We start with some introductions. I failed to note down the names of everybody present.
Afterwards, Saul says that one of the main content things we want to get done is figuring out what science we want to pursue in the telescope seasons during the next year. What's more, we want to think about the next two years, as HST proposals are going to be due at the end of the summer. We want to think about what resources we have to observe with over the course of the next few years. We'll begin with Isobel talking about Gemini, then Chris talking about VLT, and we'll mention Keck and Subaru and other potential telescopes. There will at some point be a break to figure out what things we need to look up, and then reconvene to share information.
Saul says that another task of this meeting is to do some group tuning. As the group gets larger, and there are groups forming in other parts of the world, we're going to have to formalize and write down some rules of how the group works.
Saul is hoping that tomorrow we'll finish up planning for this year and next year, and then move on to questions of where we are in different parts of the project. Also, some discussion of what things we need to do where we have nobody working on it.
Isobel, who works for the UK aspects of Gemini (UK has 25% of Gemini), is going to tell us a little about it. She has a talk which was prepared to tell people what was available for the first round of proposals. She's going to rush through it to tell us some of what the deal is with Gemini.
There are two telescope; the one in the north (Hawaii) is just about finished and starting, and the one in the south (Chile) is about a year behind. The USA has 50% of the telescope, and UK has about 25% of the telescope. Greg asks about what he saw was 30 nights which was set aside for Gemini staff in Hilo (not based on country), plus apparently a fair amount of director's time. There's a lot of engineering time in the first semester as well. There is also a host country share, which isn't counted into the country breakdown that Isobel shows us. Isobel notes that the deadline for different countries for proposals are different.... The proposals go in a couple of phases. The first one is done by country; the second phase gets the nitty gritty stuff. It sounds like the Gemini committee itself (or somebody) is going to have to merge the propsals as well (just scheduling, or actual allocation??).
The emphasis on the whole thing is image quality. They're already getting almost diffraction limit (0.07") at two microns (capable of 0.09", though that's not always achieved). Even without AO, the image quality is supposed to be better than Keck, primarily (presumably) due to thermal control. The dome is very open.The observing will be split 50/50 between normal observing mode and queue obesrving; it's not yet decided how this is split up.
She shows a list of instruments; she has a list somewhere of when they're all arriving. At the moment, there's an 1024x1024 IR camera QUIRC with an AO system, using a HgCdTl chip. NIRI will be a near infrared imager. GNIRS will be an IR spectrograph. CIRPASS will be a fiber-based integral field spectrometer; Greg recalls that this one will have a very low throughput, perhaps making it unusable for faint objects. GMOS will be an optical instrument (multi-object spectrograph, including an imaging mode with a 5' field), that's suppoesd to go to Hawaii at the end of this year. ALTAIR is an AO system that will feed any instrument (including GMOS, though ALTAIR won't work very well in the optical).
A picture of the back of the telescope shows that four or five instruments can be mounted at once. Queue nights might be able to quickly flip between them, although there are issues of calibration overhead.
She shows a dedication image of Globby NGC6934, with and without AO. (Without in optical, with at K (2 microns).) You can see the diffraction rings in the AO corrected image. The AO system has a 35-element correcting deal, plus a tip-tilt secondary. The K camera is 1024x1024, with a 20" field of view. QUIRC wasn't really designed for gemini.
To use the system, you need a guide star with mag<15 within 30" of your target (mag<18 for reduced coorection). Degradation with distance from guide star (and with guide star magnitude) is on the web somewhere. See the Gemini web page... I haven't tried to copy the table down here. Some 1-hour 5-sigma key limits include J at 23.9 (i.e. 4 hours for 10-sigma), with seeing of 0.07 (Strehl 0.06) due to AO.
Isobel notes that there is a plan to use a laser guidestar, but that those plans are off several years. Apparently it's rare (of order 1%) to find one of our supernovae within 30" of a 15-magnitude star. We may have to do a search that overfinds supernovae, to make sure we get some near a usable guide star. There are different numbers being bandied about; Saul says that you only need to overfind them by a factor of 5.
Looking at delivery times, it looks borderline at best to propose for NIRI and GMOS for Spring of 2001. QUIRC will still be online, but it's difficult to propose for doing actual photometry using AO (because it hasn't been demonstrated). Greg asks if there will be an AO system feeding a spectrometer, since we won't need photometry for that. CIRPASS is the first thing that has that, but Greg repeats his worries about the very low throughput of that instrument.
