next up previous
Next: Bibliography

Set the radiative transfer equation solver to be used.
rte_solver type
It not specified the default rte_solver is disort2. Choices for type are
If not specified the default rte_solver is cdisort. Choices for type are
cdisort

C-version of the disort algorithm, translated from Fortran by Tim Dowling. This is the recommended discrete ordinate code in libRadtran. For documentation see src_f/DISORT2.doc as well as the papers and the DISORT report at ftp://climate1.gsfc.nasa.gov/wiscombe/Multiple_Scatt/. The intensity correction can be performed according to Nakajima and Tanaka (1988) using disort_icm moments (like in the original code), or with the improvements described in (Buras, Dowling, Emde, in preparation; default). Can be run in plane-parallel geometry (default) or in pseudo-spherical geometry (using cdisort_pseudospherical).
ctwostr

C-version of the two-stream radiative transfer solver described by Kylling et al. (1995). Can be run in plane-parallel geometry (default) or in pseudo-spherical geometry (using ctwostr_pseudospherical ).
disort

Same as cdisort which is the recommended discrete ordinate code in libRadtran. Before libRadtran 1.6 cdisort invoked the Fortran code disort 1.3.
disort2

Same as cdisort which is the recommended discrete ordinate code in libRadtran. Before libRadtran 1.6 cdisort invoked the Fortran code disort 2.0.

fdisort1

The standard plane-parallel disort algorithm by Stamnes et al. (1988), version 1.3 - provided for compatibility reasons. Use only if you have troubles with the default disort or for historical reasons. For documentation see src_f/DISORT.doc as well as the papers and the DISORT report at ftp://climate1.gsfc.nasa.gov/wiscombe/Multiple_Scatt/. To optimize for computational time and memory, please adjust the parameters in src_f/DISORT.MXD for your application and re-compile. For your application please use rte_solver fdisort2 which is the advanced version, unless you e.g. want to explore how a specific feature of fdisort2 (e.g. the Nakajima and Tanaka (1988) intensity correction) improves the fdisort1 result.
fdisort2

Version 2 of the Fortran algorithm disort - provided for compatibility reasons. Use only if you have troubles with the default disort or for historical reasons. For documentation see src_f/DISORT2.doc as well as the papers and the DISORT report at ftp://climate1.gsfc.nasa.gov/wiscombe/Multiple_Scatt/ fdisort2 has several improvements compared to its 'ancestor' fdisort1 (version 1.3). To optimize for computational time and memory, please adjust the parameters in src_f/DISORT.MXD for your application and re-compile.

Note! fdisort2 is a new version of the original disort code which was implemented in summer 2009. It uses phase functions to calculate the intensity corrections by Nakajima and Tanaka (1988) instead of Legendre moments. Hence it needs cloud properties files which contain the phase functions. It is still possible to use the old version of disort2, you need to specify disort_icm moments.

sdisort

Pseudospherical disort as described by Dahlback and Stamnes (1991). Double precision version. To optimize for computational time and memory, please adjust the parameters in src_f/DISORT.MXD for your application and re-compile.
spsdisort

Pseudospherical disort as described by Dahlback and Stamnes (1991), single precision version. Warning: it is not recommended to use spsdisort for really large solar zenith angles nor for cloudy conditions. For large optical thickness it is numerically unstable and may produce wrong results. To optimize for computational time and memory, please adjust the parameters in src_f/DISORT.MXD for your application and re-compile.
polradtran

The plane-parallel radiative transfer solver of Evans and Stephens (1991). Includes polarization. The full implementation of the polRadtran solver in uvspec is quite new (version 1.4). If you find unusual behaviour, please contact the libRadtran authors.
twostr

The two-stream radiative transfer solver described by Kylling et al. (1995), in pseudo-spherical geometry.
twostrpp

The two-stream radiative transfer solver described by Kylling et al. (1995), in plane-parallel geometry.
sos

A scalar pseudospherical succesive orders of scattering code. Works for solar zenith angles smaller than 90 degrees. Can calculate azimuthally averaged radiances. Set nscat to specify the order of scattering.
tzs

TZS stands for "thermal, zero scattering" and is a very fast analytical solution for the special case of thermal emission in a non-scattering atmosphere. Please note that TZS does only radiance calculations at top of the atmosphere.
sss

SSS stands for "solar, single scattering" and is an analytical single scattering approximation which might be reasonable for an optically thin atmosphere. Please note that SSS does only radiance calculations at top of the atmosphere. This is an experimental solver - be careful!
null

The NULL solver does not solve the radiative transfer equation. However, it sets up the optical properties, and does the post-processing; useful if you are either interested in the overhead time required by a particular model input or if you are simply interested in the optical properties, as output by verbose.
Default: cdisort




next up previous
Next: Bibliography
Claudia Emde 2011-03-03