Comparison with RADEX

RADEX (van der Tak et al., 2007) is a popular algorithm that solves radiative transfer for given column density, H₂ number density, and gas temperature. Comparison of RADEX with 3D-PDR and RT-tool is only valid for:

Isothermal runs (thermal balance switched off)
Uniform-density 1D clouds
No external radiation field

3D-PDR Setup for Comparison:

Edit config.mk: set GUESS_TEMP=0 and THERMALBALANCE=0 for 1D runs

Modify params.dat as appropriate. For instance:

2)  1Dn43.dat                  !ICs file -- assuming you have already created it
...
4)  radexcomp                  !Output prefix
5)  0                          !G0 (no FUV)
6)  1.0E-17                    !Cosmic-ray rate
...
9)  1.0                        !Turbulent velocity
...
23) 30.0                       !Gas temperature
...
33) hco+.dat                   !HCO+ coolant -- e.g. if you want to include additional coolants

Run ./3DPDR

RT-tool Setup for Comparison:

Edit paramsRT.dat accordingly
Run ./RTtool

RADEX Setup for Comparison:

Set the linewidth value to that of turbulent velocity used in 3D-PDR and RT-tool
Set the RADEX “escape probability” to LVG for consistency.
Use as H2 number density and column density of the coolant, the values outputted by RT-tool.

Note

Differences are expected since 3D-PDR treats the cloud as 1D with depth-dependent level populations, while RADEX uses a 0D approach.

Example

The plot below shows a comparison between 3D-PDR & RT-tool (black lines) against RADEX (red lines). The comparison is done for three different uniform density models with \(10^2\), \(10^3\), and \(10^4\) total H-nucleus number densities at a fixed gas temperature of \(T_{\rm gas}=20\,{\rm K}\). The radiation temperatures (Tr in K) of C+, [CI] (1-0), CO (1-0), [OI] 63μm, HCO+ (1-0) and HNC (1-0) versus the column density (N) of the corresponding coolant are shown. In general there is a broad agreement between the two codes. As noted above, the differences seen in e.g. HCO+ and HNC for low densities are potentially associated with differences in the treatment of level populations.