Temporal variations in atmosphere

[teutoburg]

During the recent ScopeSim workshop in Vienna, the topic of varying sky background / atmospheric emission was brought up. Basically, for a sequence of observations, the sky shouldn't always look exactly the same (unless fixed on purpose for testing) and running a "sky subtraction" via subtracting an empty sky observation from a source observation shouldn't be "perfect". Somewhat similar in concept to the random seed in some detector noise effects. This shouldn't be too hard to implement, e.g. by varying some parameters of the skycalc atmosphere within a given range.

[hugobuddel]

The hard part would be to properly seed the randomness such that it is still reproducible. But maybe one problem at a time.