I'm reading Gaggero's Cosmic Ray Diffusion in the Galaxy and Diffuse Gamma Emission and he makes the claim,

...the definitive proof of [Cosmic Ray proton acceleration in supernova remnants] would be the observation of neutrino emission by existing or forthcoming experiments such as IceCube or NEMO.

The existence of neutrinos stems from relativistic protons colliding with ambient protons. Neutral pions are the primary decay mode for $pp$ collisions, but charged pions can also be made alongside the neutrinos.

My question is then would the non-detection of neutrinos be a statistic issue or would it suggest Supernova Remnants do not accelerate protons?

  • $\begingroup$ IceCube has recently published an article with their latest statistic. The total number of detections was small, if I remember correctly (in the teens?). I have heard from folks in the neutrino astrophysics community that the measured number of particles is at the lower end of their cosmic accelerator models, which puts quite a damper on our ability to get good statistics with experiments the size of IceCube, which were planned with a bit of optimism, and I think they are already planning a ten times larger experiment. What they have just doesn't seem to cut it in terms of aperture. $\endgroup$ – CuriousOne Dec 17 '14 at 18:28
  • $\begingroup$ @CuriousOne: Do you recall which of their recent papers it is? $\endgroup$ – Kyle Kanos Dec 17 '14 at 18:30
  • 1
    $\begingroup$ Links to the papers are on their website. Here is a recent one: arxiv.org/pdf/1406.6757v1.pdf. You can check for yourself if four years of data are good enough or not... seems a bit thin for an experiment of that size. The other papers are linked here: icecube.wisc.edu/pubs $\endgroup$ – CuriousOne Dec 17 '14 at 18:33
  • $\begingroup$ Non-detection is always harder to do definitively than detection. $\endgroup$ – rob Dec 17 '14 at 20:12
  • $\begingroup$ Hooray, a random downvote today. $\endgroup$ – Kyle Kanos Mar 16 '16 at 11:20

It is more likely that the non-detection would be associated with statistics than SNR's not accelerating protons. Fermi-LAT has already shown that $\gamma$-ray emissions from 4 galactic supernova remnants (with molecular clouds nearby them) are coming from proton-proton collisions leading to neutral pions ($pp\to pp\pi^0$, $\pi^0\to2\gamma$):

enter image description here
(source, arxiv link)

From Halzen et al (2008),

Sources producing such hard spectra extending up to 100 TeV and more are required to explain the existence of the 'knee' in the cosmic-ray spectrum around 3 PeV, with young supernova remnants being the best candidates. However, their observation is difficult as these high-energy photons are produced inside the accelerator only within the first few hundred years.
Assuming an $E^{-2}$ spectrum with a cut-off at 300 TeV (consistent with a proton cut-off at the 'knee') we demonstrate that IceCube will be able to see these sources after several years of observation.

The one year result of the IceCube Neutrino Observatory (with 40 strings) shows a 90% confidence level of no point sources of neutrinos at the PeV energy range. With the extra 46 strings of the complete observatory, it is expected that point sources of neutrinos can be discovered.

The fluxes at TeV energies of $\gamma$-ray sources is around $10^{-12}$ 1/TeV/cm$^2$/s, which is near the sensitivity level of the full 86 string IceCube observatory. Aartsen et al (2014), using data obtained from 2008 through 2011, state

IceCube recently found evidence for a diffuse flux of high-energy astrophysical neutrinos, observing a 5.7$\sigma$ excess of events between $\sim$50 TeV and 2 PeV deposited within the detector. The 37 observed events are consistent with an $E^{-2.3}$ neutrino flux at the level of $1.5\times10^{-11}$ 1/TeV/cm$^{2}$/s/sr (normalized at 100 TeV), with a neutrino flavor ratio of 1:1:1. While these events have established unequivocally that astrophysical neutrinos exist, their sources have not yet been identified. One challenge is that only $\sim$20% of the events in that sample are associated with a high-energy muon which lieaves a visible track in the detector. The remaining events without a track have a poor angular resolution of $\sim$15$^\circ$.

From a live-time of 1373 days, that amounts to less than 10 events per year. Added to that, the lack of identifying the origin of these sources, the issue is more aligned with statistics than eliminating the hadronic SNR acceleration paradigm.


Neutrinos have already been detected from the supernova remnant SN1987A, so it's a sort of moot-point. The neutrinos are actually formed in the nuclear reactions in the core and then propagated to the surface through the intervening matter rather than from other sources, which are negligibly small in comparison. Despite the low interaction probability of neutrinos with normal matter, the density of neutrinos is so high that it causes a significant change in the dynamics of the interior during the supernova explosion. Computer models of supernovae including neutrino interactions are however very difficult.

I think the statement above made by Daniele Gaggero is asserting that although cosmic ray production and neutrino interactions are both linked to supernovae, the theory leads to a conclusion that one cannot happen without the other, so the detection of neutrinos is enough to assert that cosmic ray proton acceleration also occurs. It is a bold claim!

  • $\begingroup$ Yes, we've detected neutrinos fro supernovae, but not (conclusively) from supernova remnants which is what I'm discussing here. The $\nu's$ from CCSNe arise due to the collapse itself, whereas the $\nu$'s from an SNR would be due to the production of cosmic ray protons colliding with ambient/thermal protons generating charged pions, $\pi^\pm$, which decay to muons and muon neutrinos (and antiparticles thereof). $\endgroup$ – Kyle Kanos Jan 6 '15 at 13:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.