The fates of large mass stars are quite different from those of the low mass ones. At first glance you might think that with more mass they will live longer - but no, they are just fuel guzzlers. As such, massive stars (O and B types on the Main Sequence) will have very short lives. To get a good idea of how a large mass star dies, let's look at what a really big one goes through.
Very massive stars are also very luminous, so they tend to have very strong stellar winds (since this is linked to their energy outputs). Due to this they may have a lot of gas around them that was blown off long ago. Sometimes it is easy to see these gas clouds, sometimes not. Like other mass ejection things we have already seen, the material can be spewed out in several ways, with the most common being bipolar outflow and rings. Figure 1 shows several different stages in the life cycle of stars, all in one convenient location.
Figure 1. A picture from the Hubble Space telescope showing stars in various stages of their lives. The lower right region (1) is a gas cloud from which stars form. Near the center (2) is a group of main sequence stars. Up and to the left of those is a large star (3) in the process of dying. The massive star has already ejected a ring of material as well as material out from it in a bipolar direction. In a way this image is just like a family portrait showing the oldest and youngest members of a family. Image credit: Wolfgang Brandner (JPL/IPAC), Eva K. Grebel (Univ. Washington), You-Hua Chu (Univ. Illinois Urbana-Champaign), and NASA.
Sometimes the outflow is even more energetic and at times confusing
to astronomers. Some very massive O-type stars have such strong winds
that they can completely screw up their evolutions - they lose so much
mass that you have to take that into account in your calculations when trying to figure out
how these stars will live their lives. Due to this uncertainty, we have
a rather hard time predicting what a very massive star will do or even
trying to figure out what one of them may have done in the past. Images
of two such objects are shown in Figure 2.
A recently observed massive star (with the exciting name of WOH G64)
has lost so much material that a thick, dusty ring has formed around
it. This star is very far away, in another galaxy in fact, so it
is not possible to easily see the ring, but its presence is revealed
by spectra and other instruments. You can learn about this massive
Even though it has lost a bunch of material, WOH G64 is still about 1500 times
wider than the Sun.
|Figure 2. Two very massive stars with very extreme mass loss episodes. On the left is the star WR124, a massive star ejecting material out at speeds of 100,000 miles/hr. Each blob of gas that it ejects out has a mass of more than 30 times that of the Earth's. To the right is the unusual star Eta Carina. Actually, the star is buried in the center of the two bubbles, which are thought to be from an eruption that occurred in 1847. Follow this link to see how the bubbles formed. During this outburst the two bubbles were ejected as well as a disk of material that can be seen between the bubbles. The speed of the material in this case is on the order of 1.5 million miles/hr. Image credits: Yves Grosdidier (University of Montreal and Observatoire de Strasbourg), Anthony Moffat (Universitie de Montreal), Gilles Joncas (Universite Laval), Agnes Acker (Observatoire de Strasbourg), Jon Morse (University of Colorado), and NASA.|
All of these big time mass loss episodes are just a sneak preview of what is to eventually happen to these stars. These stars are more massive than the cut off for going into Planetary Nebula stages, so the death scenarios of the massive stars go down very different paths. Just how big can a star be? We're not exactly sure, but there is a limit to how much material can come together to form a single star, and conservative estimates put that at around 150 solar masses. However in 2010 astronomers from the Very Large Telescope in Chile announced the discovery of a star that may have a mass that is currently 265 times the mass of the Sun. What is really amazing about this star, with the cute name R136a1, is that it started out with a mass of about 320 solar masses! So it probably had some serious mass loss episodes in its life. In case you were wondering, this star is not even in our galaxy, but is in a neighboring galaxy.
We'll look at the life cycle of a 25 Solar Mass star to see what happens to one of these big beasts. It is more massive, so it can go through more burning stages than a low mass star. It can ignite the more massive elements due to the greater gravitational heating in the core (more mass means more gravity). Each burning stage takes less time. The data in the accompanying table is from Chieffi, Limongi, and Straniero (1998) based upon their computer model for such a star.
|Fusion Process||Main Fusion Products||Duration of Fusion Process|
|H||He||6 million years|
|He||C, O||700,000 years|
|C||Ne, O||1000 years|
|O||S, Si, Ar||4 Months|
|Si||Fe, Cr||1 day|
When the star is burning hydrogen, it is on the Main Sequence. Every other burning stage after that has the star off the Main sequence and in the area of the H-R diagram populated by Supergiants. The stars just evolve through these various burning stages while they are Supergiants, sometimes blue supergiants, sometimes yellow supergiants, and sometimes red supergiants, so they will wander back and forth across the top of the H-R diagram during these later stages. If you click here you can see some evolutionary paths of such stars. The region that the supergiants inhabit isn't a clearly defined location on the H-R diagram, but is basically just the top part where the luminosity is very high.
