A two-stream radiative transfer code for 3D climate models and offline 1-D calculations.
📦 GitHub Repository
✉️ Eric T. Wolf — eric.wolf@colorado.edu
ExoRT is a flexible two-stream radiative transfer code designed for use with 3D climate models. It includes builds for a 1-D offline version and for direct interfacing with CESM1.2.1-ExoCAM and now with CESM3-planets.
If you use ExoRT, please cite:
Wolf, E. T., Kopparapu, R., Haqq-Misra, J., & Fauchez, T. J. (2022). ExoCAM: A 3D Climate Model for Exoplanet Atmospheres. The Planetary Science Journal, 3, 7. https://doi.org/10.3847/PSJ/ac3f3d
- see the ExoCAM Github Repository for details and instructions: ExoCAM
- Recommended version for CESM1.2.1-ExoCAM (as of September 2020):
3dmodels/src.cam.n68equiv - Files in
3dmodels/src.cam.n68equiv/are kept identical to their counterparts insource/src.n68equiv/andsource/src.main/. The only intentional exception issys_rootdir.F90, which encodes a machine-specific data path. When updating gas species insource/, always sync the corresponding files to3dmodels/in the same commit.
- ExoRT n68equiv is now natively integrated into the CESM3-planets framework as an external component.
- NEW for CESM3-planets (as of April 2026):
3dmodels/src.cam7.n68equiv - GitHub Repository: NCAR/CESM3-planets
- Project Wiki: CESM3-planets Wiki
ExoRT/
├── source/
│ ├── src.main/ # Drivers for offline calculation and shared radiation routines
│ ├── src.misc/ # Miscellaneous files and stubs from CESM origin (needed for offline runs)
│ └── src.n*/ # Radiative transfer versions (n## = number of spectral intervals + descriptor)
├── data/
│ ├── cia/ # Collision-induced absorption data
│ ├── cloud/ # Cloud optical properties (Mie)
│ ├── continuum/ # CO2, H2O continuum coefficients (from MTCKD / LBLRTM)
│ ├── kdist/ # Correlated k-distributions
│ └── solar/ # Stellar spectra
├── iofiles/ # Input/output files for the 1-D model
├── build/ # Build directory for the 1-D model
├── run/ # Run directory for the 1-D model
├── 3dmodels/ # File sets to be linked with CESM
└── tools/ # Pre/post-processing scripts (Python and IDL)
Recommended for all terrestrial planet cases as of September 2020.
The H₂O/CO₂/CH₄/C₂H₆ correlated-k data correspond to the configuration published in Wolf et al. (2022). NH₃ and CO (HITRAN 2024) were added later as additional absorbers on top of that baseline and do not affect runs that omit those species.
- Correlated-k coefficients from HELIOS-K (Grimm et al. 2015)
- H₂O: HITRAN 2016, Voigt lineshape, 25 cm⁻¹ cutoff, plinth removed; self/foreign continuum from MT_CKDv3.3 fit to Gauss points
- CO₂: HITRAN 2016, Perrin & Hartmann (1989) sub-Lorentzian lineshape, 500 cm⁻¹ cutoff, with CO₂–CO₂ CIA
- CH₄: HITRAN 2016, Voigt lineshape, 25 cm⁻¹ cutoff
- C₂H₆: HITRAN 2016, Voigt lineshape, 25 cm⁻¹ cutoff
- NH₃: HITRAN 2024, Voigt lineshape, 25 cm⁻¹ cutoff (added 2026-04-27; 3-D interface 2026-04-29)
- CO: HITRAN 2024, Voigt lineshape, 25 cm⁻¹ cutoff (added 2026-04-27; 3-D interface 2026-04-29)
- CIA: N₂–N₂, N₂–H₂, H₂–H₂ from HITRAN; CO₂–H₂ and CO₂–CH₄ from Turbet et al. (2020)
- 68 spectral intervals, 8 Gauss points
- Gas overlap via equivalent extinction absorption method (Amundsen et al. 2016): major gas treated with full 8-point correlated-k; minor species added as grey absorbers, selected on the fly
- Pressure range: 10 bar – 0.01 mb
- Temperature range: 100 K – 500 K
- Reference: Wolf et al., PSJ 3:7 (2022)
- Same as
n68equiv, with additional bins shortward of 0.24 µm - Use for F stars (6500 K < T < ~10,000 K)
Designed for Archean climate simulations.
