High energy cosmic rays are something of a conundrum wrapped in an enigma. Essentially, they can't come from very far away and still have the energy they possess. To that end, cosmic rays should originate from within the Milky Way. Yet, they seem to be coming from every direction: no matter where you look in space, you have the same probability of seeing a high energy cosmic ray. A new paper has, to the disappointment of the 90 plus authors, confirmed this uniformity to a rather high degree.
Living in a frosted fish bowl
Let's start this with an analogy. Imagine that you are inside a frosted glass bulb. When the Sun comes up, you can see light, but it seems to come from every direction evenly. There is no way to tell that the light actually comes from a single source, shining from a single direction unless the light is sufficiently bright or the frosting on the window is not too dense. Then, even though you still see light from every direction, the slight brightness increase in one direction tells you that there is a light source in that direction.
Cosmic rays with energies up to 2TeV are thought to originate from dying supernovae in our own galaxy. Observations from the Fermi satellite have confirmed that some cosmic rays do originate in supernovae, but these observations don't seem to account for the full total of cosmic rays. (Note that there are cosmic rays at much higher energies, but these certainly do not originate within our galaxy.)
Why don't we see this as a collection of point sources? Cosmic rays consist of charged particles, which scatter like crazy on stray magnetic fields. So, even though there might be only a few sources, we see cosmic rays coming from every direction. To make the story even more complex, when charged particles change direction, they radiate light and lose energy. The reverse can happen, too: charged particles can scatter off of light to produce the same effect.
Cosmic rays come from everywhere
Combined, all of this smears out the individual sources of these particles and slows them down over time. The result is not just that it looks like we are observing cosmic rays through frosted glass, but that the sources of high energy cosmic rays cannot be too far away. From our perspective, distant sources would only be observed as low energy cosmic rays because of all the energy that the ray loses between its source and us.
Scientists thus find themselves in an awkward position: apparently, the sources of high energy cosmic rays should be close—think of distances on the order of 5,000 light years, which is just a tiny fraction of the diameter of the galaxy (100,000 to 200,000 light years). But, if they are that close, we should be able to make out bright and dark spots in the sky through careful observation.
We now have seven years' worth of observational data from the Fermi Large Area Telescope, an orbiting Gamma ray observatory. Seven years of data sounds like a lot, but there aren't that many events. In any particular direction, cosmic rays with energy above 500GeV turn up about twice a year—though that is still good for some 7,000 events per year.
It all boils down to statistics
The question then becomes: do these cosmic rays preferentially come from any direction?
Answering that isn't easy. Let's think about flipping coins. For a fair coin, we expect, on average, to get as many heads as tails. But if I flip a coin 10 times, there is actually a very good chance that I will obtain a number other than five heads. In fact, let me demonstrate that: I flipped a coin 10 times, and got seven tails. Jennifer, my daughter, repeated the exercise and got four tails. Neither of these are particularly close to the average. However, we expect that for any sample of ten, the most likely result is somewhere between four and six tails.
For larger numbers, we get closer. Combining mine and Jennifer's results, we come to 11 out of 20, which is close to the average (and the most likely result is between eight and 12 tails).
The point is that for small numbers, we expect there to be large relative deviations from the average. To determine if cosmic rays are coming from specific directions, it's not good enough to detect a higher-than-average number of cosmic rays from a specific direction. Instead, we need to detect cosmic ray numbers that are larger than the average and outside of the deviations that could be reasonably explained by random chance.
The upshot is that you need to have a really good model of cosmic ray propagation and multiple methods with which to analyze the data. Each analysis should be internally consistent and, when compared using simulated data, should all give similar results. This is the core of the new paper—proving that the analysis chain is solid.
The end result is that we cannot say that high energy cosmic rays are more likely to come from specific directions. However, the statistics are such that we would only expect to detect a source if it were quite nearby and rather young. For more distant and older sources, we will require many more events.
Luckily, the Fermi satellite is still operating and collecting data. If all goes well, we can expect an even better analysis in the next five years.
Physical Review Letters, 2017, DOI: 10.1103/PhysRevLett.118.091103
reader comments
60