The Ohio State University College of Mathematical & Physical Sciences Department of Astronomy

In response to this charge, the LBT Optical/UV Spectrograph Working Group (LBTOSWG) was formed by open invitation to all LBT consortium members. The LBTOSWG met in Columbus on 8-9 March 1999. Present were Bruce Atwood (Ohio State, MODS project manager and director of the Ohio State Imaging Sciences Lab), Ralph Belville (Ohio State, Imaging Sciences Lab), Paul Byard (Ohio State, MODS team), Darren DePoy (Ohio State, representing the LBT SAC), John Hill (Arizona, representing the LBT Project Office), Mike Lesser (Arizona), Jerry Mason (Ohio State, Imaging Sciences Lab) Klaus Meisenheimer (Heidelberg), Tom O'Brien (Ohio State. MODS team), Pat Osmer (Ohio State, MODS PI), Roberto Pallavicini (Palermo), Dan Pappalardo (Ohio State, Imaging Sciences Lab), Brad Peterson (Ohio State, representing the LBT SAC), Rick Pogge (Ohio State, MODS Project Scientist), Jesper Storm (Potsdam), Giampaolo Vettolani (Bologna), R. Mark Wagner (Ohio State, representing the LBT Project Office), and Rogier A. Windhorst (Arizona State). Jill Bechtold (Arizona) participated in part of the meeting by telephone. The following is intended to provide a summary of the results of the meeting.

The major conclusions of the LBTOSWG meeting are the following:

1. There was consensus that there are strong science drivers for retaining high UV throughput to atmospheric cutoff (about 320 nm). The LBTOSWG notes that this also provides a unique niche for LBT, as the comparable facility instruments for the other 8-10m class telescopes will not work below 380-400 nm. The LBTOSWG also notes that the overall response of the UV side of the MODS spectrograph can be significantly better than it appears to be in the MODS proposal if a more appropriate grating and CCD coating are used. The MODS team pointed out in answer to the specific question from the SAC about the cost of UV response to atmospheric cutoff that extending the response from 400 nm to 320 nm is a small incremental cost because MODS is using reflective rather than refractive optics.

2. MODS will produce high-quality (D_80 < 0.25 arcsec, where D_80 is the diameter that encompasses 80% of the light in a point-source image) images over a 4x4' field, and this was deemed to be acceptable given that an "extended field" option can be implemented; with reduced image quality (D_80 < 0.8 arcsec), a 5x6' field (i.e., a factor of 1.9 larger) is usable.

   Given that many of the proposed science drivers for the spectrograph are essentially "survey" type observations, it is obvious that, all other things being equal, multiplexing greatly improves efficiency. For the very faint quasars and galaxies that are MODS targets, increasing the field size over the originally proposed 4x4' size thus has a dramatic effect. However, there are a number of important barriers to implementation of a wider field at the f/15 Gregorian focus. First, the Gregorian focal surface is curved, not flat. Slit masks larger than 5-6' (or thereabouts) would have to be curved. Second, implementation of a significantly wider field would require a completely different spectrograph from MODS. Collimator aberrations will cause poor images far from field center, though as noted above, images comparable to those obtainable in median uncorrected seeing can be obtained throughout a 5x6' field.

   Two important points must be noted:

   1. The f/4 focus is really the wide-field focus for LBT. A wide-field spectrograph ought to be designed for the f/4 focus, but not at the present as the f/4 focus will not be implemented at first light.

   2. The niches for MODS are high throughput, broad wavelength coverage, and excellent image quality. An affordable instrument at the f/15 focus is simply not going to be competitive in terms of field of view. Relative to the field sizes of other wide-field optical spectrographs for 8-10m class telescopes, the MODS field is competitive only with Gemini (5' x 5').

3. It would be desirable for MODS to have a high resolution mode in the R = 15,000 to 20,000 range. However, very high resolution (R = 30,000 to R = 50,000) will require an echelle-type spectrograph optimized for single-object spectroscopy at very high resolution; attempting to meet the MODS science goals and achieve very high resolution will result in a design that does not serve any science goals well.

   There are at least two ways that the current MODS design can achieve higher resolution. One possible solution is to simply use a higher-dispersion grating. However, this will overfill the camera and thus decrease efficiency, though the design will need additional study to quantify this. A cross-dispersed mode will also be investigated. A second possible solution is to design a longer-focal length camera, either as an upgrade option, or as the baseline camera. The latter would be possible if in the R = 8000 mode on-chip binning (2 x 2) is used; a positive aspect of such a design is that the required design change would eliminate an obstruction (the detector) in the beam, and a negative aspect would be an increased cosmic-ray problem (4 times as many cosmic rays per binned pixel) and a gap between two sections of the spectrum that would result from the necessity of using two CCDs butted together.

   We also note that the higher resolution mode also provides an upgrade path to work with the adaptive optics (AO) system. The possible adaptive correction might range from rapid tip-tilt guiding, to partial (low-order) adaptive correction. In the unbinned mode, the camera pixel scale is 0.15 arcsec/pixel.

4. It should be possible to implement in the future an integral field mode. The LBTOSWG envisages an integral-field module that can be inserted between the telescope and the spectrograph. This is regarded as a future upgrade path, and the MODS design will preserve the integral-field capability as a future option. We note that in any case a conventional long-slit mode (4 arcmin) will be available.

1. Evaluate an unobstructed longer focal length (approximately 700 mm) camera for both sides (red and blue) and performance in both binned and unbinned modes.
2. Elaborate on the "extended field" (5' x 6') concept. Include commentary on the likely availability and cost of the atmospheric dispersion corrector for this larger format.
3. Provide updated throughput and quantum efficiency curves based on a range of realistic gratings and CCD coatings.

When the MODS team response to the action items outlined above is received, the LBTOSWG will continue discussion via electronic mail. Our goal is to have written recommendations on the MODS proposal and possible modifications and/or upgrade paths ready for the next LBT SAC meeting in early May.

Compiled by Brad Peterson
peterson@astronomy.ohio-state.edu
March 26, 1999