- Species: H₂O, CO₂, CH₄, N₂, H₂; HITRAN 2004, 28 spectral bins
- H₂O and CO₂ continuum from MT_CKD2.5 (see Halevy et al. 2009)
- Mixed gas k-distributions via LBLRTM (Mlawer et al. 1997; Shi et al. 2009)
- CO₂ up to multi-bar; reasonable agreement at 2-bar dry CO₂
- CH₄ up to 0.01 bar
⚠️ Older line list misses CH₄ near-IR absorption - H₂O
⚠️ Overestimates near-IR absorption around M-dwarfs due to coarse bands - Up to 100 bar total pressure; N₂–N₂, N₂–H₂, H₂–H₂ CIA
- Reference: Wolf & Toon, Astrobiology 13(7), 1–18 (2013)
- Species: H₂O, N₂, H₂; 42 spectral bins; HITRAN 2012
- H₂O single-gas k-distributions via HELIOS-K (Grimm et al. 2015)
- Up to 10 bar total pressure; N₂–N₂, N₂–H₂, H₂–H₂ CIA
- Reference: Kopparapu et al., ApJ 845:5 (2017)
- Same as
n42h2o, extended to 68 spectral bins
cd ../ExoRT/build
# Build specific version(s)
make n42h2o
make n68equivmacOS users: the default compiler is
ifort, which Intel discontinued and never ported to Apple Silicon (arm64). On any modern (M-series) Mac you must build with gfortran by addingUSER_FC=gfortranto everymakecommand, e.g.make USER_FC=gfortran n68equiv. See macOS (Apple Silicon) below.
The executable is copied to ../ExoRT/run/.
The default compiler is ifort. Ensure nf-config is on your PATH
(loaded automatically via the NetCDF module on Discover):
module load netcdf
make n68equivifort is not available on Apple Silicon — Intel's compiler was never ported to arm64 and was discontinued in 2023. Use gfortran instead:
make USER_FC=gfortran n68equivDependencies: NetCDF-Fortran via Anaconda is recommended:
conda install -c conda-forge netcdf-fortranVerify nf-config is on your PATH before building:
which nf-configRuntime environment: The NetCDF libraries must be findable at runtime.
csh/tcsh:
setenv DYLD_LIBRARY_PATH /opt/anaconda3/libbash:
export DYLD_LIBRARY_PATH=/opt/anaconda3/libAdd the appropriate line to your ~/.cshrc or ~/.bashrc.
Note on compiler warnings: The gfortran build produces warnings from
CESM-heritage infrastructure files (infnan.F90, wrap_nf.F90,
shr_sys_mod.F90) that ifort accepts silently. These are suppressed with
-w in the Makefile and are benign for the current codebase, but are
candidates for cleanup during refactoring of src.misc.
-
Place your input file
RTprofile_in.ncin the../ExoRT/run/directory.
Template inputs are in../ExoRT/iofiles/. -
Set the solar spectrum and number of vertical levels in:
source/src.main/exoplanet_mod.F90 -
Run the executable:
cd ../ExoRT/run ./n42h2o.exe
Output is written to RTprofile_out.nc.