Take a look at the numbers in the rightmost column. Why is each stage shorter than the previous stage? There are two reasons
|Figure 2.The structure of the interior of a supergiant star when it finishes silicon fusion. The very center has the inert iron core, and it is surrounded by thin layers of fusion shells. The core size is actually exaggerated a bit here. It is actually only 1/1000 the radius of the star (about the radius of our Sun), but it contains about 1/3 of the star's mass (around 8 solar masses for a 25 solar mass star). Note how the radius of the star stretches out to the size of Jupiter's orbit. Obviously we wouldn't want a star like this in our neighborhood.||Figure 3. Betelgeuse - a truly giant supergiant. The shoulder star of Orion is so large that its size can actually be seen in the Hubble Space Telescope (most stars are so far away that they always look like dots unless they have a really huge diameter, like this one). Image credit: Andrea Dupree (Harvard-Smithsonian CfA), Ronald Gilliland (STScI), NASA and ESA.|
So now these dying massive stars will be seen as either Red Supergiants or Blue Supergiants, depending upon how hot or cool they are. Due to their very large radius they also tend to be extremely luminous. How large can the radii get? In 2005, astronomers discovered several red supergiants with radii that were much larger than that of Betelgeuse (shown above) - these stars had radii that are about 1500 times that of the Sun! To see how big these stars actually are, just take a look here. Obviously with the outer layers so stretched out, the fusion going on deep in the core is not going to be visible to anyone. Even though stars like this look like they are in the last stages of their lives, we can only see what is going on at the surface, not what is actually happening in the core. Of course the core is the interesting part!
Eventually the core of the massive star will look like a giant onion, with the densest material in the middle and the lowest density stuff on the top. Each shell of the onion will have some small amount of fusion still going on, but the energy that is being produced at this point is pretty pathetic.
All sorts of elements have been burning, and now we come to the last element to burn, iron (Fe). Does it burn? It sure does, but at a cost. Whereas previous fusion processes released energy, iron burning consumes energy. The energy that should go into holding up the star instead goes into burning the iron. Is this a problem? You bet your buttonhole it is! The iron fusion consumes energy, so there isn't enough energy to help support the star. Without support, gravity comes along and squeezes down the core. What happens when you squeeze stuff? It gets hotter. The iron core gets hotter and starts burning faster, which causes more energy to be sucked away, which removes more support against gravity, which causes the core to compress more and heat up more, which causes the iron to burn faster, which... I think you get the picture.
The core of the star collapses during the iron burning stage (since nothing is fighting the gravity), which takes only about 1/4 second. The collapsing process shrinks the core down to the size of the Earth. It gets very dense, up to the point of electron degeneracy (remember, that's what a white dwarf is like). Will the collapse stop? No, since the core is more massive than the Chandrasekhar limit (the iron core is about 1.5 solar masses in this case). The mass of the core is too much for the Chandrasekhar limit, so electron degeneracy will not stop the collapse - it will keep going. Gravity keeps crushing the star down until it reaches the point where the pieces of atoms are crushed together. This is not an easy thing to do, but as the protons (p+) and electrons (e-) are slammed together they form neutrons (n) and neutrinos (), which can be written out as
p+ + e - n +
This reaction makes sense, because you combine a positive and a negative together and just get neutral stuff out in the end. All of this compression and atomic mushing results in the core of the star ending up as a big ball of neutrons. At this point the collapse can be stopped by neutron degeneracy (1014 g/cc is the density of such material). Neutron degeneracy is much more extreme than electron degeneracy - greater density, more extreme rules of physics and so forth. Because of the degeneracy, the star will not get any denser so long as the neutron degeneracy can hold it up. In the case where the degeneracy can't hold it up, you end up with a black hole - more on this later.
