As Stormwatch team-member Steve commented to us a couple of days ago, solar storms are like buses. None for ages and then three turn up at once. This time however, we were ready for them and thanks to all of you who took the time to click on our data we were able to predict the arrival time to within 7 hours – our closest prediction yet! The Stormwatch alert warned of a storm heading in a direction that would take it within 3 degrees of Earth and that it would arrive at around 9 UT (GMT) on the 18th February.
At around 01:00 UT on the 18th, the ACE spacecraft (which sits around a million miles upstream of the Earth in the solar wind), saw a simultaneous increase in magnetic field strength, solar wind speed, density and temperature. A classic signature of a CME as detailed by Christian on this very blog a few days ago. You need to add on about an hour to these times to get the arrival time at Earth.
The ACE data for the 18th looked like this;
The heliospheric imagers on the STEREO spacecraft ‘see’ a solar storm by imaging sunlight that is scattered off the hot cloud of plasma. The more particles there are in the solar wind, the brighter it looks in our cameras. The leading edge of the storms we observe therefore corresponds with the sudden jump in solar wind density seen by ACE at around 01:00.
We are currently asking you to scale the middle of the storm trace in ‘Trace It – Incomming’ (keep doing this, it’s the best thing to do!) as it is the easiest part to track. We need to get a few more storms under our belt before we know exactly how much lead time we need to add to improve our predictions but it’s looking like 7-12 hours based on two events.
So, will this storm cause any auroral activity? Well I’d be surprised if there wasn’t anything. The efficiency with which the solar wind can dump energy into our atmosphere depends on the orientation of the solar wind’s magnetic field with respect to the Earths. If it has a northward component (Bz is positive or greater than 0) then the solar wind and Earth’s fields are similar and, like laboratory magnets, like poles repel each other and not much interaction occurs (there may be some if the magnetic field is complex, I’m just talking about the idealised situation here). If there is a southward component however (Bz negative, less than 0) the two fields can connect (space scientist call this magnetic reconnection) allowing the solar wind to enter the Earth’s magnetic field over the north and south poles where it accelerates particles into the Earth’s atmosphere, exciting the gasses there which then glow – generating the aurora.
The longer and more extreme the period of southward Bz, the more likely there is to be auroral activity and the further equatorward this activity is likely to occur.
If you look back at the snapshot of data from the ACE spacecraft (top panel, red line), you will see that initially the storm had a northward component (+ve Bz) but that this subsequently swung southwards indicating that there may be auroral activity. Keep an eye on the real-time feed at;
if you want to know how conditions evolve from now on.
The skies above Oxfordshire are once again their uniform grey colour so if you have clear skies, let us know if you see any aurora in the next few days. Midnight is a good time to be looking since this is when you will be closest to the energetic particles being thrown into the upper atmosphere by the Earth’s magnetic tail as it snaps back after being stretched to breaking point by the solar wind.
Congratulations on a very successful prediction. Let’s hope we have a few more before the STEREO spacecraft move too far from the Earth. We want to hone our skills before we start making space weather predictions for Venus, Mars Jupiter and Saturn later in the mission, but more about that later… 😉
Thanks once again for all you time, efforts and enthusiasm,
Now that we’re trying to forecast Earth-directed solar storms, I’ve been attempting to explain how we know whether our prediction has been any good or not. This involves looking at data from a NASA spacecraft called ACE (the Advanced Composition Explorer) that sits a million km upstream of Earth and measures the Earth–directed solar wind as it blows by. The ACE data are presented as a series of wiggly lines that require an experienced eye to interpret so I was very pleased when one of the world’s experts in such data, Dr Christian Mostl from the University of Graz in Austria, agreed to give us all a lesson in how to understand what’s going on with the solar wind by studying these wiggles. So, without further delay, I will hand you over to Christian (a team can never have too many Chris’s – Chris);
What’s the solar wind doing at the moment?
Here is a little tutorial for you to understand better what is going on in the solar wind around Earth at this very moment in time. After going through this text, you will able to check for yourself in real time if the next prediction which we will issue with your help was excellent, good, bad or utterly horrific!
