http://www.strw.leidenuniv.nl/~moldata/
is where all the molecular data is from. 

To plot the data used use this fragment:
window, 3, xsize=1200, ysize=700 
 plot, moldata[0,*], moldata[2,*], /xlog, /ylog, thick=2, xtitle='H2
 number density', ytitle='emission per molecule', title='solid: C^18O(2-1), dashed: N_2H^+(1-0), and dashed-dotted: NH_3(1,1)', charsize=2
 oplot, moldata[0,*], moldata[1,*], linestyle=2, thick=2  
 oplot, moldata[0,*], moldata[3,*], linestyle=3, thick=2



Email from Mark:

I have attached both the data and the code. The data is in the files
moltab.txt and kappatab.txt. The file moltab.txt contains 4
columns. The first column is the density, and the next 3 columns are
the frequency-integrated emissivity per molecule for the the
C^18O(2-1), N_2H^+(1-0), and NH_3(1,1) lines for optically thin gas at
a temperature of 10 K -- by per molecule I mean per molecule of the
emitting species, so you need to multiply by your favorite
abundance. The file kappatab.txt gives the opacities. It is in the
same format, but gives the frequency-integrated opacity per molecule
for those 3 lines.

For the code, I have attached the IDL routines emissivity.pro and
absorptivity.pro, which generated these tables. They both call the
subsidiary routine getlevpop.pro. I have also included the routine
maketables.pro, which was used to generate moltab.txt and
kappatab.txt. The routines emissivity and absorptivity return the
frequency-integrated absorptivity and emissivity per molecule. They
each take 5 arguments:

j = emissivity(n, t, upper, lower, molfile),

where n = H_2 density in cm^-3, t = temperature in K, upper = level of
upper state, lower = level of lower state, and molfile = name of
molecular data file. The molecular data file is a text file in the
format provided by the Leiden Atomic and Molecular Database. I have
not included these files in the e-mail. You can download them from the
Leiden site. Levels are ordered by energy, starting with 0 = ground
state, 1 = lowest energy excited state, etc.

Caveat: these routines do not treat hyperfine structure. For lines
that have hyperfine structure the emissivity and absorptivity returned
are its value integrated over all hyperfine components. If you want
you can handle this by using a more complex line shape function that
takes into account the hyperfine structure.

Let me know if you have any questions about these files.

--
-Mark