The core that ends up as a ball of neutron degenerate material is called a Neutron Star. This is a star so small and compact that a 1.5 solar mass neutron star would be only about 20 km in size. Think of that - an object more massive than the Sun only the size of a large city! It would be incredibly dense, 1014 gm/cc. This is sort like the density you'd get if you took 100 aircraft carriers and crush them down to the size of a sugar cube. Try putting that in a cup of tea!
In about 1/4 of a second the core has been crushed down, resulting in a neutron star, which are in most cases only a few solar masses in size. What happens to the rest of the star? Remember, the neutron star core at this point is only a small part of the total mass, so you still have quite a few solar masses to watch out for that is located beyond the star's core.
The core collapsed very quickly from a size close to that of the Sun's to only about 20 km, so there is a gap in the support of the rest of the star. What support? - there is NO SUPPORT! Nothing is holding up the rest of the star. This is sort of like when the Coyote runs off a cliff and doesn't immediately fall down - at least not until he realizes that he is off the cliff. The outer layers of the star don't really know that they have had their legs cut out from under them for a moment, but once they do - watch out. The upper layers will fall onto the ultra dense neutron degenerate core and the material will heat up to about 5 billion degrees. This high temperature and the corresponding high pressure will generate an incredible amount of energy. The energy that is generated by the slamming of the outer layers on the core is huge. This energy that is produced here in this small interval of time is the same amount as that given off by the Sun over its entire lifetime (10 billion years). This huge bottled up energy is released in a massive explosion that will blow off the outer layers - basically, the star explodes. That's how you produce a Supernova. This little animation shows a blue supergiant quickly collapsing down and then exploding as a supernova.
Massive stars are pretty rare, so on average there is only one supernova occurring in a galaxy every century.
Now I'm going to complicate matters a bit. There are actually two main types of supernovae. One is the type that I just described; the other occurs when a white dwarf is near its mass limit (the good old Chandrasekhar limit=1.4 solar masses) and is pushed over this limit when too much mass is dumped on it, usually in a binary system. You may want to refresh your memory on the stuff about novae in the previous set of notes. If the white dwarf star in the binary system is really big (close to 1.4 solar masses), rather than just going nova when mass is dumped on it, it will be too massive to hold itself up and may instead become a supernova. There are two vastly different kinds of supernovae - and they have to be distinguishable, and they are, mainly because the object that explodes in each case is very different (massive star versus a white dwarf). To distinguish between the two types, the following designations are given
Figure 4. The light curves of the different types of supernovae are shown - note that the Type Ia supernovae are brighter. Also, the rate at which they brighten and fade away is different. This helps astronomers distinguish the two types. This graph uses absolute magnitude as the brightness scale since the corresponding luminosity values are around a few billion solar luminosities.
Figure 5. The spectra of the different types of supernovae are shown. The Type Ia supernova has absorption (the valleys) and emission features (the peaks) associated with heavy elements, while the Type II has only hydrogen (H and H) showing up prominently in its spectrum. Spectra were obtained from the spectra archive at the University of Oklahoma.
In general Type Ia supernovae are brighter by about two magnitudes or so than the Type II. You would have to know the distance to the supernova if you want to use the brightness as a way of categorizing it, so that isn't a good method. Fortunately it is possible to tell the two supernovae apart by looking at their spectra. The object that goes supernova is quite different in each case, so the spectrum from each type of supernova is distinct. A white dwarf is the burned out core of a dead star, so it is made mainly of stuff like carbon, oxygen, nitrogen, etc., but not a lot of hydrogen. A massive star, on the other hand, is still mainly made of hydrogen, so when it explodes its spectrum will be full of hydrogen. It is pretty easy to distinguish the two types of supernovae without knowing their distances. You'll notice that if you do know which type of supernova you have, you can then use its typical maximum brightness (from Figure 4 above) and the apparent magnitude that you observe to get its distance.
One very unusual Ia supernova is 2006gz. This has a spectrum just like a 'regular' Ia, but it was brighter than normal. It also had too much carbon and silicon, which led some astronomers to speculate that this was actually a collision of two white dwarfs. It has also been proposed that 2006gz came from a super-Chandrasekhar limit white dwarf - an abnormally large white dwarf. So far we do not know the answer to this unusual question, but it does show that sometimes very strange things do happen. There was also a recent study by the Chandra observatory that seemed to indicate that most type Ia supernovae are actually produced by the merger of two white dwarfs, rather than a single star. So until more evidence is found (because that's how science is done), we'll just have to use the general phrase "white dwarf going over the mass limit" - which could occur due to very different reasons.
