[fitsbits] frequency definitions

Arnold Rots arots at head.cfa.harvard.edu
Wed May 6 14:25:53 EDT 2009


A couple of additional comments.

First, there is no requirement that the reference pixel is at the
center of the IF band.

Second, there is no reason to assume that if TTYPE='FREQ-LSR' the
frequency at the center of the band is at the line's rest frequency.
More likely, it will be at the Doppler frequency of the object of
interest.

Third, if TTYPE='FREQ-OBS', it says that frequencies (OBSFREQ, if you
like) are expressed in truly observed frequencies;
if TTYPE='FREQ-LSR", the frequencies are labeled as if the observer
were moving with the LSR - i.e., the observer-to-LSR component has
been taken care of already, what remains is the objects Doppler velocity.

Fourth, although it looks like you are working on Galactic stuff in
the radio (and hence the radio definition seems appropriate), it may
be good to keep in mind that you formula:

                     (      Vdop )
OBSFREQ = RESTFREQ * ( 1 - ------)
                     (        c  )

is only correct for Doppler velocities using the radio definition.
In the optical definition V = c*delta(lambda)/lambda0, rather than
V = c+delta(nu)/nu0.
And, of course, one may prefer barycentric over LSR velocities when
working on extragalactic data.

  - Arnold


Bob Garwood wrote:
> Since I'm the author of that part of that note, I should probably chime 
> in here.
> 
> As mentioned elsewhere, the primary WCS papers are referenced in that 
> previous reply and cited below.  The SDFITS note in aips++ predates all 
> of that.  SDFITS has a long history, which I won't go into here.  It 
> remains primarily a locally implemented standard with a few version of 
> "local" that aren't quite in agreement for lack of effort primarily on 
> my part to try and reach a real consensus and, most importantly, 
> exchange data between "local" implementations. 
> 
> Prior to the WCS papers, lots of people obviously described a 
> frequency-like axis.  So you're going to find that historically there 
> are keywords and columns for which you'll just have to talk to the folks 
> that wrote the FITS file to be sure you go the interpretation right - if 
> there is any ambiguity in what you find there as compared to the WCS papers.
> 
> For your specific questions ...
> 
> CRVALx - I'd say "yes"
> OBSFREQ - 'probably'.  I think "center of the IF" is possibly ambiguous 
> enough that you might want to double check with the author/writer
> RESTFREQ - no.  I'd interpret that as equivalent to the "RESTFRQ" value 
> described in the WCS papers.  It's the rest frequency of the spectral 
> feature of interest - i.e. the frequency that that feature has when 
> observed at rest with respect to the thing producing that feature.  It's 
> the lab-measured line frequency.   It may or may not correspond to the 
> center of the IF, depending on how the IF was set up and if doppler 
> tracking was being done during the observations.  It should only be 
> necessary/useful if you want to convert from a frequency to a velocity 
> (I'm 99% sure of that statement, but I'm sure someone here will correct 
> me if that isn't true) - or rather, this quantity isn't necessary to 
> recover the frequency at each pixel nor is it necessary to convert from 
> a frequency in one reference frame to a frequency in a different 
> reference frame assuming that TTYPEx is a frequency axis.
> 
> Then you ask" What is the difference between FREQ-OBS and FREQREST?".  
> But I'm not sure what you mean by "FREQ-OBS".  You should an equation 
> earlier (badly formatted in my copy, but I think I understand what you 
> meant) showing the difference between OBSFREQ and RESTFREQ, so I don't 
> think you mean FREQ-OBS = OBSFREQ.  SDFITS doesn't use FREQ-OBS so I 
> think you'll need to ask the creator of that FITS file what they 
> meant.   Unless I missed something in a quick scan of the relevent WCS 
> paper, I don't see it defined there.
> 
> Cheers,
>   Bob
> 
> 
> 
> Tom Kuiper wrote:
> > Well, not specifically about dual sideband data but about radio 
> > spectral line data in general.  The best description of the relevant 
> > keywords that I've found is in the NRAO AIPS++ documentation at 
> > http://aips2.nrao.edu/docs/notes/236/node14.html.  However, the 
> > definitions are not entirely unambiguous.  That's why I posted my 
> > interpretation, to see if anyone disagreed.
> >
> > Regards
> >
> > Tom
> >
> > David Berry wrote:
> >> It looks like Tom's query was about dual sideband data, which is not
> >> covered by paper III. OBSFREQ, etc, are not in the standard. There are
> >> probably various different schemes in use out there. For instance JCMT
> >> and CLASS software use a system based on a keyword called IMAGFREQ,
> >> which holds the image sideband rest frequency (with RESTFREQ holding
> >> the observed sideband rest frequency).
> >>
> >> David
> >>
> >>
> >> 2009/5/5 LC's No-Spam Newsreading account <nospam at mi.iasf.cnr.it>:
> >>   
> >>> On Mon, 4 May 2009, Tom Kuiper wrote:
> >>>
> >>>     
> >>>> I'd like to get it clear in my mind what the various frequency
> >>>> definitions are.  Is the following right? If there is a place where
> >>>> all this is clearly laid ou, please refer me to that.
> >>>>       
> >>> Not a spectral WCS expert, but the standard and official references
> >>> are :
> >>>
> >>> - http://fits.gsfc.nasa.gov/fits_standard.html
> >>>   the latest release of the FITS standard, in particular chapter 8
> >>>
> >>> - the full WCS paper III, either as published or
> >>>   http://www.atnf.csiro.au/people/mcalabre/WCS/scs.pdf
> >>>
> >>>   (the other WCS papers are all listed in
> >>>   http://www.atnf.csiro.au/people/mcalabre/WCS/index.html
> >>>
> >>> --
> >>> ----------------------------------------------------------------------
> >>> nospam at mi.iasf.cnr.it is a newsreading account used by more persons to
> >>> avoid unwanted spam. Any mail returning to this address will be rejected.
> >>> Users can disclose their e-mail address in the article if they wish so.
> >>> _______________________________________________
> >>> fitsbits mailing list
> >>> fitsbits at listmgr.cv.nrao.edu
> >>> http://listmgr.cv.nrao.edu/mailman/listinfo/fitsbits
> >>>
> >>>     
> >>
> >> _______________________________________________
> >> fitsbits mailing list
> >> fitsbits at listmgr.cv.nrao.edu
> >> http://listmgr.cv.nrao.edu/mailman/listinfo/fitsbits
> >>   
> >
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > fitsbits mailing list
> > fitsbits at listmgr.cv.nrao.edu
> > http://listmgr.cv.nrao.edu/mailman/listinfo/fitsbits
> >   
> 
> _______________________________________________
> fitsbits mailing list
> fitsbits at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/fitsbits
> 
--------------------------------------------------------------------------
Arnold H. Rots                                Chandra X-ray Science Center
Smithsonian Astrophysical Observatory                tel:  +1 617 496 7701
60 Garden Street, MS 67                              fax:  +1 617 495 7356
Cambridge, MA 02138                             arots at head.cfa.harvard.edu
USA                                     http://hea-www.harvard.edu/~arots/
--------------------------------------------------------------------------



More information about the fitsbits mailing list