Posted by Esther Hu on July 06, 1999 at 10:33:22:
Dear Olivia, Since there hasn't been much on the Web discussion page I thought
I would send in some commments on the WFC3 filters directly to you. o First, I suggest that you put the information on the Advanced Camera
filters (maybe even STIS) on a link to the Web page, so users can
compare the filter complements directly. o My interests are in using filters for very high redshift galaxy work.
The narrowband filter complement for WFPC2 was never well suited to
this purpose, since most of the narrow filters corresponded to strong,
zero-redshift lines and the medium bandpass filters would reduce the
detection sensitivity. (Question: my impression is that the current
manufacturing techniques should yield much higher throughput narrow
filters than was possible with the WFPC2 filters; is this correct?)
Will the narrowband IR complement of filters be better than the
filters available for NICMOS? There was an FeII filter (around 1.65
microns) where the actual line center fell where the filter's
throughput was only 40% of peak. Again, my own experience is that
recent IR narrowbands can also be made with higher throughputs (e.g.,
typically ~80% for 1-1.5% IR filters -- range of values 79% - 87%). o I'd like to see some coverage of the far red atmospheric windows
(e.g., ~8185 Ang and ~9135 Ang) because of the possibility of coordinating
programs with the new large ground-based telescopes. For faint objects
this means you can do the followup spectroscopy in relatively clean
areas free of strong night-sky, and study kinematics, profile behavior,
line ratios (in the case of doublets like [OII]), asymmetries, etc.
However, it's probably more important that you position some of the
medium (/wide) filters to cover this region, because of the difficulties
of doing deep continuum observations from the ground at these wavelengths
due to night-sky background. The range of science drivers: high-z
supernovae (H_0, q_0 determinations) now being pushed to redder
wavelengths. (IR is also of some interest for this, but I don't think
you can get the precision measurement at, say 1 micron, with the
current detector performance, even trading off against QE at slightly
lower optical wavelengths). Spectral diversity studies in outer
solar system bodies. Studies of extremely red objects and submm
sources. Planetary systems and sub-solar mass stars with molecular
band signatures.
o Some more things to think about: the Sloan z' was originally planned
(at least in part) for the high redshift quasar population (redshift
range 4-5 and above), and folding in both the atmospheric background
and detector sensitivities at long wavelengths. It may be worth
rethinking more moderate bandpass options at different (and redder
settings) for the current classes of scientific problems, and take
into account: a) the limitations in color discrimination for high
redshift galaxies and quasars (particularly beyond redshift 5.5),
highly reddened objects, and varieties of low-mass stars, where
precision photometry (from HST) can be key, and b) background sky
level vs. filter bandwidth tradeoffs are not such an issue. o More on numbers later, but for narrow-band emission luminosities,
(H alpha, [O II], Ly alpha) as a function of redshift, see Fig. 5
in ApJ 502, L99 (1998 Hu et al.)