Supershells: GRB vs Multiple SNe

Particularly novel points about GRB vs. multiple SNe supershells, raised in response to Part II of the 1998/99 AY208 final exam:

Robin McGary on shell temperature: The temperature achieved by the GRB shell will exceed that of the multiple SN shell at any phase of its evolution. If measurements could show that the temperature of a shocked region in a shell exceeds the maximum possible from multiple SNe, this argues in favor of a GRB origin.

AG: The temperature would need to be measured by using the measured intensities of shock-tracing excited lines of various atoms (e.g. C, N, O). It may actually be possible to do this. It is worth investigating how accurate current estimate of supershell-shock region temperature are. I think that this may only be a viable technique when the shells are younger than any that are observed, but someone should check.

Robin McGary, Pinaki Chatterjee, Aaron Sokasian and others on cooling rate: Several students pointed out that the cooling rate of the continuous energy injection models (i.e. multiple SNe) is different from a single burst model (e.g. GRB). For a single burst, in the Sedov-Taylor phase, T~R^-3, whereas for continuous input, T~R^-4/3. (See Shull & Saken 1995, ApJ, 444, 663.)

AG: Iím not sure we live long enough to measure the cooling rate. If one could find a way to measure temperature that was extraordinarly accurate, I suppose a 20 year baseline might give a time dependence (much the way VLBI can give useful proper motions on that time scale), but this would be very difficult. The necessary calculation here is to see what the expected difference in temperature is for a currently observed shell, over 20 years.

Robin McGary on using Ha observations: With high enough spectral resolution at Ha, it may be possible to detect multiple shocks, travelling at different speeds, within supershells. If such shocks are detected, this would surely favor a multiple SNe scenario

AG: The new WHAM data may be good enough to do this experiment. Someone should check!

Pinaki Chatterjee on shock interactions: Citing the simulations of Tenorio-Tagle, Bodenheimer, and Rozyczka (1987, A&A, 26, 145; see also several other works by this group, including Rozyczka et al. 1993, MNRAS, 261, 674.), Pinaki points out that the interaction of shocks from multiple SNe is complex, in that they can reflect off the (previous) shell boundary and also lead to Rayleigh-Taylor instabilities. The simulations show that it is very difficult to produce the largest supershells with multiple explosions because of this.

AG: Although I donít think this will settle the SN/GRB issue definitively, it would be interesting to look for evidence of the instabilities via high-resolution observations of the shells (as proposed by Rines for possible AGN supershells (see below).

Daniel Harvey, Ken Rines and others on shell symmetry: Several students pointed out that a single explosion would create a more symmetric shell than multiple explosions.

AG: This is true, but very hard to observe, given that the density structure of the ISM will ultimately dominate the detailed shape of the supershell. Not many students, except for Aaron Sokasian, appreciated how much effect cloud density structure will have on the ultimate shape of a large supershell.

Daniel Harvey on shell location in a galaxy: Several students noted that if the progenitor of GRBs was known to be from a different population than SNe (e.g. not OB assosications), then the statistics of the radial distribution of shells in external spiral galaxies could shed light on their origins.

AG: Yes, but we need to know more about GRBsí origin by other means at the same time.

Ken Rines on a search for supershells around nearby AGNe: If AGNe harbor GRB sources, one might find supershells associated with the inner reaches of AGNe. Ken proposes an interferometric study, and points out that at 7 Mpc (distance to NGC 4258) one would need a resolution of 1.5 arcsec (a 4.7 km baseline), such as at the VLA.

AG: As Ken asks, the question is whether the column density of H I expected in the shells would be high enough to detect in such an HI-VLA observation. That should be checked?it might be possible. Also, Iím not sure this study would rule out SNe, which might also occur in large numbers near the centers of AGNe, but finding supershells in AGNe would be interesting nonetheless.

Feryal Özel and others on initial conditions: Feryal pointed out that more complete surveys of the initial conditions (e.g. stellar census; gas distribution, etc.) in and around OB associations in galaxies would help establish the plausibility (or implausibility) of the multiple SNe hypothesis.

AG: I agree?and the new Hipparcos data should help with that. I am not sure, though, whether anyone is specifically conducting a systematic census for O and B stars using the Hipparcos or other new data.

Vincent Fish and several others on "young" GRB shells: If only we could find a very young shell, many ideas would work. Vincent in particular points out that a GRB should create much more synchrotron radiation than correlated SNe. So, a search for patches of intense synchrotron emission might help.

AG: Surveys of other galaxies in synchrotron emission DO exist. Many have been carried out with Effelsberg, and some with the VLA. It might be possible to look for otherwise unexplained synchrotron peaks in these galaxies and then follow them up with other observations. I seriously doubt the number of galaxies mapped in synchrotron is anywhere near high enough at present to expect to find a young GRB, but one might find other interesting "point-like" sources.

Xiaohu Wang on young shells: Xiaohu points out that a SN-generated shell will have much less kinetic energy when the shell is small than will a GRB shell, so that if one could find a small shell with very high kinetic energy, a single-explosion hypothesis would be favored.

AG: True, but we still have to find a young shell!

Return to AY208 homepage