Date: Wed, 29 Sep 1999 13:55:56 -0700 From: Marcia Rieke Message-Id: <199909292055.NAA04020@ircamera.as.arizona.edu> To: aswg_nir@hst.nasa.gov Subject: NGST Near-IR Subcommittee X-ListMember: rfosbury@eso.org [aswg_nir@hst.nasa.gov] Status: RO X-Mozilla-Status: 8001 X-Mozilla-Status2: 00000000 X-UIDL: e31f0377f2e4c156744b4f34cebb55ee Dear Near-IR Spectrometer Committee, Please forgive my slowness in getting our committee into motion. First, I want to urge all of you to read the IR-spectrometer ISIM reports. The U.S. reports are already available at http://ngst.gsfc.nasa.gov:80/science/isimpage.html and the ESA- and CSA-sponsored reports are promised by Oct. 5. I would like to schedule a telecon soon after all reports are available -- say on Oct. 8 or Oct. 11. Please let me know your availability on those dates at ~15:00 GMT. I will expect everyone to have read the reports by the telecon time so we can have a substantive discussion leading up to a request to the technical panel (membership listed at http://ngst.gsfc.nasa.gov/cgi-bin/pubdownload?Id=510 ) for an evaluation of the spectrometer designs that we are most interested in. I would also like to have the ancillary information that we identified as being needed at Hyannis in hand so if you are listed for a task below, please gather the information ------------------------------------------------------------ Task List Bob Fosbury -- Groundbased sensitivities of optical and near-ir spectrometers on 8-meter telescopes Simon Lilly -- Information on relative usage of CFHT in long-slit and multi-object modes Gardner/Ferguson/Arribas -- Consensus on faint number counts ------------------------------------------------------------ and post it to the group by sending it as an e-mail to aswg_nir@hst.nasa.gov. I will be asking the technical panel to supply us with their relative comparisons of dispersive multi-slit designs, multi- mirror slit designs, image slicers designs, and FTS/dispersive designs in terms of 1) Mass and volume estimates 2) Scattered light levels within the spectrometer 3) Ability to have both low- and high-resolution modes in one instrument. I will also ask Matt Greenhouse to ask each of the spectrometer teams to outline to us how they would observe 1) a field of ~100 faint galaxies at R=500 distributed over 4'x4'; and 2) ~10 extended sources at R=3000 distributed over 4'x4' with their outline to include all the steps needed for removing the instrumental signature, flux calibration, and wavelength calibration. I won't be sending this to Matt until Friday so if you would like the request modifed, please let me know. Last, Simon sent out a list of good issues about instrumention and I've attached the message below in case you do not already have a copy. Waiting to hear from all of you, Marcia =============Simon's message============================ (only ASCII version attached here) Dear fellow ASWG'ers, I am sure many of you felt as bewildered as I did at the complexity of the task facing us in the next couple of months. To focus my own thinking I drew up a list of twenty questions that I will myself be using to try to make my own personal decision as to what will be best as our reccommendation for the conceptual NGST instrument suite. I am attaching them to this message (as a WORD97 file) in case you are interested or want to comment and also reproduce them below as ASCII. These are also the questions that I think I would most like to hear from the Technical Panel about. Matt, if you feel it is appropriate, please forward this to the Technical Panel (in distilled form if need be). Many of them are "cost" questions. I realise that these are uncertain, obviously, but I have tried to pose them in relative terms. Furthermore, this is a cost-capped exercise so we cannot duck the question of costs, otherwise our recommendation will be meaningless. Also, if the study P.I.'s have answers to any of these, I would be interested in hearing from them as soon as possible. Any comments welcome, Simon ----------------------------------------------------------- 20 Key Questions on NGST instrumentation Simon Lilly A. The FTS as an imager 1. What is the cost (i.e. complexity) increment of a standard (i.e. non-dispersed) FTS-with-filters over that of a pure filter imager with the same field of view/number of pixels? 2. What are the failure modes of the FTS mechanisms, to what extent can these be made redundant, and how does the scientific performance degrade with these - i.e. in particular how robust is the FTS and could basic imaging performance ever be "lost" through an FTS failure? 3. What benefits (e.g. redundancy, CR removal, deep panchromatic image) come from the duplicate-detector to offset the obvious 50% "inefficiency" suffered by a dual- or single-port FTS relative to a single detector pure filter imager in the case of simple filter imaging. B. An optical channel 4. What is the cost of an optical-optimized module (say using Si PIN detectors) relative to a basic NIR-only camera (assuming the same number of pixels). 5. Is it possible to have one set of detectors operating efficiently over a decade of wavelength or does sensitivity at 0.5 mm require different physical detectors to those at 5 mm even if using the same detector material? 6. If it is assumed that you require different plate scales over the 0.5-5 mm range, then is this more cheaply done by "zoom" optics with one detector mosaic or by multiple detector arrays (modulo answer to previous question). C. The Guider 7. What are the scientific drawbacks of using the science detector for guiding (c.f. the yardstick design). 8. What is the approximate cost of implementing the science-detector- as-guider relative to the cost of a separate unit (which would presumably not to be counted as SI)? D. A Coronagraph 9. What is the cost of the simplest coronagraph that is not tied to a particular primary mirror architecture? 10. What are de-scope options for a coronagraph? E. Dispersed spectrograph questions 11. If both MOS and IFS are thought to be definitely required, can these effectively be done in one spectrograph and what are the cost implications of doing it in one or in two? 12. How serious are the "set-up" concerns, especially for non-imaging MOS (e.g. fibre slits) but for other front-end MOS concepts too. 13. Is it not possible to implement fibre bundles at the ends of a limited number of 2-d articulated arms (the dream-solution)? 14. Given that emission line searches go substantially deeper than continuum searches (W10 for R=100 and an EW=10% line) how best to do the ultra-deep searches for emission lines (omitted from the DRM)? 15. How serious is the ground competition going to be for R > 1000? F. The dispersed FTS as spectrograph 16. How much complexity is added to the imaging FTS if a MOS/IFS front-end and low dispersion prism/grism are added? What is the cost of this hybrid relative to: (a) a non-dispersed non-MOS FTS? (b) a conventional MOS/IFS? (c) (a)+(b)? 17. What are the implications of the higher multiplexing gain of the dispersed FTS for the MOS/IFS front-end? 18. Given that the attraction of dispersed-FTS is because dispersed- MOS/IFS is seriously detector limited, what are realistic expectations for dark current and read noise. G. The mid-infrared 19. How much for what, relative to near-IR instruments? 20. Is an FTS ruled out by the sharply increasing background?