You should be asking yourself Why is it 'Ia'? Why not just call it a 'I'?. Good question. A type I supernova has very little hydrogen in the spectra, but sometimes it is not because the star that exploded was a white dwarf. There are two other type I supernova, with the amazingly original names of Ib and Ic. Both are thought to come from large mass stars that have blown away most of their outer layers so there is not much hydrogen left in their spectra. The main difference between Ib and Ic is whether there is any helium left - a Ic has no helium in their spectrum, while a Ib still has some helium. Both the Ib and Ic are fainter than a Ia, and a bit rare since they only come from very massive stars. Unlike the massive stars involved in the type II supernova, these have lost too much of their mass to have a big blast. Generally when we talk about type I supernovae, we are referring to type Ia.
Recently astronomers have found that there are some supernovae that could be better described as super-supernovae - but that's sort of a silly name. The term hypernovae has been proposed for these extreme explosions. There is a bit of debate as to exactly what happens during a hypernova, since this is a relatively new idea and most of them have only been observed to occur at very great distances. One option is that a very massive star (more than 30 solar masses, possibly up to 150 solar masses) that had been previously losing mass eventually collapses in on itself, causes a massive explosion and eventually forms a black hole. Another option is the merger of unusual stars, such as two neutron stars, which results in a massive explosion (you'll learn more about neutron stars in the next set of notes). One side effect of such an event is the emission of a large amount of gamma-rays, which aren't normally observed from stars. Now a days such gamma-ray bursts can be spotted with the Swift telescope, and several possible hypernova have been discovered in that way - though it wasn't until after we look at them with other telescopes, light visible light or x-ray telescopes are we certain about their overall energy output. The overall energy of a hypernovae is generally 100 times greater than the energy given off by "normal" supernovae. Currently the record holder for the most powerful stellar explosion is the object known as SN2006gy (yes, that is a lame name), which was observed in 2006 and had an energy output that was greater than any other supernova. Here is a comparison of this objects brightness over time compared to other "normal" supernova. This is similar to Figure 4 shown above. Here is an animation showing the explosion based upon the x-ray and gamma-ray data from the region following the explosion. It is interesting that there are two large bubbles of material, which were given off by SN2006gy before the explosion occurred. This is very similar to what we see today in Eta Carina (Figure 2), an object in our own galaxy which might someday erupt as a hypernova. Several other stars are thought to have also had smaller outbursts before they blew themselves to pieces as supernova - sort of a hiccup before a really big belch! And in case you were wondering how you missed such an amazing explosion in 2006, don't worry, not too many people noticed it since SN2006gy occurred in a galaxy that is about 240 million light years away, so it was barely visible, except with the largest telescopes.
While most gamma-ray bursts are invisible to us (since gamma-rays cannot reach the surface of the Earth), it is still possible to "see" them, since the total energy given off is huge. Gamma-ray bursts give off light at all wavelengths, not just gamma-rays (remember hot objects give off light at many wavelengths even though they have only one peak for their emission). So when the Swift satellite observes a burst, the location is transmitted to regular ground based telescopes, both visible light and radio, in an effort to measure the light output in as many wavelengths as possible. In March 2008, there was a gamma-ray burst that was so powerful and concentrated that it could have been seen by the naked eye for about 15 seconds. That's pretty impressive considering that the object that produced the burst was about 7.5 billion light years away!
Supernovae are sort of rare, so astronomers are only able to observe them occurring in distant galaxies. Currently about 200 or more supernovae are observed each year in other galaxies - in 2003, nearly 330 supernovae were discovered, some nearly a billion light-years away. While this may seem okay, the problem is that since these are very distant supernova, there isn't a lot of detail visible. For the most part, astronomers can figure out the type of supernova in a distant galaxy by obtaining a spectrum, but that's about all. Astronomers over the centuries have seen supernovae, some in our own galaxy, though in the old days they didn't know what they were. By looking at the records of various cultures, astronomers have figured out that some unusual astronomical events that mystified people in the past were actually supernovae. Here are some of the more famous ones -
Kepler's supernova (as it is sometimes called) was the last supernova that was observed to go off in our galaxy. We haven't seen a star become a supernova in our galaxy in about 400 years - we're long overdue for one! Of course, it is possible that other supernovae have occurred in our galaxy since Kepler's, but they may have happened in distant parts of our galaxy, and we couldn't see them. As you'll see there is good suspicion that this is the case.
