Peter doesn't seem to be here. The word is that he's down talking to Alex Filippenko.
He's an undergrad taking a semester off who will be working with us, probably primarily on the Satellite. (Surprise.)
Alex Kim is coming back, who will be working on the Satellite, although Saul predicts he'll do a little on other stuff. Saul thinks Alex is talking about coming at the end of October or the beginning of November.
Somehow we've gotten sidelined into a discussion of how much the Satellite will cost, and how much Shank and others are claiming it will cost.
Don's telling us about the CCDs. Some details about what's happening with each chip that they're trying on. Much verbage; I didn't really distill a bottom line.
Now Saul is asking for random newsitems from people who talked in the last few weeks. It seems we have reverted to our old meeting format. This to me is a tragic development.
Matthew is talking about the magnitude system, Energy vs. Counts and a factor of 1+z. One thing Matthew points out is that the issue of 1+z lines in the zeropoints. If you plot magnitudes calculated from energy vs. counts, there's a difference of a factor of lamba within the integrals. Depending on whether you're using a count or an energy zeropoint, you do or do not get different factors of 1+z out of this. If the difference in energy zeropoint is 0, the difference in flux zeropoint won't be, for different filters. This comes in to be exactly the same 1+z factor, for basically the same reason.
Matthew asserts that it's almost impossible to do a flux calibration in counts, because things are compared to flux blackbodies. Referece to the Hayes & Latan zeropoints are in order. This does seem (to me) to be an argument that everything should be done as energy K-corrections.
It's probably worth looking at what we've done to make sure everything is consistent.
Gerson wants to say that he's fit a template ... well, Gerson and Saul are correcting each other, and I lost the flow of it. The answer is that he can get the "two" lightcurves (whatever they are) such that they differ by more than 2% over the full range of where our data and the Hamuy data overlap. What does this mean? If the introduction hadn't been such a rush of overlapping verbage, I might have some idea. Basically, this is turning into a discussion between Don and Saul and Gerson, and it's not clear why it needs to be happening in the group meeting context.
I had really thought that the group meetings had gotten a lot better, with the assigned topics, the last two or three times. I am sorely dissapointed to see them reverting to their old format.
Carl tells us that HOU got on the NSF "great accomplishments" list, based on the Kuiper Belt discovery in SCP data.
Peter tells us that the NSF zeroed out funding for Hy Spinrad, Mark Davis, and somebody else (whose name I missed). (David Branch?) This will nuke out some grad students. Peter says that there are only three people getting money from NSF for supernovae; Bob Kirshner, Alex Filippenko, and Craig Wheeler.
There is some discussion about whether or not the upcoming French CFHT search run will be searched for asteroids.
Saul has written up the upcoming proposals; they are the only ones we're doing this round, that Saul thinks we'll have to worry about. CFHT, Keck, VLT, Subaru, and HST. The first one to hit will be CFHT, probably due 8/30 or thereabouts. HST is due 9/10, and Keck is due 9/16. VLT is probably 9/30.
Reynald will write CFHT, Isobel will hopefully update VLT. Somehow we're going to have to manage Keck and HST... and Subaru. Isobel and Peter worked on the one for next semester; Saul thinks that maybe he can get Isobel to work on that one as well. That will leave us with the HST to do. The other thing to note is that the HST doesn't overlap the timeframe of the other three. CFHT and Keck, are on similar but not identical schedules, more or less February through July of 2000. VLT is April through September 2000. HST will start August or September of 2000.
Saul lists four projects; z>1, Metallicity, Other Evolution Spectra?, and UV low-z. The last one will be a second HST proposal, HST following up to get lightcurves. This will piggyback, presumably, on the low-z supernova factory. They will be easy observations; the spectra will be hard, though.
Regarding getting z>1 supernovae, NICMOS won't be back online for the HST cycle we're comparing to. This makes it difficult to follow the most different supernovae, because we're comparing to U-band lightcurves. Those are more scary than B-band, because nearby lightcurves aren't as well understood, and because it's more affected by reddening. Peter is asserting that the lightcurve stretches along with the others, but that nobody knows if there's a good stretch/luminosity relationship in the U-band. Greg says that Brian Schmidt at a meeting basically said that U-band spectra don't make good standard candles. Peter doesn't think that anybody's ever sat down with good consistent data and checked it.
There is some question of using AO and Keck in J-band, which would be as good as HST. Peter raises the question of whether or not we can actually calibrate the J-band there. Then we could say we were going to use U to get the stretch and epoch, and J to get the magnitude. There is argument over whether we could do two different filters in the IR; J and J', or J and H. Peter asserts we need these to do cosmology, in order to know the color for reddening and such.
Peter questions if AO is on Keck next year. Greg asserts that it's on next semester, though Saul thinks that that is not with the laser.
Saul wonders if it's possible with Subaru to search for supernova, for so long, that we will enough within a arcminute (thereabouts) of a magnitude 15 star. (The problem with just searching with Subaru is that we'd find roughly SNe at z>0.9, with one night at Subaru. This is based on Peter's calculation.) How many supernova candidates do you have to get to get a few within an arcminute of a magnitude 15 star? Greg points out that our 5h field is at low latitude, but in a hole in the dist distribution, so that might be an ideal situation.
Saul is raising the spectre of using Type IIs. (Basic idea: they are approximately blackbodies at early epoch. Get the temperature (and expansion veolcity?) from the spectrum. Get two points on a lightcurve, solve for luminosity and area. Make small corrections (up to 20-30%, Peter says) from blackbodies.) What would you be measuring with the HST on these guys? Peter says that they are extremely good blackbodies at early time, and only "good" blackbodies at later time. (Actually, you need two spectra.)
