SWS Note

SWS Grating resolution and instrumental profile: Current uncertainties and a request for help

To obtain spectral resolution values for the SWS full resolution grating modes
SWS02 and SWS06, most users currently adopt either the figures plotted in 
the Observer's manual, or the RESOLUTION module inside the SWS interactive 
analysis (IA) which implements a simple model of SWS geometry and diffraction.

The purpose of this note is to inform SWS observers that these values, while
being good for most purposes, may still be inaccurate at the 20-30% level
The Observer's manual plots and the IA RESOLUTION results are also 
inconsistent, since they are based on different versions of the underlying
model. A preliminary report on measured line widths (van den Ancker, Voors 
and Leech 1.7.1997) has already suggested deviations from the model, in the
sense of better than modelled resolution especially for extended sources, and
a less pronounced resolution difference between point- and extended sources.
A program is currently underway to widen the empirical database, and to
implement improved reference data for the calibration sources used.
First results confirm the discrepancies found in the preliminary report.

As long as this program is not completed, users should be aware of the 
mentioned uncertainties which may be important in special cases, e.g. for 
deconvolution or for attempts to infer source sizes from line widths.

SWS observers can help this calibration program by pointing out suitable
observations in the ISO archive, beyond those known to the SWS instrument 
team. We are currently using planetary nebulae, some PDRs, and
planets, but a wider observational basis is important to eliminate the 
effects of uncertain knowledge of the properties of individual sources. 
All we need is a pointer to suitable observations which should meet the
criteria described below, and for which the listed information should be 
available. Please send any information to lutz@mpe.mpg.de

Criteria for suitable observations
----------------------------------
Any SWS02/SWS06 observation could be used as long as
- lines are not too faint. Any observation which has lines clearly
  discernible in the dot cloud of the unprocessed AAR is ok, even if it still 
  needs e.g. flatfielding to get a good profile
- The intrinsic linewidth is small. In practical terms either a very small
  but not well known linewidth (like in some PDRs) is ok, or a somewhat
  larger linewidth up to about 50km/s FWHM, as long as it is known. 
  Some planetary nebulae are examples of this second category.
  Almost all galaxies for example are useless because of too high intrinsic
  linewidth.
- The spatial structure is known in the form of 'pointlike' (<3arcsec FWHM),
  'extended' (>1arcmin), or some intermediate diameter. Of course, this
  will be often a best effort approximation - this is why we are heading for a
  large sample!  

Information needed in order to use an observation to determine SWS resolution
------------------------------------------------------------------------------
* Source name
* TDT number of the observation (or several TDT numbers if repeatedly observed)
* Size (should be FWHM diameter in arcsec, not 'diameter of faintest wisps'), 
  and a reference for that. Of course, additional comments like 'this is an 
  east-west elongated filament, see X et al. ApJ xyz' are welcome.
* Intrinsic line width (FWHM in km/s) and reference for that.
  Again, we appreciate comments like 'Don't use the X-line because it is 
  blended with the Y-line'