Contents

1. Introduction
2. Releases
3. Column Version
4. GCM Version
5. Contact
6. References

This package contains the source code and sample makefiles necessary to run the latest version of RRTMG_SW, a correlated k-distribution shortwave radiative transfer model developed at AER for application to GCMs. This version of RRTMG_SW utilizes a two-stream radiative transfer method as implemented at ECMWF. This code has also been modified to utilize updated FORTRAN coding features. Two modes of operation are possible:

1. RRTMG_SW can be run as a column model using the input files and source modules described in the column version section, or
2. it can be implemented as a subroutine into an atmospheric general circulation model or single column model.

The version of RRTMG_SW provided here utilizes a reduced complement of 112 g-points, which is half of the 224 g-points used in the standard RRTM_SW, and a two-stream method for radiative transfer. Additional minor changes have been made to enhance computational performance. Total fluxes are accurate to within 1-2 W m^-2 relative to the standard RRTM_SW (using DISORT) in clear sky and in the presence of aerosols and within 6 W m^-2 in overcast sky. RRTM_SW with DISORT is itself accurate to within 2 W m^-2 of the data-validated multiple scattering model, CHARTS. Required absorption coefficient input data can be read in either from data stored within the code or from a netCDF file as selected in the makefile.

This model can also utilize McICA, the Monte-Carlo Independent Column Approximation, to represent sub-grid scale cloud variability such as cloud fraction and cloud overlap. If the McICA option is selected to model a cloudy profile in column mode, then the model will run stochastically, and the output fluxes and heating rates will be an average over 200 samples. In GCM mode, the code will calcualte a single column per profile, and the statistical basis is provided by the spatial and temporal dimensions of the 3-D calculations. Several cloud overlap methods are available for partial cloudiness including maximum-random, exponential, and exponential-random. Without McICA, RRTMG_SW is limited to clear sky or overcast cloud conditions.

For more information on the model, see the Wiki Description Page.

Version 5.0 is the latest version of the model

Releases before Version 5.0 are not publicly available.

The following text files (in the doc directory), along with this README provide information on release updates and on using and running RRTMG_SW:

The following source files (in the src directory) must be used to run RRTMG_SW in stand-alone mode as a column model (the utility files are stored separately in the aer_rt_utils directory):

The following module files (in the modules directory) must be used to run RRTMG_SW in stand-alone mode as a column model (these must be compiled before the source code files):

The following file (in the data directory) is the optional netCDF input file containing absorption coefficient and other input data for the model. The file is used if keyword KGSRC is set for netCDF input in the makefile.

The following files (in build/makefiles directory) can be used to compile RRTMG_SW in stand-alone mode as a column model on various platforms. Link one of these into the build directory to compile.

Several sample input (and output) files are included in the run_examples_std_atm directory. Note that user-defined profiles may be used for as many as 200 layers.

1. In the build directory, link one of the makefiles from the makefile sub-directory into build/make.build. To use the optional netCDF input file, switch the keyword KGSRC in the makefile from dat to nc. Compile the model with make -f make.build
2. Link the executable to, for example, rrtmg_sw in the run_examples_std_atm directory
3. If the optional netCDF input file was selected when compiling, link the file data/rrtmg_sw.nc into the run_examples_std_atm directory.
4. In the run_examples_std_atm directory, run the UNIX script ./script.run_std_atm to run the full suite of example cases. To run a single case, modify INPUT_RRTM following the instructions in doc/rrtmg_sw_instructions.txt, or copy one of the example input_rrtm* files into INPUT_RRTM. If clouds are selected (ICLD > 0), then modify IN_CLD_RRTM or copy one of the in_cld_rrtm* files into IN_CLD_RRTM. If aerosols are selected (IAER > 0), then modify IN_AER_RRTM or set it to the sample file in_aer_rrtm-aer12.
5. In column mode, if McICA is selected (IMCA=1) with partial cloudiness defined, then RRTMG_SW will run the case 200 times to derive adequate statistics, and the average of the 200 samples will be written to the output file, OUTPUT_RRTM.

The following source files (in the src directory) must be used to run RRTMG_SW as a callable subroutine:

NOTE: Only one of rrtmg_sw_k_g.f90 or rrtmg_sw_read_nc.f90 is required.

The following module files (in the modules directory) must be used to run RRTMG_SW as a callable subroutine (these must be compiled before the source code)

The following file (in the data directory) is the optional netCDF file containing absorption coefficient and other input data for the model. The file is used if source file rrtmg_sw_read_nc.f90 is used in place of rrtmg_sw_k_g.f90 (only one or the other is required).

1. The module rrtmg_sw_init.f90 is the initialization routine that has to be called only once. The call to this subroutine should be moved to the initialization section of the host model if RRTMG_SW is called by a GCM or SCM.
2. The number of model layers and the number of columns to be looped over should be passed into RRTMG_SW through the subroutine call along with the other model profile arrays.
3. To utilize McICA, the sub-column generator (mcica_subcol_gen_sw.f90) must be implemented in the GCM so that it is called just before RRTMG_SW. The cloud overlap method is selected using the input flag, icld. If either exponential (ICLD=4) or exponential-random (ICLD=5) cloud overlap is selected, then the subroutine get_alpha must be called prior to calling mcica_subcol_sw to define the vertical correlation parameter, alpha, needed for those overlap methods. Also for those methods, use the input flag idcor to select the use of either a constant or latitude-varying decorrelation length. If McICA is utilized, this will run only a single statistical sample per model grid box. There are two options for the random number generator used with McICA, which is selected with the variable irnd in mcica_subcol_gen_sw.f90. When using McICA, then the main module is rrtmg_sw_rad.f90. If McICA is not used, then the main module is rrtmg_sw_rad.nomcica.f90, though the cloud specification is limited to overcast clouds.

Atmospheric and Environmental Research 131 Hartwell Avenue, Lexington, MA 02421

Original version: Eli. J. Mlawer, J. S. Delamere, et al. (AER) Revision for GCMs: Michael J. Iacono (AER)

Contact: Michael J. Iacono (E-mail: [email protected])

• AER Radiative Transfer Models Documentation
• Github Wiki
• RRTMG_SW, RRTM_SW
  • Iacono, M.J., J.S. Delamere, E.J. Mlawer, M.W. Shephard, S.A. Clough, and W.D. Collins, Radiative forcing by long-lived greenhouse gases: Calculations with the AER radiative transfer models, J. Geophys. Res., 113, D13103, doi:10.1029/2008JD009944, 2008.
  • Clough, S.A., M.W. Shephard, E.J. Mlawer, J.S. Delamere, M.J. Iacono, K. Cady-Pereira, S. Boukabara, and P.D. Brown, Atmospheric radiative transfer modeling: a summary of the AER codes, J. Quant. Spectrosc. Radiat. Transfer, 91, 233-244, 2005.
• McICA
  • Pincus, R., H. W. Barker, and J.-J. Morcrette, A fast, flexible, approximation technique for computing radiative transfer in inhomogeneous cloud fields, J. Geophys. Res., 108(D13), 4376, doi:10.1029/2002JD003322, 2003.
• Latitude-Varying Decorrelation Length
  • Oreopoulos, L., D. Lee, Y.C. Sud, and M.J. Suarez, Radiative impacts of cloud heterogeneity and overlap in an atmospheric General Circulation Model, Atmos. Chem. Phys., 12, 9097-9111, doi:10.5194/acp-12-9097-2012, 2012.
• Full list of references