As pointed out earlier, this is the site to see the current space weather environment around Earth – this is one of our “weather stations” in space:
These strange, wiggly lines tell you the state of the solar wind for the last 7 days, observed by the Advanced Composition Explorer or ACE spacecraft, situated about 1.5 Million kilometers away from the Earth, in the sunward direction. So, first question: why are these lines so freakishly trembling? Answer: the solar wind is just a very turbulent medium. Even the “slow solar wind” (by definition around 400 kilometers per second – that’s over 1 million kilometers per hour!) is flowing away from the Sun so fast that its velocity always exceeds its own speed of sound by about a factor of 5. This makes it a constantly supersonic flow, and thus much, much more turbulent than, say, a wild river. By the way: why exactly the solar wind is so fast is one of the great mysteries of astrophysics!
Most of the time, the ACE plot will show you plain, normal, slow solar wind. I summarize here some of its parameters, which are shown in the plot, from top to bottom:
Bt (its total magnitude): around 5 nT, but always less than 10 nT.
Bz (its north-south component): between +/- 5 nT; be aware that this parameter is like the on/off button for Earth’s magnetosphere: the more southward or negative it is, the more energy will be transferred to the magnetosphere, resulting in auroras and magnetic storms. In contrast, for a positive or northward Bz, very little happens.
lets skip the next panel for simplicity, so lets go to…
N is the number density: between 1 and 10 protons/ccm. On the plot, there actually are two horizontal dashed lines at these levels, so you can easily spot outsiders. Don’t worry about data points which are below 1 p/ccm, this happens quite often. But be aware that this is a logarithmic scale, so the y-axis goes up all the way to 100, and intervals which appear slightly above 10 p/ccm can actually indicate much higher density!
V is the solar wind speed: between 300 and 450 km/s. Note that the solar wind flow points all the time almost perfectly in the direction away from the Sun.
T is the temperature: most of the time the protons have less than 100 000 K – again there is a horizontal line to guide the eye.
Now, if there is a slow solar wind, there must be a fast one too, right? Every few weeks on average, a gust of high speed solar wind, between 500 – 800 km/s, will hit Earth. They are strong enough to produce minor magnetic storms and are very good at creating beautiful auroras, but they usually lack the punch of a strong solar storm (a CME) to knock out any technological infrastructure. You can identify a high speed stream from these parameters:
Bt, Bz: a peak of 10-25 nT followed by values around 5 nT.
N: a peak of 10-20 p/ccm followed by low values of 1-5 p/ccm.
V: 500-800 km/s.
T: greater than 100 000 K, so hotter than slow solar wind.
Also, the magnetic field should be very wiggly throughout the interval where the speed is high.
Usually, N will peak first, followed by B, T, and V. This order is a consequence of a fast stream compressing the slow wind ahead of it.
Now, I have a little exercise for you: can you spot whats going on in the plot below?
The answer is: There is a long interval of slow solar wind followed by a high speed stream, starting on day 31. Great!
Finally, we come to the cherry on the cake: An “interplanetary coronal mass ejection” or ICME is the name given to a solar storm or coronal mass ejection (CME) you have been tracking in Solar Stormwatch when it is observed directly, on-spot or in situ by a spacecraft in the solar wind like ACE. ICME parameters are in principle elevated compared to the slow solar wind, but their signatures can vary greatly from one to another. Also, there are different regions inside an ICME. So to make it as simple as possible, watch out for intervals where B, N, V and T are much higher than normal, and yes – this can be difficult to distinguish from a high speed stream. Here is an example:
On day 06 there is a very abrupt and strong upward jump in all the parameters – this signals the arrival of the shock wave driven by the CME! That’s the time which should be compared to the forecast which was issued by Solar Stormwatch with your help.
While not well seen in the example above, there can be intervals following the shock wave, where the Bz is very smoothly changing and T is very low. With this we can identify an ICME with certainty, and this part of it is called a “magnetic cloud”. It signals that a CME hit Earth’s magnetic field head on! Most scientists think of “magnetic clouds” being at the core of many, if not all, ICMEs, and there the magnetic fields are the strongest, giving rise to the strongest geomagnetic storms.
So, now you should be able to spot for yourself if a solar storm has just hit Earth. But beware! Often it is easy to identify a shock wave driven by the ICME by very abrupt and strong upward jumps in all the parameters, but to see if a magnetic cloud has hit Earth you will have to wait for sometimes up to at least another day. Slow ICMEs do not drive a shock and can often be identified with certainty only in retrospect.
Has it really been over four years since I watched the STEREO spacecraft rocket into the sky over Florida? The two spacecraft used lunar swing-bys to put them into Earth-like orbits, one drifting away ahead of the Earth and one behind, each retreating from the Earth at an angle (with respect to the Sun) of 22.5 degrees per year.