Peter asserts that you would have to redo the way you do your search to find enough II's at CFHT and Subaru.
Saul wonders if we could combine the U-band Ia and II proposals. They are very different redshift objects.... Peter points out, however, that the ultimate goal is the same, and that we're looking at objects we'll find in the same datasets at the same time. Greg says that people will get mixed up about what data is needed etc. if the proposals are put together. He and Susana argue for two proposals... which now seems to be what Peter prefers too.
So now we're looking at 3 proposals. UV, with Andy Fructer. II, with Ed Baron and David Branch. Finally, the z>1 stretch/epoch U-band deal, which will be the "SCP" proposal.
There is also the possibility of doing 1% (!) photometry on supernovae at a redshift of 0.5.
Back to the upcoming Spring, for which we will use the current HST stuff. Should we ask to start trying out AO ideas? Perhaps we could do a Subaru search, with Keck AO followup. This would be important to verify that we can do it before we have the HST depending on all of it.
It looks like we're going to be doing a standard sort of high-z search. However, the issue of doing AO IR does raise it's scary head. You would really need either NICMOS or that in order to get B-band lightcurves of these puppies.
5-minute break. Or more.
Metallicity effects in Type Ia supernova. That's a redshift 1/2 deal, but you need very good spectra. You won't get enough z=0.5 supernova with a CFHT run (or will you?), but you would with a Subaru run. Or, presumably you could just structure the CFHT run to find z=0.5 supernovae. This won't work for next semester, since we are supposed to do z>1 supernovae at HST. Also, we need J-band photometry at Keck, since that's what we said we'd do in the HST proposals.
Peter asserts that at Keck we need spectroscopy from the things at CFHT. There's 2 or 3 nights right there. We also want a night of IR with AO. Hopefully some of this might come from Richard Ellis. Saul talks about putting in two proposals; one for Spectroscopy, one for AO at IR. One of these would be an Ellis proposal.
There is a whole 'nuther host morphology deal that Richard may do with some of his time at Keck. (He may get 6 or 8 nights at Keck next semester.)
There is the question of NIRSPEC. How low can you go in wavelength? It might make more sense to do our spectra with that than with LRIS. How about sending in the prposal saying we'll use NIRSPEC, but if NIRSPEC isn't available we'll do it with LRIS. (Which is natural, since they're both on Keci-II... if we don't get NIRSPEC, we just will get LRIS (if we get any time at all).)
Greg questions if we want access to an optical spectrograph.
As to when we'd do the search, there's a question of when the HST servicing mission is. Would we search in March or April? Or May? It sounds like we want to do April, unless HST makes that untenable. We would ask for VLT time for I-band followup and maybe even J-band AO (which may not be ready). We could follow up the lower redshift (i.e. 0.8) supernovae with VLT, in R and I.
WRT VLT, there's the question of how well you can do J from the ground. The other group is going to be trying to do it... do they know something that we don't?
Saul is proposing that the VLT proposal would be more or less just like this semesters, where we say we have this big continuum of objects from 0.8 to 1.2, and it's a big concerted effort, and we're gonna get some of our followup from VLT. We may ourseles keep in mind that the 0.8's are the main focus of VLT, but don't make it sound like second-class objects in their proposal.
Now there's a question of whether or not one of our CCDs might be in any of these cameras next semester.
Saul wonders if we want to try to do the Metallicity study.
WRT convincing people that we can do Type II, Peter's talking about a proof of concept 1-week baseline search for Type II supernovae, in the V-band. Greg asserts that this is silly to send to Subaru, because the exposures are short and the fields are huge. Greg suggests moving the deepsearching over to Subaru and the Type II to CFHT... eventually. This semester? Unclear.
How about sneaking in the Type II project rather than writing a proposal for it. Greg worries with this about the confusion with telescope operators and exposure times and filters.
Perhaps we could ask for the Type II search at CTIO. You would need at least 2 nights, 7-10 days apart. Followup...sneaked in, or maybe, or maybe, or maybe... we'll never settle and focus. Maybe you'll get the spectra at VLT. Or, it can be a separate Richard Ellis proposal.
Oh it's messy. We also need to talk to Richard Ellis to find out what it is that he's interested in doing. We're up to something like three proposals. J AO photometric followup, Spectra of z>1 supernovae, and then CTIO Type II supernovae. We may put in two ourselves; Saul asserts that he's told by Joe Miller that if we put in two strong proposals, we will get more time. Rob doubts it on general pessimism. Brenda doubts it on historical precedent.
Re: the Type II again (these are as big a mess as the discussion), do we want to get spectra at Keck or at VLT, perhaps in coordination with Mark Phillips.
Greg points out that if we want to use the results of this search for the next round of proposals, we have to run this search in February.
Manpower issues... maybe we should let the Type II thing slide, and propose it for Fall of 2000.
Or maybe we try to weasel time out of Subaru for a pilot study, and maybe have Richard have some spectra time right about then. That way, we can do the search without having to send somebody to Chile for 10 days.
So.... we talked a lot and I didn't take notes. If we're gonna put in for HST, proposal due Sepetember 2000, to do II's, we should do the pilot study in the Spring. So we have to decide if we need HST, or if it's good enough. Peter has to talk to Ed and do the work to find out how good the Type II photometry has to be.
Metallicity. It sounds like we don't know the nearby supernovae yet to do much with a high s/n 0.5 redshift spectrum. However, Saul thinks we should mention this to Richard as something he might be interested in looking into.