Montana State University

MSU Solar Physics Group

Montana State University

Coronal Mass Ejection Prediction Page

Frequently Asked Questions

Back to the Main Page

Answers to some frequently asked questions.

Q: How much matter is contained in a typical CME ?

A: Typical values are 1 to 10 billion tons of material (1016 to 1017 grams), in the form of a large cloud of charged particles.

Q: How fast do the CMEs travel ?

A: A coronal mass ejection can make the 93-million-mile journey to Earth in just three to four days. This implies an average speed of about one million miles per hour.

Q: Do all CMEs affect the Earth ?

A: No. Because the Sun can eject matter in any direction, only some of the CMEs will actually be directed towards Earth. (See the next question regarding "CMEs" versus "halo CMEs".) But even among those CMEs which are Earth-directed, some will affect Earth more than others. The different qualities of a CME which can determine its "geo-effectiveness" include the amount of material ejected, the speed at which it travels, and the strength and direction of the magnetic field carried by the cloud of charged particles.

Q: What's the difference between a "CME" and a "Halo CME" ?

A: The difference is primarily a matter of perspective. Coronal mass ejection (CME) is the name given to an ejection of a large amount of matter from the Sun's outer atmosphere. These ejections typically comprise millions of tons of material in the form of charged particles, and can be seen because the material reflects sunlight. When one of these ejections is directed towards the Earth (or conversely, directly away from the Earth), it looks like a roughly circular "halo" surrounding the Sun. The "halo CMEs" then are those CMEs which are more likely to impact the Earth than those which are shot out at right angles to the Earth-Sun line.

Q: How often do halo CMEs occur?

A: That depends on the phase of the solar cycle. The Sun's activity is not the same year after year, but waxes and wanes in a cycle that lasts about eleven years on average. For example, the years 1990-1991, and 2000-2001 saw a large amount of activity (solar maximum: many flares and CMEs), whereas 1995-1996 was a time of minimal activity.

According to researchers using data from the SOHO solar observatory satellite, 25 Earth-directed halo CMEs were detected during the last eight months of 1997, and now in 2007 we again are in a time of low activity. But the fact that the Sun's activity is cyclic means that in the next 1-2 years we should expect to see CMEs becoming more frequent.

Q: Who is more likely to be affected by the impact of a CME with the Earth ?

A: A coronal mass ejection can affect various systems in different ways. In extreme cases, electrical currents can be induced in long metal structures, like power lines and oil/gas pipelines. Additionally, the high-speed charged particles of the CME can cause the buildup of electrical charge in metal structures in satellites; such buildups of electrical charge can be damaging to the sensitive electronic systems common in telecommunications satellites. Because of the possibility of damage to these vital and very expensive systems, advance knowledge of the likelihood of a CME is potentially helpful to the operators of such systems.

However, not all the effects of CMEs are bad. The beautiful aurora borealis and aurora australis (Northern and Southern Lights) are caused by the influence of CMEs with the magnetic field surrounding the Earth. Roughly speaking, this interaction with the Earth's magnetosphere forms a kind of magnetohydrodynamic dynamo -- like an electrical generator -- which can drive a flow of charged particles down into the atmosphere. A small part of the energy carried by this current of particles is dissipated in the generation of auroral light. (Most of the dissipated energy goes to heating the atmosphere, and to breaking apart the atoms and molecules in the atmosphere.)

Q: Why are the patterns S-shaped (and inverse S)? What properties of the magnetic loops in the solar corona make groups of them form an S?

A: The patterns are S (and inverse S) shaped because the magnetic field lines that comprise them are twisted, rather like a slinky, but not so densely coiled. The field lines are three-dimensional, but they look like an S, in projection.

Q: How big are these S-shaped regions? Bigger than your average sunspot? How long do they typically last -- days, weeks?

A: Sigmoids are typically 50,000 to 100,000 miles long, but they vary widely in size. They are easily 10 to 20 times bigger than an average sunspot. The Sun rotates roughly once a month, and we have seen regions come back month after month with the same sense of twist, in the most extreme cases.

Q: How long do coronal mass ejection events last? Is it a single burst of particles ejected from the Sun? Or does the event last for some time, like minutes, hours or even days?

A: Coronal mass ejections take off on the scale of hours, build up their speed to about a million miles an hour. The structure that passes the Earth does so in hours, just because it is a cloud that is fairly thick in its extent in the Sun-Earth dimension.. The event that it causes at Earth last longer, because it sets forces into motion at Earth that have their own timescale.

Q: Just to confirm that I'm understanding what you've written... the sigmoids appear in the corona, not in the photosphere (where sunspots are seen). Right? Is there any link between sunspots and the S-shaped patterns?

A: Yes, sigmoids are observed in X-rays, which are emitted by plasma that is at temperatures of one to several million Kelvins. These are coronal temperatures -- the photosphere, where sunspots are seen, has a temperature of only 6000 Kelvins.

As for the link between sunspots and sigmoids -- there is one. At times we see whorls in sunspots, and we know that they are related to sigmoids. We have shown that they imply the same sense of the slinky-type structure that makes the sigmoid -- right-handed versus left-handed, S versus inverse-S.

Q: If we know a CME is coming, what can we do about it? Turn off satellites so there's less of an electrical problem? Monitor power grids more closely when the CME hits, to reduce the chance of a massive failure like hit Hydro-Quebec on March 13, 1989?

A: You can turn off high voltages on your satellite, so it won't zap itself under the severe conditions when the plasma cloud hits it, then turn it back on when the danger passes. Several satellites do this during major meteor showers, for example. Better to lose a little data than lose a satellite.

Pretty much the same answer applies to transformers in power systems too. Better to shut off some part of the power grid briefly when there is a surge of current coming -- which you can't do without planning ahead, if half your crew is on vacation -- than lose major components.

Q: Earlier papers about the sigmoid structures have implied that if you see one, there's already an eruption of some kind in progress. Are you saying that you see these structures before an eruption of any kind takes place? How long do you see them before they erupt.. hours? days?

A: Yes, that's exactly what we've found. We have found that regions that are sigmoidal typically stay that way for their lifetimes, which range from weeks to months. So what your're asking is really equivalent to "How long is it between eruptions of a sigmoidal region?" We have not yet done a careful study of that question, but our impression is that the answer will be a few days, in the case of the regions that erupt most frequently. We cannot recognize a typical region as sigmoidal when it first appears on the Eastern limb of the Sun, when we are looking at it edge-on. However, by the time it gets within 45 degrees of the central meridian, its sigmoidal structure is clearly visible.

Q: You mentioned that in extreme cases, you've seen sigmoids in the same region come back month after month. Is this the same sigmoid, or is it something that appears, disappears then another one re-appears in the same place? What is a more typical value for how long each sigmoid lasts?

A: The Sun's corona is completely filled with strands of magnetic flux, only some of which are heated enough, at any given moment, to become bright in X-rays. In the simplest cases, all these strands are sigmoidal, so even though Mother Nature may choose to light up a different one every few minutes or few hours, they all give us the same message. If a person has wavy hair, you can get that message by studying one strand, or several, or even a very large number.

Contact Information

For more information, please contact

Thanks to Barbara Thompson, Simon Plunkett, Tim Slater, and Sally Stephens, who helped compile some of these.

Back to the MSU Solar Physics Press Page