Improve Rayleigh Correction for thin cirrus to avoid brown clouds.

Feature Request

Is your feature request related to a problem? Please describe.

The true colour RGB (here shown for FCI, 2025-02-26 12:00 in the Norwegian Sea north of the Shetland Islands) contains a correction for Rayleigh Scattering (using pyspectral). It uses the red channel to avoid overcorrecting clouds, that have less atmosphere above them than the surface. This correction does not work very well for thin cirrus, which can lead to brown clouds at high satellite and solar angles.

This might risk users confusing such clouds for containing dust.

Describe the solution you'd like

The default true colour RGB should use an improved correction to avoid clouds turning brown.

Describe any changes to existing user workflow

It will change the looks of the true colour RGB and other RGBs that use it.

Additional context

Additional discussion on Slack.

For reference, this may be improved but not totally fixed by my (slowly progressing) work on new LUTs using libradtran in mystic mode to allow for a fully non-spherical atmosphere. The original LUTs were made using a pseudospherical calculation and hence do not extend to very high solar or viewing angles, such as above.

However, the primary problem here is that the cloud is quite dark despite being high and therefore gets "over-corrected" by pyspectral.