NIRI will have three image scales, giving 120", 50", and 23". There will be a low resolution spectroscopy mode, that may be of interest to us. (R~1000-9000) At the low end of that resolution, you'd get the whole J-band at once (assuming a resolution element is no more than 2-3 pixels). Peter notes that at moderately high redshifts, the silicon feature would be smack in the middle of that.
CIRPASS will be an OH rejection spectrograph, an IFU with 500 fibers. The resolution will be 3000-26000 (with a 2K array). Image scales vary from 0.05" to 0.36". Isobel notes that GMOS is going to have a fiber array as well, theoretically due to arrive at the same time as GMOS. Greg guesses that the throughput will be circa 50%, but we shall see.
The preducted S/N~5 in 1 hour is J~19.4 for the central fiber on a point source with 0.36" lens scale. But is this per resolution element? If so, then it's actually pretty good.
GMOS will be a 0.36-1um (though yukky beyond 0.8um) imager and spectrograph with MOS capabilities. It is the same sort of thing as LRIS is now. We should get better image quality (maybe 0.3" vs. 0.6"). Of course, one limitation with all of this is that we can't currently centroid our objects that terribly well. The imaging field of view will be 5.5'x5.5'. The pixel scale is 0.72"/pixel, in three 2kx4k chips (with vignetting at the ends, and 37?-pixel (1/2mm) gaps in between the chips). (Greg wonders if the optics could be adjusted to center the image on two chips rather than three, so that there is only one gap and so you have one fewer chip to reduce.) There will be a flexure control system; in principle, where the CCD is mounted can be moved to compensate for flexure. The IFU will have a lenslet array with 1000 elements, plus a 500 element background field. The FOV will be 5"x7".
Greg wonders what will be considered a "big" proposal - will one night be a lot? Or will we need to apply for something like a few hours? Greg notes, for instsance, that 3 nights seems to be a hard limit at Keck; we never get more than that.
Chris gives a lowtech presentation using last years slides....
He starts talking about La Silla, which is of the most interest to the low redshift search. He lists telescopes and instruments. He talks about the ones of interest to us. SOFI at NTT (3.6m) is the instrument the other group used to follow up their SNe in the J-band (back in 1998). This was before ISAAC was online at VLT; that is now the one to use. Image quality is better at ISAAC too.
The 3.6m EFOSC II is one that we've used. It has good sensitivity in the blue, and Chris expects good results in the blue end spectroscopy from the low-z supernovae that were observed with this camera.
What happens to the 2.2m is what may interest the low-z work the most. Current, there's a wide field imager (WFI) with a 40' field of view on there, using 2x4 chips (2kx2k each) This is scheduled to go offline in 2002, when the VST (VLT Survey Telescope) is supposed to come on. It's not clear if the WFI will even exist after that. The 2.2m team leader doesn't know what's going to happen to the WFI. (The VST will be a 2.5m telescope with a wider field (1 sq. degree?) and better instrument quality.) It may be that the 2.2m will become completely free in 2002. Chris says that it has the best image quality of everything at La Silla, except for the NTT. Chris notes that ESO is looking for people who can come in and afford to run telescope; the 90cm scope was closed in 1998, but somebody came in this year to open the thing back up and run it. (The 2.2m is probably something like a million dollars a year to run.)
Chris shows the same VLT transparency as last year. FORSI is the instrument on UT1 that we used last year to monitor Ravel (z=0.87 if memory serves), and that we've used this year for spectra of Beethoven. It has a Tek2048 CCD, which isn't very sensitive in the red. However, the fringing with this CCD is less than with the modern thin CCDs, and the flexure of the instrument is very small. By moving the object on the slit, you can make a good fringe frame. (At Keck, this is harder to do, because the flexure moves the fringes a little.) In the future there will be FORSII. It's a copy of FORSI, and will be on the 2nd 8m telescope. ESO is looking to upgrade the CCD in that one to something that's more red sensitive (e.g. an EEV 2kx4k CCD). They're targeting this for perhaps March/April next year.
ISAAC is the other instrument on UT1, a 1-5um imaging spectrograph. It will use two different arrays for different wavelength regions. The 3-5um region is just coming into service now. This is the instrument which is being used to take images of Beethoven. He thinks we got S/N between 10 and 20 (with seeing 0.5"), z=0.55, Tint=1.5h, with the first observation of Beethoven. (There's a couple more coming.) There will be better for this; Conica will be on UT3, available in mid-2001. This will be an AO infrared instrument (using NAOS, the adaptive optics instrument), so the image quality can potentially be a whole lot better. The field of view will vary from 10"x10" up to 1'x1'.
