This article describes the detector background subtraction phase of the data processing. The FIMS and SPEAR procedures were conceptually similar but less detailed procedure.
On this page...
FIMS Procedure
The FIMS-SPEAR data are affected by instrumental (detector) background due to cosmic rays, radioactive decay within the detector, and thermal charged particles entering the instrument. Somewhat counter-intuitively, the detector background is neither temporally constant nor spatially uniform across the face of the detectors, as described below. Since the detector background is independent of the size of the shutter aperture, while the astrophysical signal is not, the background can be determined by comparing the data observed with both 100% and 10% shutter apertures toward the same positions on the sky. The team thus determined the detector background assuming that the intensities measured with 100% and 10% shutter apertures are:
I100% = Isky + Idet
I10% = a Isky + Idet
Here, a, Isky, and Idet denote the scale factor between two shutter apertures, the sky intensity measured with 100% shutter aperture, and the detector background, respectively. Figure 1 shows the sky areas for the L-band where observations were made in both the 100% and 10% shutter modes. The same is shown for the S-band in Figure 2.
Figure 1 - The L-band count rate rate maps used to calculate the detector background of the L-band. [Left] The 100%-observation map and [Right] 10%-observation map of the sky region where both 100% and 10% modes were used to observe.
Figure 2 - The S-band count rate maps used to calculate the detector background of the S-band. [Left] The 100%- observation map and [Right] 10%-observation map of the sky region where both 100% and 10% modes were used to observe.
In Seon et al. (2010), the detector background was assumed to be constant across the detector plane. However, in the most recent analysis, the team found weak dependencies on time (orbit) and wavelength. Figure 3 shows the dependence of the scale-factor a on wavelength. Figure 4 shows the detector background spectra, represented in two different units, derived using the scale factor. In the present version of FIMS data products, the dependence of the detector background on time was ignored. The figure shows that the scale factor differs from the originally designed value of 0.1 and varies according to wavelength. The variation in the scale factor might be due to differences in photon paths or the dependence of grating reflectance on the angle (the wavelength direction). However, at the moment, it is unclear why the scale factor is not a constant across the detector plane. The team has not examined the possibility of detector background variation along the spatial direction. The detector background for L-band shown in Figure 4 is more or less consistent with that estimated in Seon et al. (2011), but slightly higher at the longest wavelengths.
Figure 3 - The scale factor between the 100% and 10% shutter modes for (a) L-band and (b) S-band.
Figure 4 - Detector background spectra for (left) L-band and (right) S-band. The upper and lower panels are shown in units of ct s−1 Å−1 and continuum units (CU = ct s−1 Å−1 cm−2 sr−1), respectively.
SPEAR Procedure
The SPEAR team followed a similar but less detailed procedure, which is not documented in detail, using smoothed fits to the background. For emission line maps, the weighted sum of dust scattered stellar continua that was subtracted was in all cases much larger than the detector backgrounds.