University of Wisconsin-Madison

2008 IceCube Update - Section 2

II. COSMIC RAYS

A major reason to build IceCube is to find the sources of high-energy cosmic rays [2-4]. Cosmic-rays were first observed almost 100 years ago by Victor Hess. Over the past decades, many experiments have observed the cosmic-ray energy spectrum and composition, from GeV energies up to 3×1020 eV. The flux drops rapidly with energy, reaching 1/km2/century at the highest energies. Cosmic-rays have a mixed composition containing mostly nuclei from proton to iron, with at most a small fraction of heavier nuclei and photons.

Despite the decades of effort, we still know very little about the origin of cosmic-rays. At energies up to 1015 eV, cosmic rays are strongly bent in galactic magnetic fields. They likely originate in our galaxy. Supernovae remnants are the most likely sources. Their strong magnetic fields and shock waves can accelerate charged particles.

Galactic magnetic fields are too weak to confine more energetic particles, which are thought to be primarily extra-galactic. Possible sources are active galactic nuclei (AGNs, galaxies with central supermassive black holes) which emit jets of relativistic particles along their axes. Or, cosmic-rays might be accelerated by the sources of gamma-ray bursts (GRBs). GRBs are believed to originate in the collapse of supermassive stars and/or mergers of black holes and/or neutron stars. Either of these sources may provide appropriate conditions to accelerate nuclei to ultra-relativistic energies.

The most energetic cosmic rays have limited ranges. At energies above about 4×1019 eV, cosmic protons are excited by collisions with the 30K microwave background radiation, creating a Δ resonance. The decaying Δ emits a lower-energy proton. This energy loss limits the range of more energetic protons to about 100 Megaparsecs [5]. Heavier nuclei are photodissociated by interactions with the microwave background; this leads to a similar range limitation.

Further, all but the most energetic cosmic-rays are bent in the intergalactic magnetic fields and so do not point back toward their origins. At energies above 6×1019 eV, bending by interstellar magnetic fields may be tolerable. The Auger collaboration has found evidence that some cosmic-rays may point toward nearby (within 75 Megaparsecs) AGNs [6]. However, the Fly's Eye collaboration does not observe this correlation [7].

In the absence of definitive correlations, we must consider other messengers. TeV photons have been observed from some nearby sources, such as supernovae and some nearby AGNs. At energies above a few TeV, photons interact with interstellar photons, forming e+e- pairs; like protons and heavier nuclei, these photons also have a limited range.

In contrast, neutrinos have very small cross-sections and so can freely travel cosmic distances. They are the only particle able to probe high-energy accelerators out to cosmic distances. Here, we focus on the neutrinos with energies above about 100 GeV which are most relevant for understanding cosmic-ray acceleration. These neutrinos are produced in π± decays, π± −−> μ± νμ; , followed by μ± −−> e± νμe, producing a 2:1 ratio of νμ;:νe. IceCube cannot differentiate between ν and anti-ν, so we will combine the two particles. Over long distances, neutrino oscillations change this 2:1 ratio into a 1:1:1 ratio of ντμ;:νe. The charged pions are produced in incidental 'beam-gas' interactions, whereby the nucleons under acceleration interact with either gas or photons present in the accelerator. If cosmic-rays are heavier nuclei, νe may also be produced by nuclear beta decay of unstable isotopes produced in spallation.

The neutrino flux from cosmic-ray accelerators has been estimated by two methods. The first uses the measured cosmic-ray flux and the estimated photon and matter densities at acceleration sites. The second extrapolates the measured TeV photon flux to higher energies, assuming that the photons are from π0 decay. That leads to an estimate of the number of π±. Both approaches find similar neutrino fluxes, and both lead to a similar conclusion: that a neutrino detector with an area of ∪ 1 km3 is needed to observe neutrinos from astrophysical sources.