Example included in:
This directory includes an example, how to perform a line-by-line calculation with DISORT using absorption coefficients generated by the ARTS model (http://www.sat.ltu.se/arts/).
ARTS is installed in the directory
the executable is accessable as
In order to have ARTS default settings please set the following environment variable in your shell:
The file gen_moltau.arts provides an ARTS control file example which can be run as follows:
With this setup ARTS generates line-by-line absorption coefficients in the wavelength range from 10.00 to 10.01 microns using 1000 grid points (equally spaced in frequencies). Please see also the comments in the file, ARTS writes a molecular_tau_file called moltau_example.nc in netcdf format.
Try arts -d … to get help for the individual ARTS functions. The ARTS user guide is available at
uvspec_lbl.ir.disort.inp is an example input file for uvspec, which uses the molecular_tau_file from ARTS. Running this file should produce uvspec_lbl.ir.disort.out.
Note that the altitudes, pressures, and temperatures in the atmosphere_file must correspond to the atmosphere that was used in ARTS. ARTS also includes the FASCOD standard atmospheres (as libRadtran) but only up to an altitude of 95km. For this reason, the file afglms.dat is included here, which is the standard midlatitude-summer atmosphere from the libRadtran/data/atmmod directory cut above 95km. (If you want to use an atmosphere different from the standard atmospheres please contact me, I have a script to easily generate consistent atmosphere files.)
— Claudia 2010/07/28 14:46
Here a comparison for TOA nadir radiances is shown for DISORT, tzs (thermal zero scattering) and ARTS. Differences are mostly below 0.1% which is ok. The calculation was made for the standard midlatitude-summer atmosphere.
Here a comparison for BOA nadir radiances is shown for DISORT, and ARTS. For the standard midlatitude-summer atmosphere the differences are as much as 5%!! Doing the same calculation with 1000 layers the models approach each other in the middle which shows that the difference is due to different vertical interpolation schemes. In ARTS T and babs are linearly interpolated, and the Planck emission is calculated from the interpolated values. In libRadtran an average absorption coefficient is taken for each layer and the Planck function is linearly interpolated.