Chapter 3 provides details on several IRAF/STSDAS tasks useful for displaying and analyzing spectroscopic data; these tasks include splot, mkmultispec, fwplot, and sgraph. Following are brief descriptions of several more tasks that are useful for investigating or improving FOS data quality. In all cases, the on-line IRAF/STSDAS help file contains additional practical information.
33.6.1 deaccumThis task unbundles individual groups of an ACCUM multigroup spectrum. In ACCUM operation the data are readout at regular intervals and the output after each readout is stored as a separate group containing the accumulated sum of all readouts. The product of the deaccum task is a multigroup spectrum in which each individual group contains only the signal accumulated during each individual readout interval. This data format is very useful when the groups are plotted together for inspecting the effects of noisy diodes or poor guiding during the exposure. This check is always a good idea to do as it allows inspection of the entire spectrum as opposed to the more limited paper product group count diagnostic.
We recommend running deaccum with all of your ACCUM mode .c1h files. It will also work with statistical error (.c2h) files and correctly unbundles these data in quadrature. This task will not work for data files that contain wavelengths (.c0h) or data quality values (.q0h or .cqh files). The group parameter EXPOSURE is updated in the output image headers to reflect the change in exposure time associated with each data group. For example, to unstack the calibrated flux data in the image y3ci0203t.c1h, and write the results to the new image y3ci0203u.c1h, use the command:
fo> deaccum y3ci0203t.c1h y3ci0203u.c1h
33.6.2 gimpcorThe size of geomagnetically-induced motion can often be important when analyzing photometric losses due to image drift off the diode array and when assessing the widths or wavelengths of spectral features.
Since the calfos (pre-onboard) GIM correction is applied on an integral pixel basis in the x-direction only, uncorrected motions of up to +/- 0.5 pixels can occur in X. This effect is much smaller for observations corrected by the onboard GIM correction. For observations obtained prior to the implementation of the onboard GIM correction (the presence of header keyword YFGIMPEN set to "T" tells you if the onboard correction has been performed), the actual decimal magnitude of GIM motion can be evaluated with STSDAS task gimpcor. This information can be useful for evaluation of the GIM-related uncertainties in observing modes to which the correction is not applied, as well, such as pre-onboard GIM-correction spectropolarimetry and IMAGE mode.
This task calculates the x-direction GIM correction in fractional diodes predicted by the FOS GIM model. Rounding of the output to the nearest integer gives the actual correction that would have been applied by calfos in normal pipeline processing if the OFF_CORR switch were enabled.
The size of the y-motion, which is important for the evaluation of photometric losses, is also calculated. Recall that no GIM correction is made in the y-direction prior to the implementation of the onboard GIM correction.
gimpcor is essentially a small version of calfos, but gimpcor does not actually perform any corrections. All inputs required for running calfos are also required for running gimpcor. The image motion corrections are output in diodes, not pixels, for x and Y-base units for y. For example, the following command can be used to show the magnitude of the GIM correction for the observation y0nt0303t (an observation made in RAPID mode, with each line corresponding to each readout of the science diodes). Note that components of the geomagnetic field are also given along the V1, V2, and V3 axes.
fo> gimpcor y0nt0303t
Y0NT0303T XOFF YOFF B(V1) B(V2) B(V3)
33.6.3 unwrapThe FOS accumulators will overflow if more than 65,535 counts are recorded in any individual readout interval. In many cases an overflow can be corrected by the unwrap task. All pixels greater than a threshold value will be examined for wraparound. This algorithm is not foolproof, especially if more than one wraparound has occurred. Some experimentation with the threshold value usually results in a good correction for reasonably smooth continua or emission lines. An alternate IDL algorithm that is somewhat more robust than unwrap is available on the FOS WWW page in the Calibration Tools section.
fo> unwrap y20vk10gr.d0h uw10g.d0h thresh=20000.
33.6.4 waveoffsetDetermination of wavelength offsets between two spectra can be made with the STSDAS routine waveoffset.
This routine computes the offset of one spectrum from a reference spectrum, as a function of wavelength and diode position in the frame of the reference spectrum. The input spectra may be divided into a number of equally sized bins. Each bin in the target spectrum is cross-correlated with the corresponding bin in the reference spectrum and an offset is determined. The offsets for all bins are placed in an output table as both diode and wavelength offsets.
The following example will compute the offset between upper and lower apertures for the spectra in file h57red. The upper aperture is in the first group of the file and the lower aperture in the second group. The wavelengths for the first group are stored in wh57red (group 1). Results are written to a table called offsets.tab.
fo> waveoffset h57red wh57red h57red tab=offsets
Copyright © 1997, Association of Universities for Research in Astronomy. All rights reserved.
Last updated: 01/14/98 15:41:23