On Sunday February 6th 2011, just after 17:08 GMT the two spacecraft will have drifted to the point where they are on exactly opposite sides of the Sun from each other. This is a momentous moment as it will be the first time we have been able to see the entire Sun. All very interesting you may think, but why is that important? Well, it is true that the the Sun rotates just like a planet, taking around 27 days to complete one rotation so we could just wait for it to roll past. Unlike a planet however, the Sun is a continuously churning magnetic fluid that rotates at different rates at different locations. The up-shot of all this motion is that the magnetic fields get stretched, tangled and knotted, causing the vast eruptions that solar stormwatch has been designed to study. These magnetic fields connect different regions of the Sun and, while we have been able to image the Earth side of the Sun since the start of the space-age, we have never been able to image changes on the far side that may trigger eruptions towards Earth.
While the two STEREO spacecraft will image the whole sun for a fleeting moment, as they continue on their paths towards the far side of the Sun from the Earth, Earth-orbiting spacecraft like the Solar Dynamic Observatory will fill in the gap and allow at least 8 years of observations of the entire Sun.
It’s going to be a fascinating time and, by participating in Solar Stormwatch, you will be helping us to understand the complex and mysterious life of the Sun, which in turn will help us to understand the many millions of stars that adorn the night’s sky.
Thanks again for all your time, effort and enthusiasm.
Thanks to everyone for their efforts in tracking our first real-time Earth-directed solar storm. We have been very impressed by the care with which you have all been analysing our data. So, what happened? Well, from your clicks, we predicted a storm would reach Earth around 07:00 on the 13th of December. Meantime the official spaceweather forecast issued by NOAA gave the following warning;
Space Weather Message Code: WARK04
Serial Number: 1681
Issue Time: 2010 Dec 12 1708 UTC
WARNING: Geomagnetic K-index of 4 expected
Valid From: 2010 Dec 12 1710 UTC
Valid To: 2010 Dec 13 0700 UTC
Warning Condition: Onset
The K-index is a measure of the amount of disturbance in the Earth’s magnetic field (as it is buffeted by the solar wind). A little later NOAA sent confirmation that the Earth’s field was indeed being affected;
Space Weather Message Code: ALTK04
Serial Number: 1488
Issue Time: 2010 Dec 12 1742 UTC
ALERT: Geomagnetic K-index of 4
Threshold Reached: 2010 Dec 12 1741 UTC
Synoptic Period: 1500-1800 UTC
Active Warning: Yes
(I’d encourage you to keep an eye on the professional forecast issued by NOAA at http://www.spaceweather.gov)
So, it seems that something arrived at Earth around 14 hours before our prediction. The effect wasn’t very big but was this disturbance caused by the same storm that we were tracking? I think so, and I’d like to talk you through why I think that.
It is true to say that not much happened at Earth as a result of this storm but just imagine standing in front of someone who was about to throw a baseball at you from a distance. The difference in angle that the pitcher would have to throw that ball between it sailing harmlessly by and alternatively hitting you on the nose is very small but the two scenarios would have dramatically different consequences! OK so this analogy breaks down when you consider that the storm expands as it travels out from the Sun and is very large by the time it reaches the distance of Earth but the principle stands. It doesn’t have to miss by much for it to have little or no effect at all. To stretch this analogy a bit further, I think this storm just brushed by our ear! It’s also difficult to tell ahead of time just how effective a storm will be (in my now horribly tortured analogy, we don’t know whether it’s a baseball or a beachball – OK, that poor analogy has suffered enough, I’ll stop hurting it now).
The reason I think our timing was out for this event was that we ask you to scale the middle of the features in the j-maps (keep doing this, it’s the best way!) so this would mean we are not tracking the absolute front of the storm (which would arrive at Earth a little earlier). Once we have a few predictions under our belt, we’ll know what this offset is (and 12 hours is not a bad estimate from the experience we have so far) and will compensate accordingly.
