Previous: Observing with FOS
Up: Observing with FOS
Previous Page: Observing with FOS
Next Page: Spectropolarimetry
Data reduction typically involves the following steps:
- Debiassing and flat-fielding the data frames. Flat-fielding is particularly important for removing the many poorly-resolved sky lines redward of 7200 Å (for this reason, the resolution of ISIS is preferred if wavelength coverage and throughput are adequate). The tungsten lamp is so red that flats are saturated in the red before enough signal is obtained in the blue; twilight flats are better.
- Correction of curvature, extraction and sky-subtraction of first- and second-order spectra for each object on slit. Sky-subtraction can be `top-hat', using a few equally-weighted columns (preferably on either side of the object spectrum) for sky and a few equally-weighted columns for object; ot it can be optimal, with the software weighting each object column according to the strength of the signal. In practice, optimal extraction typically yields only a few 10s of % extra signal-to-noise for an unresolved object.
- Wavelength calibration, via arc exposures (CuAr or CuNe, former probably better, has lower density of lines). FOS is very stable, and wavelength calibration can anyway be checked against the numerous sky lines, but it's a good idea to take a few arcs per night.
- Correction for absorption bands due to the atmosphere (especially the Fraunhofer A and B bands at 7594 and 6867 Å), via observation of a star with a featureless spectrum, usually a B star such as Feige 34.
- Correction for atmospheric extinction at the airmass observed, using the theoretical curve from La Palma technical note 31.
- Correction for the spectral response of FOS, via observation of a spectrophotometric standard star with a wide slit, to give e.g. Jy through the slit as a function of wavelength. FOS' wavelength response is surprisingly bumpy, with features of strength 5% on a scale of 200Å.
- Smoothing of second-order spectrum to resolution of first-order, and merging of first- and second-order spectra into one.
Thus, as well as spectra of targets, observers may want to take bias frames, twilight flats, tungsten-lamp flats, arc-lamp exposures and spectra of B-stars and spectrophotometric standards.
Much of the above can be carried out using FIGARO software. The software available for spectrum extraction (SCP and SAM) is described in Starlink User Notes 148 and 149.