RFM for Planetary Atmospheres | ![]() | ![]() |
30JAN25 |
Symbol | Value | Units | Description | Application |
---|---|---|---|---|
rE | 6367.421 | km | Radius of curvature | Circular geometry |
g | 9.80665 | ms-2 | Gravitational acceleration | Plane-parallel geometry |
Mair | 28.964 | kg kmol-1 | Molar mass of air | Plane-parallel geometry |
Cp | 29012.0 | J K-1 kmol-1 | Molar heat capacity of air | Cooling rate |
Tsp | 2.7 | K | Cosmic background temperature | Microwave calculations |
The atmospheric profile within the RFM is entirely user-specified so the main adaptation for non-terrestrial atmospheres is straightforward.
However, depending on the type of calculation, there are other parameters for which the RFM assumes terrestrial values by default, listed in the table on the right.
These can all be altered within the *PHY section of the RFM driver file.
The default values themselves are set in module phyadj_dat.f90
For plane-parallel atmospheres, the default behaviour of the RFM is to assume hydrostatic equilibrium to derive absorber amounts in each layer from pressure and mixing ratio supplied in the *ATM section, independent of altitude coordinate (any user-supplied *HGT profile being ignored). The conversion requires both gravitational acceleration and the molar mass of air so, for other planets, these two values should be altered. However, using the HYD Flag, it is also possible bypass the hydrostatic assumption and use the user-supplied *HGT profile instead, in which case the absorber amount can be calculated independently of these parameters (presumably you will have used appropriate values of g and Mair to match your height and pressure profiles beforehand).
Cooling rate calculations (COO Flag) are, effectively, a conversion from radiation loss to temperature change for an atmospheric layer, and therefore require some value of the molar heat capacity of air.
The cosmic background temperature is used to calculate the Planck function representing cold space. By its very definition it is not specific to the Earth, but is included here since it can be adjusted in the same way as the other parameters.
The function refrac_fnc.f90 returns the refractivity (=refractive index - 1) using a modified form of Edlen's formula printed in Kaye & Laby, which is a function of (p/T) (i.e., density) and wavelength. In reality, at microwave frequencies, the refractivity also depends on water vapour, but this is not included in the refractivity model. The wavelength dependence is weak, refractivity increasing by only a few percent between long wavelengths and the visible region of the spectrum, most of the variation being at short wavelengths.
Rayleigh (i.e. molecular) scattering, which again only becomes significant at visible wavelengths (varies as 1/λ4), is handled in subroutine spcrex_sub.f90, and based on Bodhaine et al (1999). Apart from reproducing the Kaye & Laby refractivity formula, this also includes the King Factor (or depolarisation) calculated in subroutine rexpth_sub.f90, which is based on atmospheric composition. This subroutine includes default concentrations for four components (N2, O2, CO2 and Ar) which may be replaces by any user-supplied atmospheric profiles. However, the influence of the King Factor is generally close to 1.
RFM v5.22 allows line parameters to be supplied in an alternative format which allows 'self' as well as 'air' terms for additional line parameters. This may be useful for atmospheres where the main absorber is also the main constitutuent (eg CO2 on Mars).
There is a HITEMP version of HITRAN, for transitions which become significant at higher temperatures, and the GEISA database contains some extra molecules. See RFM Spectroscopic Data for use of these alternative datasets
The HITRAN Collision Induced Absorption data has recently been extended to allow for a number of different broadening gases. RFM v5.10 onwards allows for a completely general specification of CIA molecules.