👍 to this in principal.

This is a pretty big job to do right, though, because it's highly wavelength-dependent. For example at CTIO, the sky varies by ~2 mag from the blue to red sides of just the optical, but on the blue end it can easily be 5 mags from new to full moon.

More broadly, it might be useful to make this a part of a full sky brightness calculator that includes both moon and site-specific sky brightness (which would need to be provided/customizable by the user).

I hear that @yoachim has suggestions for how to do this, since he has implemented similar functionality for LSST. Any comments @yoachim?

Did someone say sky brightness model?
Here's the model for LSST: https://github.com/lsst/sims_skybrightness
hHere's a notebook showing it in action: https://github.com/lsst/sims_skybrightness/blob/master/doc/Sky_Model_demo.ipynb

It's mostly interpolating a set of spectra from the ESO calculator, plus an analytic model I made for twilight.

I guess no-one wants to duplicate this in astroplan, right?

In that case, how about adding a section "related packages" to the astroplan docs and pointing people to the LSST code?

What we really need is for ESO to open source the code behind this web page:
http://www.eso.org/observing/etc/bin/gen/form?INS.MODE=swspectr+INS.NAME=SKYCALC

I'm pretty sure the code was written in a way that it could be modified to other locations (and elevations), but they haven't exposed that on the web API.

It wouldn't be too painful to just use the ESO calculator for a few select pointings and fit some simple functions to the sky brightness for a set of filters.