Pre-Defined Background Subtraction

  While the pre-launch version of Calnic C used a pre-computed model of the background in space, it has since been replaced by empirical estimates for each grism. These new estimates were computed by medianing several sparse images together. Calnic C and NICMOSlook scale these pre-defined backgrounds using a scaling factor that is calculated by taking the mean flux values of a region surrounding each spectrum, and dividing it by the mean of the background image in the same region. The scaled background estimate is then subtracted from the grism image prior to extracting each spectrum.

The region used to scale the background is the same one that is used by the Interpolated Background Subtraction method described in section 3.4.2. Figure 3.2 shows an example of how Calnic C chose the region in a complex situation.

The uncertainty in the background estimate is computed from the error plane of the grism image. The uncertainty is the square root of the sum of the errors divided by the square root of the number of pixels.