Note:
pverp = pver + 1(interface levels = midpoint levels + 1)
Gas species variables are optional: if a variable is absent from the input file, ExoRT sets that species to zero and prints a clean diagnostic rather than an error. Only TS, PS, the pressure/temperature/height arrays, the albedos, coszrs, mw, and cp are required.
| Variable | Dimension | Required | Description |
|---|---|---|---|
ts |
(1) | yes | Surface temperature (K) |
ps |
(1) | yes | Surface pressure (Pa) |
tmid |
(pver) | yes | Temperature at layer midpoints (K) |
tint |
(pverp) | yes | Temperature at layer interfaces (K) |
pdel |
(pver) | yes | Pressure thickness of each layer (Pa) |
pint |
(pverp) | yes | Pressure at interface levels (Pa) |
zint |
(pverp) | yes | Height at interfaces (m) |
asdir |
(1) | yes | SW albedo, direct |
asdif |
(1) | yes | SW albedo, diffuse |
aldir |
(1) | yes | Near-IR albedo, direct |
aldif |
(1) | yes | Near-IR albedo, diffuse |
coszrs |
(1) | yes | Cosine of solar zenith angle |
mw |
(1) | yes | Molecular weight of dry air (g/mol) |
cp |
(1) | yes | Specific heat of dry air (J/kg/K) |
h2ommr |
(pver) | optional | H₂O specific humidity, kg(wv)/kg(air) |
co2mmr |
(pver) | optional | CO₂ mass mixing ratio (dry) |
ch4mmr |
(pver) | optional | CH₄ mass mixing ratio (dry) |
c2h6mmr |
(pver) | optional | C₂H₆ mass mixing ratio (dry) |
nh3mmr |
(pver) | optional | NH₃ mass mixing ratio (dry) |
commr |
(pver) | optional | CO mass mixing ratio (dry) |
o2mmr |
(pver) | optional | O₂ mass mixing ratio (dry) |
o3mmr |
(pver) | optional | O₃ mass mixing ratio (dry) |
h2mmr |
(pver) | optional | H₂ mass mixing ratio (dry) |
n2mmr |
(pver) | optional | N₂ mass mixing ratio (dry) |
Use tools/makeColumn.py to generate input files. Species with VMR set to 0.0 are automatically omitted from the output file.
Contains all input variables plus the following radiative flux and heating rate fields.
| Variable | Dimension | Description |
|---|---|---|
LWUP |
(pverp) | Longwave upwelling flux (W m⁻²) |
LWDN |
(pverp) | Longwave downwelling flux (W m⁻²) |
SWUP |
(pverp) | Shortwave upwelling flux (W m⁻²) |
SWDN |
(pverp) | Shortwave downwelling flux (W m⁻²) |
LWUP_SPECTRAL |
(ntot_wavlnrng, pverp) | Spectral LW upwelling flux (W m⁻² per interval) |
LWDN_SPECTRAL |
(ntot_wavlnrng, pverp) | Spectral LW downwelling flux (W m⁻² per interval) |
SWUP_SPECTRAL |
(ntot_wavlnrng, pverp) | Spectral SW upwelling flux (W m⁻² per interval) |
SWDN_SPECTRAL |
(ntot_wavlnrng, pverp) | Spectral SW downwelling flux (W m⁻² per interval) |
LWHR |
(pver) | Longwave heating rate (K/Earth day) |
SWHR |
(pver) | Shortwave heating rate (K/Earth day) |
| Directory | Contents |
|---|---|
data/aerosol/ |
Aerosol optical constants and CARMA production rates |
data/cia/ |
Collision-induced absorption data |
data/continuum/ |
CO₂ and H₂O continuum coefficients from MTCKD (LBLRTM) |
data/kdist/ |
Correlated k-distributions |
data/cloud/ |
Cloud optical properties (Mie scattering) |
data/solar/ |
Stellar spectra |
Python scripts in tools/ handle the primary 1-D pre/post-processing workflow. Legacy IDL scripts (.pro) remain for reference.
