In this chapter, we provide a few references and examples of non-standard data types which highlight particular, but not frequently utilized, capabilities of FUSE. This information is primarily relevant to "Advanced" users even though "Intermediate" users might find it useful. This information might not be highly relevant to the "Casual" user.
Table 9.1 gives the measured full widths of Ly β airglow lines in the LiF1A segment. An extended source in the HIRS aperture has nearly the same resolution as a point source. It is a little less because of the astigmatic curvature orthogonal to the dispersion, which degrades the overall resolution. As with point sources, the exact resolution depends on wavelength and segment.
|Aper.|| Δ v|
| Δ λ|
|a For Lyβ on LiF1A|
When interpreting observations of very large extended sources (such as the Cygnus Loop or large reflection nebulae), it is also important to keep two additional facts in mind. First, for TTAG data there may be useful information in apertures other than the one specified by the observer. Spectra from these other apertures will not be extracted by CalFUSE by default, and one must return to the IDF file to extract these spectra. Second, to interpret the data from the other apertures, one needs to know their physical relationships (see Table 2.1 and Fig. 2.2). Together with the aperture position angles, these values can be used to determine the locations of the other apertures on the sky. Examples of the analysis of extended source observations can be found in the papers by Blair et al. (2002), Dixon et al. (2006), Sankrit et al. (2007a) or Ghavamian et al. (2007).
Extremely faint extended and point sources often require additional analysis, in particular to perform the most careful background processing and subtraction. Fechner et al. (2006) provide an example of detailed analysis of a very faint point source, while Sankrit et al. (2007b) and Danforth et al. (2002) examine the problems encountered with faint extended sources.
Moving targets were also observed by FUSE during its mission. Detailed descriptions of the data acquisition and data analysis specific to moving targets can be found in papers by Feldman et al. (2002) and Weaver et al. (2002).
Q: What is the best way to combine spectra of bright targets?
A: For bright targets, the goal is to optimize spectral resolution, so you will want to cross-correlate individual exposures on narrow absorption lines before combining the spectra.For TTAG exposures of moderately-bright objects, it is occasionally beneficial to splitlong exposures into shorter time segments and coalign the individual segments before re-combining.
Q: What is the best way to combine spectra of faint targets?
A: In this case, it is best to optimize the background model, so you will want to combine IDF files for the individual exposures prior to extracting the spectrum.
Discussions regarding both techniques (bright and faint targets) are
provided in the document "FUSE Tools in C" at
Q: How do I improve the background subtraction at the short-wavelength
end of the SiC1B channel?
A: The LWRS spectrum of the SiC1 channel falls very near the edge of the detector (see, for example, Fig. 4.1), where the background rises steeply. The scattered-light model used by CalFUSE, when scaled to match the counts in the center of the detector, does not accurately reproduce the background near the edge, and can cause an over-subtraction of the background, with the shortest wavelengths on SiC1B being most susceptible. In principle, one could use the dark exposure data taken near the end of FUSE operations to perform a custom background fit, but this would require specific expertise and would be a difficult task. In practice, one can compare the data from the SiC1B and SiC2A channels for consistency. Only very faint targets should have a concern with this issue.
Q: What if the stim pulses are missing for one or more of my exposures?
A: In most cases the STIM pulses were turned on briefly at the beginning and end of each exposure. However, some exposures don't include any STIMs at all. If that's the case, the cf_thermal_distort routine of CalFUSE will output a warning message to the trailer file (see Appendix C of this document). Since the thermal correction of the detector pixel scale is based on the STIM positions, CalFUSE has to make an approximate correction in these cases, based on typical STIM positions. This correction will only be approximate, and thus the wavelength scale may be slightly incorrect for such exposures. The errors will be largest at times when the temperatures were changing rapidly, such as after the detector high voltage state has changed. Unless you are really interested in the most accurate wavelengths possible, you should not have to worry about this. Alternatively, one could choose to drop such exposures from the total observation when the individual exposures are combined to assess any possible impacts.
Q: How do I find and retrieve the special case of bright-earth observations?
A: There are two types: (1) The S100 program, executed in the fall of 2007, obtained several weeks of downward-looking airglow spectra. (2) Throughout the mission (particularly early on), we obtained approximately 1300 bright-earth exposures during earth occultations of some 200 targets. All of these data have the program ID of the science program and target, but have exposure numbers of 901 and above (see Chapter 4). These data can be retrieved as special cases from the MAST archive.
Q: How do I extract spectra from non-target apertures?
A: This can only be done for TTAG mode data and the exact technique depends on whether the new target is an extended or a point source:
• Extracting Spectra from Non-Target Apertures: Extended Sources
• Extracting Spectra from Non-Target Apertures: Point Sources
Q: How to learn how to get the most out of the FUSE IDF files?
A: The best introduction to the IDF files is probably the User's Guide to cf_edit
http://archive.stsci.edu/fuse/calfuse.html, along with the discussion of the file format provided in Chapters 4 & 5 and Dixon et al. (2007). If heavy data processing is anticipated, "FUSE Tools in C" documentation should be consulted as well.