IRAF tasks for error propagation and atmospheric model fitting??
Frank Valdes wrote on Oct 27, 1998
----- Begin Included Message -----
From: colgan@cygnus.arc.nasa.gov (Sean W. J. Colgan)
Subject: some basic iraf questions
Date: 22 Oct 1998 17:18:18 -0700
Hello -
I am part of a team building an echelle spectrometer (AIRES) for
SOFIA. I am looking into using IRAF for all or part of our data
reduction. AIRES will work in the far-infrared utilizing two
dimensional detector arrays and will output FITS files. Because our
detector arrays are made of discrete detectors, each pixel will
have a different sensitivity. Hence, the FITS file will contain
a data array, an error array, and probably a data quality array
to keep track of cosmic ray hits. I have looked through the IRAF
materials on the web and talked to some colleagues. The only
IRAF tool I found which propagates error arrays while carrying
out basic arithmetic operations is msarith. Before focussing in too
much detail on msarith, I was wondering:
--> Are there iraf tools besides msarith which do basic arithmetic
operations on FITS images while propagating an error array?
As part of our operation, we will measure water lines and then fit
these data with an atmospheric model. The atmospheric model takes in a
set of parameters (eg. how much water is above the observer), and
produces a transmission spectrum. This spectrum can't be easily
represented by a simple mathematical function, but is better given
numerically as a series of (wavelength,transmission) pairs.
--> Is there an IRAF tool which fits a series of numeric models to data
(in a FITS file) and chooses the best numeric model?
Thanks!
+----------------------------------+-------------------------------+
| Sean W.J. Colgan | |
| MS 245-6 | colgan@cygnus.arc.nasa.gov |
| NASA/Ames Research Center | Phone: (650) 604-0420 |
| Moffett Field, CA 94035-1000 | FAX: (650) 604-6779 |
+----------------------------------+-------------------------------+
----- End Included Message -----
----- Begin Included Message -----
>From valdes Tue Oct 27 10:02:03 1998
Date: Tue, 27 Oct 98 09:56:37 MST
From: valdes (Frank Valdes)
To: colgan@cygnus.arc.nasa.gov
Subject: Re: some basic iraf questions
Hello Sean,
> --> Are there iraf tools besides msarith which do basic arithmetic
> operations on FITS images while propagating an error array?
As you realize there are several major subpackages in IRAF. In
particular, NOAO, STSDAS, and XRAY. I am mostly knowledgable about the
IRAF core system and NOAO (iraf@noao.edu is the support site for the
IRAF core system and the NOAO packages but not for the HST and x-ray
packages). In these packages there is currently little that propagates
error arrays. We have a future initiative to address this. The NOAO
package is the most general. It does support spectroscopy, including
echelle data, and is works with most generic data. The other packages
are more instrument specific; i.e. for HST, ROSAT, AXAF, EUVE. These
may be more likely to support error arrays but possibly be fairly
specific to their data. So I think the answer to your question is that
currently there are no generic or "off the shelf" tools that handle
error propagation in a general way for you to use. But as a project I
would think you would want to build your own package anyway and IRAF
would be a good choice. The examples of NICMOS, or AXAF, or FUSE
would be helpful. You might want to query the authors of MSARITH
(really HST) about their experiences and whether they know of useful
tools.
> --> Is there an IRAF tool which fits a series of numeric models to data
> (in a FITS file) and chooses the best numeric model?
Again there may be something well suited for this in one of the other
major packages. In fact I would be surprised if there isn't something in
the XRAY package. But in the IRAF core system and NOAO there is nothing
that would directly deal with this. There are some sophisticated software
libraries for fitting data that could be used for this. Existing tasks
could be easily copied and modified to include numeric model fitting.
This is more a packaging issue as to how you would want to input the models
and how you would want to present the results. As in the previous question,
you will not find a package the perfectly suits your needs and you would have
to develop some software yourself. If you were to develop a good fitting
package for atmospheric modeling in a general fashion, it would be valuable
contribution that would find uses for many IR systems around the world.
IRAF contains a very large body of software and you will need to search
it yourself as I'm not sure anyone would know all of what is possibly of
use. Talking to the various major groups is probably the best way to
go about this. I will post your question and my reply to the
adass.iraf.applications newsgroup which, I believe, will reach software
developers at the other sites as well as some users around the world.
(See the IRAF home page http://iraf.noao.edu/ for information about the
ADASS newsgroups and the IRAF subgoups.
Cheers,
Frank Valdes
NOAO/IRAF Group
----- End Included Message -----
Last post on Oct 27, 1998