tools/makeColumn.py — generates RTprofile_in.nc from a P-T profile and user-specified gas mixing ratios. Edit the USER SETTINGS block at the top or pass CLI flags:
cd tools/
python makeColumn.py --defaults # show current settings
python makeColumn.py --output RTprofile_in.nc # use USER SETTINGS
python makeColumn.py --profile TS273K \
--co2vmr 400e-6 --nh3vmr 1e-6 \
--output RTprofile_in_nh3.nc
python makeColumn.py --zero-h2o --output RTprofile_in_dry.ncGas species with VMR = 0.0 are omitted from the file; ExoRT reads absent species as zero. Available P-T profiles (from profile_data.py):
| Tag | Description |
|---|---|
TS273K |
Moist-adiabat Earth, Ts = 273 K, 300 levels |
US1976 |
US Standard Atmosphere 1976, 49 levels |
smart_2bar_t250 |
2-bar dry CO₂ Mars atmosphere, Ts ≈ 250 K, 69 levels |
tools/plotspectra_1D.py — plots SW and LW spectra from RTprofile_out.nc. The spectral grid is auto-detected from the file. Supports side-by-side comparison of two runs.
cd tools/
python plotspectra_1D.py # interactive display
python plotspectra_1D.py --save # save spectra_sw.png / spectra_lw.png
python plotspectra_1D.py --f2 ../run/RTprofile_out_ref.nc --label2 "no NH3"SW panel: wavelength axis (µm), W m⁻² µm⁻¹. LW panel: wavenumber axis (cm⁻¹), W m⁻² cm. Also prints a band-by-band flux table to stdout.
ExoRT applications
- Wolf, E.T. & Toon, O.B. (2013). Hospitable Archean Climates Simulated by a General Circulation Model. Astrobiology, 13(7), 1–18. https://doi.org/10.1089/ast.2012.0936
- Kopparapu, R.K. et al. (2017). Habitable Moist Atmospheres on Terrestrial Planets near the Inner Edge of the Habitable Zone around M Dwarfs. ApJ, 845:5. https://doi.org/10.3847/1538-4357/aa7cf8
Correlated-k / line-by-line
- Grimm, S.L. & Heng, K. (2015). HELIOS-K: An Ultrafast, Open-source Opacity Calculator for Radiative Transfer. ApJ, 808:182. https://doi.org/10.1088/0004-637X/808/2/182
- Mlawer, E.J. et al. (1997). Radiative transfer for inhomogeneous atmospheres: RRTM, a validated correlated-k model for the longwave. JGR Atmospheres, 102(D14), 16663–16682. https://doi.org/10.1029/97JD00237
- Shi, G. et al. (2009). An extension of the correlated-k distribution method to non-homogeneous atmospheres. JQSRT, 110(18), 2268–2278. https://doi.org/10.1016/j.jqsrt.2009.05.005
Gas overlap
- Amundsen, D.S. et al. (2016). Accurate treatment of transmission, emission, and scattering in parameterised radiation schemes for a general circulation model. A&A, 564, A59. https://doi.org/10.1051/0004-6361/201323169
Line shapes / CO₂ broadening
- Perrin, M.Y. & Hartmann, J.M. (1989). Temperature-dependent measurements and modeling of absorption by CO₂–N₂ mixtures in the far line-wings of the 4.3 µm CO₂ band. JQSRT, 42(4), 311–317. https://doi.org/10.1016/0022-4073(89)90077-0
Continuum absorption
- Halevy, I., Pierrehumbert, R.T. & Schrag, D.P. (2009). Radiative implications of CO₂ cloud formation in a Neoproterozoic snowball Earth. JGR Atmospheres, 114, D18112. https://doi.org/10.1029/2009JD011915
CIA data
- Turbet, M. et al. (2020). CO₂ condensation is a serious limit to the deglaciation of Earth-like planets. Earth and Planetary Science Letters, 531, 115759. https://doi.org/10.1016/j.epsl.2019.115959
- Gordon, I.E. et al. (HITRAN collaboration) — CIA data retrieved from the HITRAN database.