scientific statement I will make, I know!), but I knew that if I stayed in I wouldn’t see anything!

In addition, there was quite a lot of cloud forecast, and a near full moon would be hanging in the sky. These factors would potentially hinder visibility - and are a reminder of the other things you need to take into account when aurora chasing.

But out I was heading anyway. The magnetometers had been twitching since the late afternoon, and I suspected that there may well be something visible as twilight ended. And I was right!

As I got into position at Slemish, in fact I was about 20 minutes too late. I managed to catch the tail end of an aurora sub-storm with some pillars reaching up into the sky. And then the display settled down a bit.

There was a distinct enhanced band in the sky to the north. And it was elevated. In fact, a closer inspection showed in was a double arc aurora. A lower one just about poking over the horizon to the north, and a higher one drifting up through gaps in the cloud high up into the night’s sky. I could just about make out a faint green tinge to them with the naked eye, despite the moonlight.

I took some time lapse shots. But then the cloud closed in. As it cleared, the aurora had all but gone. And if you look at the magnetometers, they too died down. As the substorm died back, the aurora threw off some "blobs" (known as homogenous patches) which moved up into the sky above, before fading from view.


This seems to be around the time when many of my fellow photographers got into position around the country. Messages came in: I’m not really seeing anything. And that quickly became the dominant narrative on social media. This was a show that had failed to deliver.

Lessons to learn

Did this show fail to deliver?

Not really. A combination of over-hyped forecasts of ovals and Kp indices raised expectation levels too high during the afternoon. That, and the fact that the show was all but over by around 20.00, along with the bright moon, meant that the display - that was most certainly there - did not live up to the levels of hope raised during the day preceding it.

We need to be aware of the nature of the forecasts: forecasts are not guarantees!

We need to pay attention to both the magnitude of the solar wind (its speed and density) and its magnetism (Bt and Bz).

The sweet spot is when a fast and dense wind with good magnetism strikes. Those are shows you’ll never forget!

Use the magnetometers over the ovation ovals or Kp forecasts

The oval is good and has its uses, more so than the Kp forecasts. But there is no substitute for the real time data that comes in from the magnetometers. Use these and respond to the information there. On the 13th, they were clearly showing activity as we headed to darkness. If you wanted to capture the show that evening, you should have been out early!

There still is no substitute for putting in the hours

Even with all this information and real time data, the aurora rewards those prepared to put in the hours, standing in dark, often cold, sometime rainy places. You can’t expect to get a ping on your phone, pop out and see the show as soon as you step out of your car!

When you get it right, the rewards are still so worth it all!

Last Thursday, I saw a variation of an aurora display I’ve never seen before. I’ve seen pillars and blobs, but I’ve not yet seen an aurora throw off so many gentle arc/blobs as this one did. And so I’ve not only got some photos, but I’ve added to my bank of experience of aurora chasing. I call that a result!


Thanks as always to the good folks at Aurora UK on Facebook for help in interpreting the data on this one, especially James Rowley-Hill and Tracey Snelus. Any errors that may be above are entirely my fault and nothing to do with them!

Comment