Astronomers have long assumed that supernovae are the source of at least most of the cosmic rays that hit Earth.
Woah, slow down... cosmic rays? Right, you hear the term all the time, but do you really know what they are? They are charged particles that rain down on Earth from space. Really! Kinda cool, huh? There are charged particles— mostly protons, or hydrogen nuclei, but with some heavier ions mixed in— smacking into our atmosphere all the time. Some of them have extremely high energies, higher energies than those to which we can accelerate particles in our best particle physics accelerators. Of course, the very highest energy cosmic rays are the rarest.
Thanks to a recent study by the Chandra Space Telescope, we have direct confirmation of the model that cosmic rays are produced in supernovae.
In fact, the Solar System is awash in charged particles, many of which come streaming off of the Sun. A "cosmic ray" proper, however, has a higher energy than most of the particles coming off of the Sun, and comes from outside the Solar System. The production mechanism for these cosmic rays is called the "Fermi mechanism," and involves compressing magnetic fields in supernova remnants.
The path of a moving charged particle will be bent in the presence of a magnetic field. Indeed, magnetic fields can "capture" moving charged particles (both from the Solar wind and cosmic rays), causing them to spiral about it. We have bands of charged particles, the so-called "Van Allen Belts", around the Earth; these are particles trapped in the Earth's magnetic field. As the particles spiral along the fields, they crash into the atmosphere near the North and South Poles, where the magnetic field lines dip down into the earth. The result of the collisions of these particles with the atmosphere is what can be seen on earth as aurorae.
Supernova remnants have two things. First, they have strong magnetic fields; we've known this for a long time. Second, they have expanding gasses. In general, if you have a gas whose particles are partly charged, magnetic fields will move along with the gas. As the high-velocity gas in a supernova expands into the interstellar gas around it, you get shocks where the expanding gas collides with the ambient gas. You will also have magnetic field lines getting compressed, as the magnetic fields in the expanding gas plow in to the ambient magnetic fields out there.
Charged particles moving about the field lines of expanding gas will bounce back and forth between the expanding magnetic fields and the ambient magnetic fields. As the two field lines come closer together, the magnetic fields pick up energy. It's similar to bouncing a tennis ball between two trucks coming towards each other. Each time the tennis ball collides with one track and bounces back towards the other, it picks up a bit of the kinetic energy of the oncoming truck, getting faster and faster and faster.
Normally, the charged particle would stay trapped in these strengthening magnetic fields (the compressing magnetic field lines) forever. However, there is enough junk there that eventually the charged particle— potentially moving quite fast now— will bounce off of something and get scattered out of the supernova remnant. At that point, it goes flying through the Galaxy as a cosmic ray.
The new observations show hot spots in X-rays appearing and disappearing in the shocks at the edge of a supernova remnant, which results from the sporadic production and release of the charged particles, some fraction of which will run into planets and be observed by the residents there as cosmic rays.
(Hat tip: Roger Amdahl of the Second Life "Astro News" group.)
Any guesses as to what fraction of the high energy (say >1GEV) may be coming from other sources? Things like Neutron stars, magnetars, less violent stellar outbursts, collisions of gas clouds ...
Nice result, but I have a couple of nits to pick.
First, these are presumably the "Galactic Cosmic Rays" (GCRs). There are other cosmic rays, mostly of lower energy, which are produced from accelerating solar wind particles at the heliospheric termination shock, by a similar mechanism as you describe for the GCRs. The highest energy cosmic rays (above ~10^12 eV) are of extragalactic origin.
The other point happens to be my area of expertise:
We have bands of charged particles, the so-called "Van Allen Belts", around the Earth; these are particles trapped in the Earth's magnetic field. As the particles spiral along the fields, they crash into the atmosphere near the North and South Poles, where the magnetic field lines dip down into the earth. The result of the collisions of these particles with the atmosphere is what can be seen on earth as aurorae.
The problem is with the last sentence. The particles that cause aurorae are not the radiation belt particles (which include electrons of 1-10 MeV and ions of >10 MeV/q) but rather 1-10 keV electrons which are accelerated a few thousand km above the Earth's surface. The latter are much more numerous, and they deposit their energy at higher altitudes where "forbidden" oxygen transitions (particularly the 557.7 nm line which is the green of the aurora) can take place. The radiation belt particles contribute to the airglow, but much less because there are relatively few precipitating particles at those energies.
The best site I know of for the basics of magnetospheric physics is the online textbook at Oulo University (http://www.oulu.fi/~spaceweb/textbook/). Oulo is in northern Finland, far enough north that aurorae are visible most clear nights (at least from September to April when midnight sun is not an issue).
It's been a couple of years since I've checked, but re: the extragalactic cosmic rays, has a GZK cutoff been observed? There has been this problem with the very highest-energy comsic rays that they have enough energy relative to the local CMB background photons that the Universe is "optically" thick (electronically thick?) to them, and they shouldn't be able to travel very far-- limiting potential extragalactic sources of those CRs.
re: the extragalactic cosmic rays, has a GZK cutoff been observed?
The latest I've heard was a year or two ago at a department colloquium. At the time there was evidence of a GZK cutoff, but it wasn't yet a slam dunk. Apparently some earlier studies failed to see the cutoff. It's hard to tell because fluxes are so low (<1 particle per km^2 per year) at the relevant energies, whether or not the cutoff is there.
Mr Lund or Dr Knop, can you please clarify "forbidden" transitions for me (i.e. as opposed to "allowed" ones, and why the "forbidden" ones are "permitted" in the aurora)?
There are various mechanisms by which there can be a transition between two quantum states in an atom. The most likely transitions go by an "electric dipole" transition. There are various "selection rules," but ultimate it has to do with the difference in the quantum numbers of the two states: the spin of the electrons, the orbital angular momentum, the total angular momentum, the z-component of the angular momentum, etc.
"Permitted" transitions are transitions that have a strong coupling between the two states. As such, they have a very high transition rate, or equivalently, the upper state has a very low lifetime.
A "forbidden" transition does *not* have an electric dipole transition; it is "forbidden" by the selection rules. However, there are other kinds of transitions -- magnetic dipole, electric quadrapole, etc. They *can* happen, but the probability (thus the transition rate) is much lower, and thus the lifetime of the state is much higher.
In a high density gas-- a gas such as is found in gas vapor tubes we make on Earth-- the lifetime of these "forbidden" states is so long that a collision with another atom or electron is very likely to happen before the photon-emitting transition has time to happen. As such, that transition is more likely to be collisionally de-excited than radiatively de-excited, and we never see the radition from it.
However, many astronomical gases are so low density that the time between collisions can be longer than the lifetimes of these "forbidden" transitions, so we *do* see the radiation from these forbidden transitions.
In other words, they're not the most probable transitions, and they are forbidden *to first order*-- and we don't observe them in the lab because other things cause the atom to leave the state before it can radiatively decay. In environments that are much lower density than what we observe in the lab, though, these "forbidden" transitions really do happen.
Actually Oulo is quite a way south of the quiet-time Auroral Arc (the location of where aurora occurs most days). We work with the VLF group in Sodankyla (maybe 300 km further north and they are part of a geophysical observatory run by the Oulu University) and I have been lucky enough to get there for a workshop (long way from my home in New Zealand), out of the two weeks I was there we had maybe 2 nights of aurora (which I missed - but I am going on reports of the rest of the people from the workshop), you normally have to go further north to Norway to get it more often (and even then it was too cloudy for me to see anything)
But I agree with what you said about the space physics textbook very worthwhile if you need information about this sutff.