How do you align a telescope to make sure it’s pointing in the right direction? Piece of cake: you use any cheap sky mapping software that could give you the position of a star in the sky at a given time for your location, you point your telescope in that direction and make sure that you see the right star. Done!
Now, how do you align a neutrino telescope like IceCube to make sure it’s pointing in the right direction? Well, as you may imagine, this is a little bit harder. With no high energy neutrino sources detected so far, we can’t really use the same procedure as with an optical telescope, but there’s a way: using cosmic rays.
Every second, IceCube detects about 2000 muons coming from the interaction between cosmic rays (protons, most of the times) and the Earth’s upper atmosphere. These protons have energies of 20 TeV on average, about 6 times the energy of the protons going around the LHC ring. Although protons of this energy can’t be used for galactic astronomy since their trajectories are bent by the interstellar magnetic field over scales of tens or hundreds of parsecs, they can travel on pretty straight paths over shorter distances like those characteristic of the inner Solar System.
Just as there are no neutrino point sources that have been observed so far, due to these magnetic deflections there are also no point sources of cosmic rays in the sky, but in this case the idea is not to look for a source of cosmic rays, but for an antisource. The Moon, conveniently located only 360000 km away from our planet, is a very good absorber of cosmic rays. Because of that, the Moon should be casting a cosmic ray shadow, blocking all cosmic rays that come from its direction, and then preventing us from detecting the muons associated with those ill-fated protons in IceCube. Also, since the distance is small the shadow is not destroyed by magnetic field deflections for cosmic rays of TeV energies, so the effect should really be observable by IceCube.
If we map the cosmic rays detected around the location of the Moon, we should see a deficit in the number of cosmic rays that are coming from that part of the sky. If we don’t see such a deficit, then that means that there’s something wrong with the directional reconstruction algorithms that we use for cosmic rays and neutrinos. This is actually an old idea, first proposed by G.W. Clark in 1957 and used in many occasions by different experiments to characterize the angular resolution of their detectors.
In IceCube, this work was performed by Laura Gladstone, David Boersma and two undergraduate students: Jan Blumenthal and Hugo Stiebel. Here’s how IceCube “sees” the Moon. The map shown below corresponds to the Moon shadow observed with IceCube in its 59-string configuration that was operated between May 2009 and May 2010. During this period, the detector recorded 22 million events in an angular window of 8 x 8 degrees centered around the position of the Moon.
The map below shows one result from the analysis. The color scale indicates the total number of events shadowed by the Moon at each position in the map. The deepest deficit corresponds to a total of 8192 events blocked by the Moon, and the location of this deepest point agrees with the expected position of the Moon to within 0.1 degrees. The statistical significance of the detection is around 14 sigma, where the usual rule in particle physics is that anything above 3 sigma indicates “evidence” for something being observed, and 5 sigma indicates a “discovery.” Probably it is too late to claim discovery of the Moon given that people have even walked on it, but those 14 sigma tell us that we are very confident that the shadow we see is not a fluke.
We’re right now writing a paper that will include all the details of the Moon shadow observation.
Not only the fact that we see the Moon is important for IceCube, also its width tells us something about the precision of the reconstruction methods that we use to determine the arrival direction of the cosmic rays. The width of the shadow is of the order of 1 degree, which agrees with simulations that we have of the detector.
Besides its use as a sanity check for the pointing capabilities of the detector, the Moon shadow can also be useful for physics since it can provide a measurement of the antimatter content in cosmic rays, a value that has implications for dark matter since this antimatter could be generated in WIMP annihilations.
Due to the magnetic field of the Earth, the proton shadow of the Moon should be slightly shifted to the left of the expected position of the Moon in the map above. This effect is stronger at lower energies, where the offset can be as large as ~1 degree of offset for cosmic rays of ~1 TeV. For IceCube energies the effect may be too small to be measured, but other experiments working at lower energies actually see this deflection happening. What’s important about this deflection is that if there’s a small fraction of antiprotons in the cosmic ray flux they would produce a shadow that’s deflected by the same amount as the normal proton shadow but in the opposite direction from the Moon due to their opposite electrical charge. At the moment I’m writing a cosmic ray propagation code to see how much and in which direction we should see the shadow shifting when observed from the South Pole (see an example of the propagation code output in the image below).
If such an “antishadow” is observed, a direct measurement of the antiproton-to-proton ratio in the cosmic rays can be done by comparing directly the strengths of both shadows. If such a shadow is not observed, then a limit could be set on this ratio. This is what several experiments have done. Most recently by ARGO-YBJ, a cosmic ray detector located in Tibet. The current status of the antiproton/proton ratio measurements is summarized in the plot below, which I took from a conference proceeding by ARGO.
Most of the direct antiproton measurements have been performed only up to 100 GeV, with a ratio indicating that there’s only one antiproton for every 10000 protons in the cosmic ray flux. At higher energies only limits have been set so far with values of about 0.1 in the ratio, or equivalently that there’s less than 1 antiproton for every 10 protons.
The recently launched AMS-2 cosmic ray detector that’s now taking data attached to the International Space Station should be able to extend the direct measurements up to 10^3 GeV.
I’ll keep you updated about any news on the IceCube moon shadow front. For the moment, I leave you with a song appropriately called ‘Moonshadow’ by Yusuf Islam (aka Cat Stevens.) I’m sure he was thinking of muons and antiprotons as he was writing it!