As I mentioned in the Can neutrinos travel Faster Than Light thread, there was a paper written 14 years ago looking at the neutrino observations from that supernova. The author did some calculations on the different arrival times of the mass eigenstates, due to their different velocities. For a given initial neutrino energy, the different mass eigenstates will have different velocities, and over 170,000 years, that difference will lead to different arrival times for the different mass eigenstates, each of which will be a mix of the flavor eigenstates.
My question is, given the above, given our current best estimates of neutrino mass differences, and given current bounds on their possible masses, should there have been a double bang (or triple bang) of neutrino detections in the neutrino detectors operating at the time? Either within a single detector, or collectively for the three mentioned in the article? When should the other two bangs have occurred relative to the main one detected, and how many neutrinos would be expected to have been detected?
Also, is there any way to account for the five events observed 4.7 hours prior to the approximately 25 in the main set (preferably without any neutrinos traveling FTL)?
Using our current knowledge of neutrino masses and mass differences, and taking 20 MeV as a typical energy for a detected supernova neutrino, the arrival times for the different neutrinos would differ by about 15 microseconds for the larger mas gap and about 0.5 microseconds for the smaller mass gap. The 24 neutrino event candidates seen in Kamiokande II, IMB, and Baksan spanned about 12 seconds, and deviations on the microsecond scale would be unnoticeable.
There has been a trickle of papers over the past 25 years attempting this. Most hypothesize some new supernova explosion dynamics that introduce a two-stage effect in the core collapse, ranging from rotation-driven intermediate equilibria (e.g., core collapse first to a neutron star, then again a short while later to a black hole) to QCD phase transitions occuring en route to a blck hole. It is worth noting that supernova models over the past 25 years have gone from adorable in their primitiveness to (in the past five years or so) impressive computational tours de force, although there is still much left to be desired. So, the double-pulse idea is an open question, and everyone is very anxious for a galactic supernova, which would bring an absolute deluge of new information into the fields of neutrino physics, astrophysics, and gravitation (the latter especially if Advanced LIGO is running before the 'splosion).
A supernova within our Galaxy might even be picked up by GEO, Virgo, or one of the other detectors, even if we’re so unlucky as that it happens before the Advance LIGO upgrades are finished. And for that matter, the typical time between supernovae in a galaxy is pretty long compared to the timescale at which gravitational astronomy is currently progressing, so it’s pretty plausible even that we’ll have something like the Einstein Telescope, or something generations beyond that even, up and running before we get another supernova.