Fringe Maps

One of the unfortunate aspects of a thinned CCD is the problem of interference fringes. When light of certain wavelengths hit the CCD, some of the light is reflected and interferes with itself. The scale of the fringe pattern is wavelength dependent, and like the rainbow on an oilslick, you get a fringe pattern like that in Figure 7.3.

The night sky is the greatest source of this fringe pattern. Because most of the night sky brightness is contained in certain emission lines, some of these lines create a strong fringe pattern. Light from the sun, moon, and stars are broadband, and so will not show the fringe pattern. However, since we are running an unfiltered system, the fringing subtly effects the photometry of the stars in unpredictable ways.

One cannot leave the fringe pattern in the sky flat, as in Figure 7.2, as this would amplify the effect of the fringing on the photometry. At the moment, we have decided to find the fringe scaling factor and subtract the fringe pattern. This helps, but is not perfect. One solution (to many problems!) would be running with an R-band filter, but this would greatly decrease our sensitivity to faint sources.

Figure 7.3: A sample fringe map. The image has been arbitrarily normalized. This pattern is consistent from image to image, although the normalization changes.