Overview of the study

Details for radiative transfer intercomparison

/home/users/emde/model_intercomparisons/SWIR_RT_intercomparison/

• All calculations performed using libRadtran (www.libradtran.org)
• Contribution to accurate RT cases, scalar (DISORT and MYSTIC), vector (MYSTIC)
• for DISORT 32 streams and second order intensity correction are used
• CPU times

Comments

• For MYSTIC with polarization the noise for Q in clear-sky atmosphere is large, because photons reflected at the surface only contribute to I, not to Q –> include some importance sampling method to enhance Rayleigh scattering?
• Optical properties (aerosol_files can be used for scalar calculations) can not be specified in the required way (spectral ssa, spectral altitude dependent tau, constant phase matrix)
  • best solution now: optprop_file including spectral ssa, const phamat, ext=1, use wc.dat to specify tau (but this could only be done sequentially for all wavelengths, much overhead).
  • may be better: make a aerosol_files option for phase matrix (but this requires too much memory).

• Band 1: O2 A-band, Window 1
• Band 2: O2 A-band, Window 2
• Band 2: 1.61 micron CO2 band, Window 1
• Band 3: 2.04 micron CO2 band, Window 1
• Band 3: 2.04 micron CO2 band, Window 2

First results: libradtran.tar.gz

1. DISORT:
   • Cases A-E, scalar, no sun glint
2. MYSTIC:
   • Cases A,B,E, scalar, no sun glint
   • Cases A,B,E, vector, including sun glint

• MYSTIC scalar for sun glint to be done
• DISORT for sun glint to be done
• MYSTIC ice clouds with more photons