But how do we know that there wasn’t much of an effect at Earth? Well, one way is to look at data from the Advanced Composition Explorer (ACE) spacecraft. This sits about a million miles upstream from Earth in the solar wind and ‘tastes’ the wind as it goes by. A solar storm usually contains a giant bubble of magnetic field that sweeps up material ahead of it while material from the solar atmosphere expands into the void behind. ACE has many detectors but a useful summary of solar wind conditions over the last seven days can be found at;
This is a rolling summary of solar wind conditions just upstream of the Earth. These plots are a somewhat confusing collection of wiggly lines that, between them, can tell us if anything unusual is happening. The top panel contains the total magnetic field strength (Bt) and the size of the component that is aligned to the Earth’s magnetic field (Bz). If a magnetic bubble goes by, we’d expect Bt to go up and if Bz was negative at the time (southward) it would mean that the solar storm and Earth’s magnetic fields were aligned in opposite directions – ideal for the two fields to merge and let the solar wind into the Earth’s atmosphere to cause an aurora. On the afternoon of the 12th, the total field does indeed increase in strength but the Bz component is positive for the most part, meaning that the Earth and solar wind magnetic fields were aligned – remember that two aligned magnets repel each other while opposites attract. Not surprising then that there wasn’t much activity.
As a storm goes by ACE, we’d also expect a sudden increase in the solar wind speed and density as a storm front passes (two of the lower panels are marked speed and density). While the speed and density did increase, they didn’t go up by very much and there were no sudden jumps indicating that the storm was not travelling fast enough to generate a shock ahead of it.
Another way of investigating whether the Earth’s magnetic field was rattled by the event is to look at measurements by the aurorawatch team at;
Their page shows the current state of the Earth’s magnetic field over northern Europe. When a solar storm hits Earth, the very least it does is cause the Earth’s field to be compressed and wobble. This shows up on the ground by very small compass movements. The aurorawatch team run a series of sensitive magnetic field monitors (called magnetometers) and present their data as a series of graphs. Basically, a flat line means a quiet magnetic field and a wiggly line means that the Earth’s field is being buffeted by the solar wind. On the day of our prediction the aurorawatch plots showed little or no disturbance to the Earth’s magnetic field over Europe (in contrast to the activity seen in Boulder, Colorado).
So, while this was not exactly ‘the perfect storm’, I think it did enough to show that our techniques work and that you are all doing a fantastic job in your careful analysis of the STEREO HI data.
In my geekier moments (which occur far too often to be healthy) I like to think that the pioneering space-weather forecasters of today are not only helping us to protect our technological infrastructure here on Earth but are also blazing the trail for the future exploration of the solar-system. Your efforts with solar stormwatch are an important part of this, and I’m really grateful for your time, careful analysis and enthusiasm in helping us with our work.
Tin hats on everyone! Stormwatchers have been tracking a solar storm launched from the Sun on the 14th October. Analysis of your results shows that this storm, travelling at 244 km per second, will cross Earth’s orbit at 14:43 GMT on the 21st October just 37 degrees ahead of our planet. While this is not expected to be a direct hit, we may still suffer a glancing blow so look out for auroral displays around this time. Stormwatchers have enabled this forecast by participating in the ‘Incoming -Trace it’ activity and have been doing such a great job that we can now start to issue warnings ahead of their arrival at Earth. Thanks to everyone for your hard work and patience. The more people we have looking out for and tracking these storms, the better our forecasts will be, so tell your friends!
Space Weather News for August 1, 2010
GLOBAL ERUPTION: During the early hours of August 1st, NASA’s Solar
Dynamics Observatory recorded a complex global disturbance on the
Earth-facing side of the sun. Most of the sun’s northern hemisphere was
involved in the event, which included a long-duration C3-class solar
flare, a “solar tsunami,” and a massive filament eruption. As a result
of these blasts, a coronal mass ejection (CME) is heading toward Earth.
High-latitude geomagnetic storms and auroras are possible when the cloud
arrives a few days hence. Check http://spaceweather.com for movies and
Thanks to Neel for his ‘life as a scientist’ piece. While Neel is busy writing up his results, Stormwatch is gearing up for the next release of activities with which we hope to be analysing each storm in much greater detail. As we accumulate data from all your dedicated analysis, we are starting to get some preliminary science results that, at some point, when we are confident that we are interpreting them correctly, we will want to share with you so that you can see the outcome of your hard work. the question is, how would you like us to do this? Obviously there is the blog and the forum but are there any other ways you would like us to communicate the science with you?
Meanwhile, Stormwatch continues to attract media attention. This month’s (June) Sky at Night magazine contains an article summarising several astronomy projects that rely on the input from many members such as Galaxy Zoo, Moon Zoo and Solar Stormwatch (the cover disk also contains the April edition of the TV programme which is all about the Sun).
Those of you who have been looking at the real-time data will know that the Sun appears to be waking up (slowly mind you!) and this means that it is becoming increasingly likely that a CME will occur. Keep up the good work, and look out for new activities in Solar Stormwatch soon.