UV Photo-rates¶

There are several parameters to control how ProDiMo calculates the photo-rates

.true.      ! Rphoto_from_RT    : calculate photo-rates from rad. transfer?
.false.     ! Rphoto_bandint    : use UV-optical colours from I_\nu^star?
20          ! NLAM              : total number of lambda-bins
8           ! NUV               : number of lambda-bins in the UV 91.2 nm - 205 nm
2           ! N1UV              : number of lambda-bins used for the UV band 1 (91.2 nm - 111 nm) 

where NLAM is the total number of bands used in the continuum radiative transfer (see Continuum Radiative Transfer), NUV is the number of bands out of the NLAM bands, that should be used in the UV range (91.2 nm - 205 nm) and N1UV is the number of bands out of the NUV bands that should be used for UV band 1 (91.2 nm - 111 nm).

The first versions of ProDiMo had Rphoto_from_RT=.false., in which case the pre-integrated UMIST photo-rates were used and scaled in a sophisticated way to become consistent with the local strength of the UV field chi (91.2 nm - 205 nm) as resulting from the radiative transfer.

However, this assumes that the "color" of the stellar UV is similar to the color of the standard Draine UV field, which can be wrong by orders of magnitude. We recommend using Rphoto_from_RT=.true., in which case detailed photo-cross-sections from the Leiden database are used, which are then integrated "on the fly" over the local mean intensity $J_\nu(r,z)$ as result from the radiative transfer.

Interpolation method for the radiation field¶

In case Rphoto_from_RT=.true. the radiation field $J_\nu(r,z)$ is taken from the few band-integrated lambda-points of the radiative transfer, the number of which is tunable with NUV. However, those values have to be interpolated to properly integrate over the provided photo cross-sections.

Spline-interpolation (Default)¶

If Rphoto_bandint=.false. (the default), ProDiMo will make a spline-interpolation in log(frequency) to get $J_\nu$ at every frequency $\nu$ needed during the integration of the photo-rates. This may be the best option if the input UV (and dust opacities) are smooth. However, sometimes the spline interpolation produces overshoots. If you use a highly variable stellar input UV, this is likely to occur. Examples are in StandardObjects/EarlySun (very strong Ly alpha) and for A-stars which have an enormous drop from soft to hard UV.

The spline interpolation can be verified with the IDL routine idl/RTinterpolation.pro or the with prodimopy.plot.Plot.plot_starspec routine. In both tools required the output file RTinterpolation.out which is created if verbose_level>=0.

Both methods compare the band-integrated $J_k(r,z)$ to the spline-interpolated $J_\nu(r,z)$. If serious overshoots occur, one can try to increase NLAM and/or NUV/N1UV to get a better sampling of the UV spectrum.

Band-integrated method¶

Another option is to use Rphoto_bandint=.true.. In this case, ProDiMo uses a different strategy to get $J_\nu(r,z)$ from the few band-integrated lambda-points:

$$J_\nu(r,z) = Istar_\nu * J_k(r,z)/Istar_k$$

It will use $Istar_\nu$ (the stellar radiation field) as a "pattern" of the UV field, and shift it up and down to match the band-integrated radiative transfer results. $k$ is the index of the spectral band that $\nu$ belongs to. $Istar_k$ is the band-integrated $Istar_\nu$, and $J_k(r,z)$ is the band-integrated radiative transfer result.

This strategy is actually used everywhere in the program, for example also for providing the background radiation field for the non-LTE modelling of atoms and molecules. In the near IR, where thermal emission from dust starts to dominate, the upper method makes no sense and is switched off for lambda $<2\,\mu m$, falling back to the spline-interpolation method.

My recommendation is to use Rphoto_bandint=.true. whenever you see problems with the Spline-Interpolation method. There is a second drawback with this method though, because, at large radii, the UV field (in particular around 100nm) may be dominated by the interstellar component, in which case Rphoto_bandint=.false. is obviously the better option. ProDiMo will state at which radius this is expected to occur, and will warn you if that radius < Rout.

Choice of cross-section data¶

There are different versions of the photo cross-section data included in ProDiMo. One choose by setting the Parameter

2           ! photorates_version   : use cross-section data from Heays+ 2017

Currently one has options 1 and 2.

• Version 1: Uses the data from data/ChemicalNetwork/pd . These cross-section data are from van Dishoeck+ 2006 and references therein.
• Version: Uses the data from data/ChemicalNetwork/pd2017_version2. The reference for this data is Heays+ 2017, which also includes a detailed description.

The data files are from the Leiden database.

In both directories, there is a file pp_list.txt which makes the connections from the cross-section data to the actual photo-reaction. This file can also be edited if necessary (see Adding photo cross-sections). The data directories also contain README files that provide further information.

Additions to the Heays+ 2017 data¶

Here we list cross-section data that does not come from Heays+ 2017, but have been added later on. Please cite the respective references if this data is relevant for your work.

• UV photodissociation spectrum for CH3+ (giving CH2+ and H) has been added. Cite: del Mazo-Sevillano P. et al. (2024)

Treatment of the EUV wavelength range¶

The data from Heays+ 2017 also includes the EUV range ($\lambda < 91.2\,\mathrm{nm}$), for some photo-reactions. By default, this wavelength range is not included in ProDiMo. However, this can be activated by setting

.false.     ! noEUV    : include EUV cross-sections for the photo-reactions

We want to emphasize that the EUV is not included in the radiative transfer and an extrapolation scheme for the local radiation field is used. Furthermore the impact of the EUV cross-sections also depends on the chemical network used. For example UMIST2013 still included some photo-reactions that had only cross-sections in the EUV, however, this was maybe not intentional and it seems those reactions have been removed in UMIST2022. So in general we recommend for the moment to always use

.true.     ! noEUV    : Recommended value, don't include EUV cross-section for the photo-rates

However, one can still have stellar EUV radiation included in ProDiMo by using the ProDiMo X-ray chemistry and/or X-ray radiative transfer.