The Rousselot (2000) night sky list is now available here: https://people.ast.cam.ac.uk/~optics/dazle/sky-data/Rousellot_list_v2.0.dat . To be downloaded and placed under git.

For Patat (2008), the night sky spectra (as opposed to line list), are available here: http://www.eso.org/~fpatat/science/skybright . The list of OH lines does not seem to be available in machine-readable form. Investigating further.

I don't think we need machine-readable Patat. Almost all the lines are in the Rousselot list, and Patat will help us identify the other lines (there are other references).

Let's start with these two sources and get the line lists into the system.

[ I thought we already had that Rousselot file somewhere, but really wanted one annotated with the transitions? And that someone had that list available. ]

I took a quick look at your new version – looks much better.  Thanks for fixing the OI wavelengths.

Please change the 557.8887 intensity to say 100

The OH lines are visible above a quoted intensity of c. 0.40.  There are a lot of doublets with the identical intensity and splittings of c. 0.003nm, e.g.

 724.68910    2.657e+00   OH    0 # [5]
 724.69160    2.657e+00   OH    0 # [5]

Please merge these into a single line with the average wavelength (and do NOT label it as a blend – add a new flag if we need one). They must be a transition to a split lower state, so the intensity ratio will be stable and this is OK. No need to do the ones with intensity < 0.10

There are other obvious blends, but let's get these issues cleaned up first.

I've done the work for this. Stand by for a push once I've gone through and checked for other problems.

Effectively reviewed by RHL and Price, so elect to skip a formal PR in this instance.

Who is going to merge this?

I am. One sec.