Prospects for early detection of supernovae via neutrino experiments

by Alex Habig
May, 1997

This item posted to VSNET on May 14, 1997.

Fraser Farrell writes :

> How accurate (and precise) a position would be provided by this
> system?  Would an Alarm Report read "possible SN in Centaurus" or
> "possible SN at RA 13h19m11s +/-5s  Dec -55d20'09" +/-02"  ?

From comparisons of timing between the neutrino arrival directions at the various experiments, some position can be given. However, with only two experiments, that information will be poor. Geometrically, two timing points gives you a ring on the sky, whose width depends upon the angle between the SN position and the two experiments. That line could be 10 degrees or so at best.

Once more experiments come on line, that produces the intersection of two such cones, so we'd be able to say "SN between RA 13h19m and 13h45m, Dec. between -55d and -50d".

Also, neutrino experiments are only sensitive to galactic (and LMC/SMC) SNae. That further narrows the search, in a painfully loose sort of way.

If the large water cerenkov experiments (SuperK and SNO) can be talked into providing the pointing information available from the measurement of the nu->e scattering interactions, then we'd also have a couple of cones on the sky of radius 10 degrees or so to add to the timing information.

> Would the system give any indication of expected visual magnitude?  Even
> an approximate guess would help; for example, a 50000-neutrino pulse
> should be a brilliant SN, but a 50-neutrino pulse would warn us to get
> the binoculars and starcharts out...

It will be in our galaxy, so star charts shouldn't be too faint. However, if the SN lies on the other side of a big cloud of dust, then optical observers are simply out of luck. Let's hope that mother nature is kind.

We hope to include some estimate of the relative intensity of the neutrino signal, however. It will be a subjective scale, though, since it's not the coincidence network's job to analyze the data from the various experiments at the event level. After any initial alert from the coincidence network, the individual experiments should chime in with more carefully analyzed reports which should help out.

> How many hours/days between the neutrino pulse and the visual outburst?
> Or is this piece of information one of the things you hope to discover
> from this project?

This varies widely with the progenitor star (and whose models you believe). SN1987A had a delay of about 1h40m (although this was modeled after the fact, instead of being directly measured). Some models of other compact blue giant stars produce shock breakout times of as low as 35m. Red Supergiants, though, are so extended that it'll be hours before anything gets to the surface. If Betelgeuse goes, we've got some time to prepare for the show :)

Reconstruction of this time for SN1987A depended heavily upon chance observations (both positive and negative) by an amateur astronomer named Jones (apologies, I'm forgetting your first name right now!).

If the world has warning and is out looking, we'll get a lot of observations at early times, and be able to well determine the breakout time. This gives us a very good handle on what the progenitor was like.

> You should post any Alarm Report into vsnet-alert and the International
> Supernova Network (  A lot of astronomers and
> institutions subscribe to these two mailers.  On numerous occasions,
> these have spread some vital news worldwide before any official messages
> from CBAT.

Will do!

       Alec Habig, Boston University Particle Astrophysics Group

Back to Michael Richmond's home page