This article describes the derotation and photon masking phases of data processing. The FIMS and SPEAR teams both followed a similar procedure for these steps.

On this page...

Derotation

To minimize the effects of the detector anode wire differential non-linearities (DNL; see Flat Fielding), the anodes were mounted at a 15 degree angle with respect to the dispersion direction. Spectral features thus cross several anode DNL bands, and, to first order, the effects of the DNL average out to zero. However, at any given location of the detector, a spectral feature may be shifted by the local DNL (spectral lines look like twisted wires; see figure below). Dividing by the flat field removed these local shifts. Before collapsing the 2D XY data into 1D spectra, the data first needed to be scaled such that a pixel in X is the same physical size as a pixel in Y, then rotated such that the dispersion and imaging axes are orthogonal. A rotation angle of 14.9 degrees was empirically determined to yield the sharpest spectral features in the center third of the detector (where optical aberrations are at a minimum).

Auroral spectral features on detector.

Spectral lines in raw L band observations of the terrestrial aurora. The dispersion axis goes diagonally down-right and is labeled with "λ". The imaging axis goes diagonally up-right and is labeled with "φ".

Photon Masking

The columns BAND ID and MASK were added to the events file. For stim signals, BAND ID = MASK = 0 was assigned. BAND ID = 1 and BAND ID = 2 were assigned to the long-band and short-band photons, respectively. For photons incident onto the predefined, good locations on the detector were assigned to have MASK = 1 for long-band photons and MASK = 2 for short-band photons. 

Data Use | Acknowledgements | DOI | Privacy

Send comments & corrections on this MAST document to: archive@stsci.edu