While it is difficult to see a supernova in great detail today (since most are so far away), it is sometimes easy to find the material left over from the explosion. This is because there is not only a lot of material, but it stays relatively hot for quite a long time. The gas cloud left over from the supernova explosion is known as a Supernova Remnant (SNR). Like an explosion in a fireworks display, it takes a long time for the cloud to fade away, though in the case of a SNR, the cloud can hang around for thousands of years.
|Figure 6. Various Supernova Remnants are shown. To the left is the Crab Nebula, the remnant from the Supernova observed by the Eastern astronomers in 1054 A. D. This visible light image is from the Very Large Telescope. In the center is the Cygnus loop, as seen by an x-ray telescope. This is a very old remnant, and it is also very large. The line in the lower right part of the picture is the width of the Full Moon. On the right is an x-ray image of the remnant that was Tycho's supernova (observed in 1572 A. D.). If you click on the Crab or Tycho images, you'll be able to see the Chandra x-ray image of those remnants. (Image credits: VLT, ROSAT, MPE, NASA, NASA/CXC/ASU/J. Hester et al., NASA/CXC/Rutgers/J. Warren & J. Hughes et al.).|
These aren't just neat things to observe; they can provide useful information about supernovae. This is done by measuring the velocity of the gas and combining that with the size of the object. Doing this provides astronomers with the age of the supernova (since velocity = size / time since explosion). Even though we might not have seen the explosion, we can estimate when it occurred. Many SNRs seen today are still rather hot even though they may have "gone off" hundreds of years ago. They often have x-ray emissions - which tells us that they are still really hot. The supernovae described earlier, like the one in observed in 1054 and those seen by Kepler and Tycho, all left behind large expanding gas clouds. The one seen in 1054 can be observed today as an object called the Crab Nebula (or, by its catalog name, M1).
|Figure 7. Cas A, a supernova remnant. The image to the left is from the Very Large Array radio telescope (NRAO). The image to the right is from the Chandra x-ray telescope (NASA/CXC/MIT/UMass Amherst/M.D.Stage et al). Click on either image to see a larger view.|
One rather confusing SNR is Cas A, a SNR in the constellation of Cassiopeia (that's where the "Cas" comes from in its name). This is a very strong radio source, as well as a strong x-ray source (see Figure 7). Obviously it is still rather hot. It is also relatively small. Combining this information with velocity and size data tells astronomers that this object blew up about 300 years ago. We haven't seen a supernova in our galaxy for about 400 years! Somehow, this thing blew up and no one noticed it! It also looks like we missed another supernova that was even more recent - only 150 or so years ago! Unfortunately it was in a rather messy part of the galaxy, so it would have been difficult to see even if we knew it was going on.
Another neat SNR is the Gum Nebula. This is a large gas cloud, and it is really large because it is really old. It is estimated that this object blew up sometime around 10,000 BC. Not only is it rather old, but it is also relatively nearby. Taking this into account, when this thing blew up, it would have been as bright as the Full Moon!
Let's see what we have got so far. Astronomers can study supernovae in other galaxies, but they are so far and faint that only the largest telescopes we have can see most of them. They can study supernova remnants to try to figure out what happened in the past when these things blew up. That's sort of boring, eh? It isn't all of the time. Actually, things in the astronomical community got rather crazy not too long ago, on February 23, 1987, to be exact. On that night there was the event of a lifetime - Supernova 1987A, or SN 1987A for short. Supernovae are named for the year (1987 in this case) and a letter for the order of their occurrence in the year. The first one of the year is A, the second is B, etc. In 1987, the first supernova seen was labeled SN 1987A.
Figure 8. Supernovae 1987A - image of the supernova at its brightest in the part of the sky it appeared in. In the upper left of the image is the Tarantula nebula, a large H II region. These objects are actually thousands of light-years away - around 160,000 light-years in fact. Image © Anglo-Australian Observatory, Photograph by David Malin.
What was the big deal about SN 1987A? This was the first naked eye visibility supernova observed since 1604 - you didn't need a telescope to see it; that's how bright it was! While it did not occur in our galaxy, but in one of our neighboring galaxies (the Large Magellanic Cloud is the galaxy it happened in), it was still close enough that detailed studies of it could be carried out. The supernova was discovered, or perhaps a better word is "noticed," by Ian Shelton on the night of Feb 23, 1987. It is quite likely that others saw it but they did not recognize it for what it was.
