One last question - why include the MK expected values in a comment rather than in the data array? Not only does it break the paradigm started by the Flatau data, but it would be nice to print out the difference between the values we're getting from MK calculations and what's expected from their table...

I will add these to the arrays. I noticed that no other field from the "data" arrays (except for the first two) were used in the script. Therefore, I decided not to put these values in the array (as they are not used) but specify them in comments so that we know what should we expect without referring to the paper.

Also, I noticed that, the dt calculation in polysvp looks like the following:
dt = T-T0
I modified it to:
dt = max(T-T0, -80.0)
so that it is exactly the same as in the code. Should I push this change as well?

About including expected MK values in a comment rather than including them in the data array like Flatau: I made my comment after seeing those later entries used in the print statements... but closer inspection now shows that those later-entry print statements are actually commented out. So I don't care if you fix this or not. Just including the 2 entries that are actually used is actually kind of nice.

About dt = max(T-T0, -80.0): I don't think it matters since we never call the python with T<-80C(?). I guess we might as well make the change for consistency. I didn't realize we were thresholding T to be warmer than -80C in the model. I guess that should prevent the negative T errors we've been seeing in polysvp, huh?

Sorry, for some reason, I didn't read your first comment earlier and went ahead to fix the code. The code now has complete data arrays with None inserted for the missing data. I have also added max for -80C.

MK tables starts from 150K (-123.15 C), so I had to add max(T-T0,-80) to get values for Flatau (similar to the values produced by the C++ code).