Peter notes that we ought to try to do photometry with an AO system in the coming year to prove that we can actually do it. Chris says that Adonis can do AO on the ESO 3.6m.
UT2 started observing this year in April. It uses FORS2, which doesn't do polarimetry but does do cross-dispersed Echelle stuff, but is otherwise identical to FORSI.
UT3 and UT4 will open next year sometime; FORS2 will move to UT4.
UT3 is going to get an instrument called VIMOS, which is a multi-object focal reducing spectrograph. Its field will be 14'x14'. It's like FORSI, but with 4x the field of view and many more slitlets. It's basically a redshift machine, but also with an imaging capability. This one is going to be commissioned within the next few months. No longslit capability, only MOS. Chris isn't sure if there are plans to have an integral field capability for this one. With the active optics (which corrects for flexure, but not atmosphere), seeing could be 0.3-0.6", though it remains to be seen if that works over the whole field.
On UT3, there will be a visitor focus, for people who come and bring an instrument to run with for a little while. This will go on the Nasmith focus, and will have a field up to 20'. The telescope will provide rotation and guiding.
All of these telescopes will be done using service observing, as is done now with ESO.
Chris shows the schedule of Oct-Dec 99, when we were doing observations for Ravel with VLT. In blue, he shows visitor mode (not available for service observations). Green and red were service modes; red is when conditions didn't fit our requirments (clear, moon not to close, good seeing). Green are the nights that matched what we wanted. In October, there were only two nights which were good enough (and October was a good month). We got a 1h exposure. In November, we got two more exposures (2h each). The point is that even though you think that service mode gives you the flexibility to obesrve whenever you want, that isn't the reality of the situation. He points out that if this had not been a Class A service program, it simply wouldn't have been done. Chris warns us that service mode is as not flexible as you might think, and that we should keep this in mind while planning how we want things done.
Chris notes that visitor mode doesn't need to be allocated in full night chunks, but can be done in 1/3 or 1/2 nights.
Chris notes that we haven't had very good luck getting time at ESO in the last year or so. Even the Ravel program got rejected, but the director general still put it into the queue. We've had very little feedback on why we were rejected. Apparently the last time around we did mostly imaging, and they wanted us to do both higher redshift objects and also spectroscopy. In other words, they were trying to rewrite our science program.... The other supernova group hasn't gotten time on VLT either (although they did get some when we observed Ravel). So far we've had two proposals; the first was rejected but done anyway, the second was rejected. There's now a third proposal in right now (what is this one?).
It would be interesting to reduce the ravel results and see what we got out of it.
We've having a long discussion about how we might be able to get time from ESO, which I didn't archive here.
We're going to break for lunch and come back at 2:00. First, though, in the name of listing the telescopes we're going to discuss at 2:00, we'll discuss some of the other telescopes. Andy Howell mentions the Habey-Eberly telescope (with an effective aperture of 9.2m). It's right now being used at the 50% level, but the data is pretty bad (you can only go down to about 20th magnitude). The problem is stacking the 91 mirrors; the stack currently degrades after an hour, and it takes 30 minutes to restack them. They are going to get edge sensors so that they can keep them in alignment (as is done with Keck). The telescope also can't move all around the sky. It has a tracker, and can only track an object for about an hour. The soonest we'd consisder it for our high redshift work is a year and a half from now, though that may be a little bit optimistic. The seeing at that site tends to be pretty bad; typical is a little over an arcsecond, Andy says. Thus this telescope is probably off of the table for the current discussion, but might be good for z=0.5-0.7 work a year and a half or two years from now.
Saul says he just E-mailed Momoru (spelling?), the contact at Subaru. Subaru recently sent out a request for proposals for a short semester beginning in January. Momoru is puzzled about this; he isn't convinced that the mirror or the camera (Suprime cam or some such) will be ready by then. There is a possibility that something will happen this fall in coordination with what we're doing, but don't count on it until a half of an hour before the observation. There was some indication somewhere about some time being guaranteed for supernovae, and Momoru didn't know about that. Apparently, though, he has guaranteed for us two fields times one hour per field, separated by a month. (I may not have gotten that down right.) Apparently we're supposed to be putting in proposals with him.
Regarding Keck, Deimos is going to start up next spring. That team has 20 nights a semester. They're going to be concentrating on two fields, so if we ask for time when their fields aren't up. Their spring field is at 16h; we should figure out their fall field. (Mark Davis can tell us this.) Another constraint is that we want to be early in the queues. We also need to avoid the HST ACS servicing mission. All of this is pointing to early in the semester.
There are still questions as to where will Keck's AO be, as compared to Gemini and VLT.