Add a fringing surface-layer operator

[kadrlica]

As one of the inputs to #827 we need to add a way to calculate and applying fringing.

Some documents for reference come from David Kirkby:

• talk
• notebook

[rmjarvis]

Another worthwhile link for people with access to LSST DESC confluence pages:

HyeYun's animation of fringes as a function of wavelength (using TS3 data for ITL-3800C-107 from BNL)

[rmjarvis]

Also this:

Andy's movie of (PRNU variations and) fringes as a function of wavelength for e2v-CCD250-088

[kadrlica]

Some investigation of Gaussian random fields. The basic idea is that eventually we will want to adjust the height field of the CCD surface layers to adjust the physical scale of the fringing.

The fringing obviously depends on wavelength, but we are going to need to decide how much the spatial pattern and amplitude of fringing can be separated from the other surface-layer operations (conversion depth and incident angle). In reality they are all bundled up, but we may need to see how far we can get assuming they are separable.

[rmjarvis]

Copying the final image from Alex's notebook here, so the link stays live: