According to Thomas Gaisser in his book 'Cosmic Rays and Particle Physics' the measurement of neutrinos from an astrophysical beam dump give information that complements the photon signal (can not be measured simultaneously at the same site because one has to look up for photons and down for netrinos), for exemple ICECUBE experiment (south hemisphere ) for neutrinos and MAGIC experiment (northen hemisphere ) for photons, you can see the book at page 188

Yes, provided that you have sufficient neutrino events (particularly neutrinos scattering on electrons) you should be able to "point back" to the source. However, 100 Mpc is really far, usually well below our ability to detect such events.

Thanks Ahmed and Joseph for your entries.

The recent sky surveys in the optical band (e.g. Lick Observatory Supernova Survey, LOSS) are detecting SN events at nearly 140 Mpc (see e.g. Smith N. et al. 2011, MNRAS).

It would be very interesting to know the furthest reaches available to the current neutrino detectors available to date, given their detection sensitivities.

Is it like 10 Mpc? 50 Mpc?

Concerning SN neutrino detection, the current state of art detectors can point back to the source up to distances of some 100s of kpc's.

According to the following paper , If Super Kamiokande pointed back to the galaxy of Andromeda (~770kpc) in search for SN neutrinos they would get 1 event above background..

http://arxiv.org/pdf/astro-ph/0701081v1.pdf

No, IceCube reach is about 100Kpc or less!

If the no goes to my reply, I'd like to remind you that Ice Cube has higher energy threshold than SK . Also, even the Deep Core module doesn't have the pmt coverage of the SK resulting in more poor angular resolution at Supernova neutrino energies.

no, sorry Emmanuel I was answering to Ehsan about the 100/50Mpc question.

ok! One silly question: How do we determine the neutrino telescope's reach exactly?

I take it that one computes the angular resolution for the desired energy interval,and then matches this resolution to the angular size of an object to the sky. The longest the distance, the smaller the size. So therefore by the angular size- distance relationship, one says " I can resolve a 1.1 deg object with my telescope and this corresponds to (..)kpc distance" right?

In the case of IceCube what we see is an increase in the overall noise level. Neutrino flux diminishes as the square of the distance so only very near sources give a rate high enough as to be statistically over that noise level. Directional capabilities are very limited in this case, although not impossible.

Well, there was a SN event that can be used as a comparison, the SN 1987A in the LMC. As you probably know for that SN neutrino's bursts were detected by three experiments: Kamiokande II saw 12, IMB in Ohio 8 and Baksan 5. They all together collected 25 neutrinos above bckgd compatible with SN event.

Since flux scales as the inverse of squared distance, the flux of neutrinos at Earth of a SN occurring at distance R would be then F = F_o X (R_o/R)^2, where F_o and R_o are the flux of SN 1987A at Earth and R_o the distance of LMC

The neutrino count rate in a detector is roughly dN/dt = F X AO X e, where AO is the detector geometric acceptance (which is a combination of detection area and solid angle aperture) and e is the detection efficiency (which is a combination of volume, mass, material and cross section).

Since flux scales as the inverse of squared distance, the count rate for a source at distance R would be dN/dt = F_o(R_o/R)^2 X AO X e

If the source was at 2R_o, nu's counts would have been lowered by a factor 4: 3 in KamII, 2 in IMB, 1 in Baksan. At 3 R_o, we would have had a factor 9 less: 1 for KamII and 0 for the others. That means no detection above bckgd.

R_o is about 50 kpc, so it looks like that beyond 100 kpc might be difficult to detect neutrinos. At least with those detectors (in 1987).

Conversely, to see the same number of nu for a source at R = 2R_o, one should increase AO X e by a factor 4. If correct, it looks unlikely that we will be able to detect neutrinos form other galaxies (implying Mpc distances).

Alos, to do neutrino astrophysics we would need an energy spectrum of SN neutrinos and this implies to detect hundreds, if not thousands, of nu's in many different energy bins. This is not realistic right now. Also, to measure very accurately the arrival direction is mandatory to point back and match the optical counterpart, if visible (say, if you see nu's coming from the opposite side of the sky wrt the SN, then you might suspect that they are uncorrelated).

Dear Emanuele,

after this answer I understand completely the distance measurement in nu-telescopes.

In the 1987 observations how many neutrino events were the background?

Typically one should set a coincidence window between the optical and neutrino signals coming from the same point in the sky. Do you know exactly what was the time coincidence of the SN measurements?

Hi Emmanuel,

from what I remember, neutrinos from SN 1987A arrived about three hours before the light flash. This is ok, because the light flash starts to be visible when the shock wave coming from core collapse emerges from the star envelope, while neutrino escape promptly the star. In a word, light emission is delayed, while neutrino emission is prompt. The time coincidence was made after the flash light was seen.

Actually I dont know the details of how many bcgkd nu were expected. I know that the event identification as nu event was pretty robust (Ihave the right paper on an other pc right now...if you wish I can add it here). Also, solar neutrinos should have much less energy (few MeV). In the attachment there is the energy-arrival time scatter plot (stolen from http://nu.phys.laurentian.ca/~fleurot/supernova/).

Thanks Emanuele for the plot.

What actually attracted my attention is the time lag in the middle of the plot for the detection of neutrinos by Super-K.

Is it just by coincidence?