rundisort

A codebase for running DISORT from PYTHON to perform radiative transfer calculations for layered model atmospheres, including absorption and scattering. Inputs include gaseous optical depths, atmospheric parameters, and cloud properties. Radiative transfer is performed using DISORT 2.0 Beta (from ftp://climate1.gsfc.nasa.gov/wiscombe/Multiple_Scatt/). Using f2py, inputs and outputs can be passed directly to and from the DISORT fortran code. This now uses PYTHON versions 3 and up. This is a work in progress, and may have bugs. It is shared without guarantees of any kind. Improvements are ongoing, and we welcome feedback.

Dependencies:

Install gfortran via, e.g. brew install gcc.
Install Python (version 3+) on your computer if you don't already have it.

Setup and Testing:

Install the dependencies (above).
Clone the repository or copy the files in the source directory onto your computer.
Within the installation directory, modify the makefile as needed for your gfortran compiler. Run "make" to compile disort_driver_py.so (a python library).
If the output file is not named disort_driver_py.so, but rather something similar but with additional characters, link to disort_driver_py.so: e.g. $ ln -s disort_driver_py.cpython-310-darwin.so disort_driver_py.so.
(Optional) Test fortran by compiling tests/fortran_test/fortran_test.f and running the output executable. It should print "Fortran says hello" four times.
To prepare for testing f2py, run make in tests/test_f2py/test_f2py.py. This should create a file times3*.so which is used to test f2py (along with running other tests) in the step below).
Run the tests by running pytest from the main directory.

Authors

Penny Rowe https://github.com/prowe12
See also credits below

License

Acknowledge use by including the statement "By Penny Rowe - Own work, GNU GPLv3, https://github.com/prowe12/rundisort.

See the [LICENSE] LICENSE file for details.

Acknowledgments

(please let us know if any references are missing):

If this code is used in work leading to a publication, please reference: Rowe, P. M., Cox, C., Neshyba, S., & Walden, V. P. (2019). Toward autonomous surface-based infrared remote sensing of polar clouds: retrievals of cloud optical and microphysical properties. Atmospheric Measurement Techniques, 12(9), 5071–5086. http://doi.org/10.5194/amt-12-5071-2019
DISORT references:

Stamnes, K., Tsay, S.-C., Wiscombe, W. J., and Jayaweera, K.: Numerically stable algorithm for discrete-ordinate-method radiative transfer in multiple scattering and emitting layered media, Appl. Optics, 27(12), 2502–2509, doi:10.1364/AO.27.002502, 1988.

Stamnes, K., Tsay, S.-C., Wiscombe, W., and Laszlo, I.: DISORT, a general-purpose Fortran program for discrete-ordinate-method radiative transfer in scattering and emitting layered media: Documentation of methodology, Tech. rep., Dept. of Physics and Engineering Physics, Stevens Institute of Technology, Hoboken, NJ 07030, 2000.
For use of the temperature-dependent single-scattering parameters of liquid water (pmom files), a paper is in progress. In the meantime, please reference:

Rowe, P.M., S. Neshyba, and V. P. Walden, Radiative consequences of low-temperature infrared refractive indices for supercooled water clouds, Atmos. Chem. Phys., 13, 11925–11933, 2013. www.atmos-chem-phys.net/13/11925/2013/ doi:10.5194/acp-13-11925-2013

as well as referencing Zasetsky et al., 2005 and Wagner et al., 2005 as given within.
For use of the single-scattering parameters for liquid water at 300 K, please reference Downing, H. D. and Williams, D.: Optical constants of water in the Infrared, J. Geophys. Res., 80, 1656–1661, 1975 for the indices of refraction and please acknowledge Neshyba, Rowe, and Walden for creating the single scattering parameters (paper coming soon).
For use of the single-scattering parameters for spherical ice: Warren, S. G. and Brandt, R. E.: Optical constants of ice from the ultraviolet to the microwave: a revised compilation, J. Geophys. Res., 113, D14220, doi:10.1029/2007JD009744, 2008. (For other ice habits, we use Ping Yang's database. Please email penny@nwra.com for data and references.)
For use of the solar irradiance spectra: Kurucz, R.L., Synthetic infrared spectra, in Infrared Solar Physics, IAU Symp. 154, edited by D.M. Rabin and J.T. Jefferies, Kluwer, Acad., Norwell, MA, 1992.
For use of cloud_2012012006_cld1.mat (used by sample_run.m), please reference: Cox, C., Rowe, P. M., Neshyba, S., & Walden, V. P. (2016). A synthetic data set of high-spectral resolution infrared spectra for the Arctic atmosphere. Earth System Science Data Discussions, 1–29. http://doi.org/10.5194/essd-2015-40, in review for Earth System Science Data. Please also see our database of atmospheric profiles and cloudy and clear sky up and downwelling radiances characteristic of the Arctic at the Arctic Observing Network (AON) Arctic data repository (at https://www.aoncadis.org/dataset/AAIRO_spectra.html; doi:10.5065/D61J97TT).

A note on precision

Some variables in DISORT are single precision. Sensitivity studies indicate that best accuracy is achieved when all layer optical depths in typical model atmospheres are above 10^-5. For example, for a total optical depth of 0.5 and temperatures near 240 K, our studies indicate that including a layer optical depth of 10^-6 results in round-off errors in zenith downwelling radiance of ~0.2 mW/(m2 sr cm-1), whereas omitting this layer from the calculation causes an error of only ~0.004 mW/(m2 sr cm-1). Furthermore, omitting extremely thin layers saves computational time. For this reason, the code excludes upper atmospheric layers with optical depths below 10^-5 (this is done on a wavenumber-by-wavenumber basis). If such thin layers need to be included, it is possible to compile entirely in double-precision at a computational cost probably less than 20% (see DISORT documentation). As described by Istvan Laszlo (personal communication), "To run DISORT in double-precision you should use DISORTsp.f, which is currently only available in pre-version 3 distributions, like in DISORT2.0beta. To create the double-precision version it is, however, not sufficient to simply auto-double DISORTsp.f at compile time. You would first need to change all instances of R1MACH to D1MACH. Then you should compile all files with the auto-double option, except one file: RDI1MACH.f should be compiled without this option. Finally you would need to link all compiled files to create the executable."

Credits:

The DISORT code is DISORT2.0beta, from ftp://climate1.gsfc.nasa.gov/wiscombe/Multiple_Scatt/. Please be sure to acknowledge use of DISORT. Although provided here, it is identical to the original code except for two changes. In disort.f, the maximum number of layers, MXCLY has been increased from 6 to 120. A header that was printed to the screen has been suppressed.
Development of this Python suite of codes for running DISORT for layered atmospheres including both clouds and gases is by Penny Rowe, with much help from Von Walden and Steven Neshyba, as well as helpful insights provided by Christopher Cox. While a small project, work has proceeded over several years, and thus was funded from a variety of sources, including the National Aeronautics and Space Administration (NASA) Research Opportunities in Space and Earth Sciences program (contract NNX08AF79G), the National Science Foundation (NSF) Idaho Experimental Program to Stimulate Competitive Research (EPSCoR), and NSF award ARC-1108451. Rowe also acknowledges support from USACH-DICYT and Neshyba acknowledges support from the University of Puget Sound and the National Science Foundation under grant CHE – 1306366.
Python code is written by Penny Rowe, with contributions from Steven Neshyba (except as noted within).

Please email me with any questions or problems: penny@nwra.com

Thank you and hope it is useful! Dr. Penny M. Rowe