When SN1987A went off, virtually every telescope that could observe it was used to study it. This included telescopes located in countries such as Chile, South Africa, and Australia, mainly because the Large Magellanic Cloud and the supernova are very southern objects. Satellites were also used - including the ultraviolet satellite, IUE (which made over 600 observations) and a variety of rockets that obtained x-ray images. The Hubble Space telescope wasn't in orbit at the time, but it has been looking at it since.
The brightness variation of SN1987A over time. This chart shows
thousands of observations by both professional and amateur astronomers.
The time scale at the bottom is in days, but uses a rather strange
system. The entire graph covers a time corresponding to about 3.3
years. The magnitude scale is on the sides. Remember, the faintest that
you can see with your eye is magnitude 6, so SN1987A was visible to the
naked eye for about a year. Chart is courtesy of the American
Association of Variable Star Observers (AAVSO).
Another added advantage to this supernova over others that are studied is that we knew which object blew up. This area of the sky is fairly well studied, so many stellar surveys were done and the star in question, Sanduleak -69 202, was cataloged and its characteristics (color or temperature and luminosity) were known before it became a supernova. Why was that important? We knew what type of star exploded (temperature, luminosity, likely mass, composition, etc.), so information about it helped astronomers refine computer models of supernova events. The supernova was bright for a long time (visible to the naked eye for about a year), so its evolution was followed closely (and still is being followed) to see if our theories about supernova remnants are correct. The rate at which the supernova brightened and is fading away can also provide information for various computer models and can be compared to other supernovae.
Figure 10. SN1987A and the star that eventually went supernova, Sanduleak -69 202. It was a relatively bright star to begin with, but when it went supernova (as is seen on the left), it got very bright. Image © Anglo-Australian Observatory, Photograph by David Malin.
The spectrum of the explosion confirmed the heavy element production (elements heavier than iron). This was seen in the way that various elements would appear and then decay over time. This was exactly in line with the theories about the heavy element production - a nice example where the theories that have been around for a long time finally were supported by observations. Also, by measuring the expansion velocity, we can determine the distance to the Large Magellanic Cloud galaxy in a very accurate manner - something that is very difficult to do. Probably one of the more exciting discoveries was of the neutrinos from the supernova. All of those neutrino detectors around the world were set up to look for neutrinos from the Sun. On the day of the supernova, these detectors were practically flooded with neutrinos. By "flooded" I mean they detected maybe six or eight neutrinos, which is considered a flood when compared to the normal number of neutrinos that are detected. This was another case of theory and observations coming together!
When you look at the supernova today, you can see several rings of material around its location. This material is not part of the supernova explosion, but was blown off by the star years before (you may want to go back up to the top of the notes to look at those big stars that are currently blowing off mass, especially Eta Carina). It wasn't until the supernova went off that the ring actually did something - in this case the rings lit up, because the light (energy) from the supernova traveled through them and heated up the gases.
Figure 11. How the area around SN 1987A looks today. The supernova is in the center of the multi-ring structure. The inner ring is less than a light-year from the supernova, while the other rings are a few light-years away. The rings are not part of the supernova, but are made of material that was blown out from the star before it went supernova. To see why the rings have the orientation that they do, just click on this link. Image credit: P. Challis (CfA).
Today the supernova is too faint to see with the eye, but its evolution is still being followed, particularly what is going on with the rings around it. The distance from the supernova to the nearest ring is less than a light-year. The shock wave from the supernova started hitting the ring material in 1998, which is causing it to light up again. It will take some time for the shock wave to travel through the ring, and you can expect more neat pictures from the Hubble Telescope showing that. Here is the latest movie showing the ring lighting up due to the shock wave. Even though the supernova did happen quite some time ago, it is still one of the major events in modern astronomy.
|Figure 12. The inner ring around SN 1987A is starting to light up due to the collision of the shock wave from the supernova. The yellow areas in the picture on the right are where the collisions are currently occurring. Image credit: P. Challis and R. Kirshner (Harvard-Smithsonian Center for Astrophysics), P. Garnavich (University of Notre Dame) and Z. Levay (STScI).|