RTE Benchmarks

Radiative Transfer Equation (RTE) benchmarks using ARTS (Atmospheric Radiative Transfer Simulator). The benchmarks compute broadband shortwave (SW) and longwave (LW) radiative fluxes for several standard atmospheric test cases and compare them against reference flux data.

Version

v1.4

Test Cases

Setup	Description
ckdmip	CKDMIP (Correlated K-Distribution Model Intercomparison Project) atmospheres
rce	Radiative-Convective Equilibrium atmospheres
rfmip	RFMIP (Radiative Forcing Model Intercomparison Project) atmospheres

Repository Structure

data/            # Input atmospheric profiles and auxiliary data (ARTS XML)
results/         # Reference and computed flux outputs (NetCDF)
  <setup>/LW/
  <setup>/SW/
scripts/         # Python scripts and helper modules
  cache/         # LUT caches per setup / spectral band

Scripts

File	Purpose
convert_rte-examples2arts2.py	Convert rte-examples netCDF data to ARTS ArrayOfGriddedField4 XML
rte_benchmarks.py	Run single-profile SW/LW benchmark simulations (slower, more flexible to adjust outputs)
rte_benchmarks_batch.py	Run batch SW/LW benchmark simulations (faster, less flexible)
rte_benchmarks_overview.py	Generate comprehensive flux analysis plots and statistics
rte_aux_functions.py	Auxiliary functions (thermodynamics, unit conversions, flux computations)

Dependencies

• pyarts — ARTS Python interface
• pyarts-fluxes >= 0.8 — flux computation
• numpy, scipy, xarray

Workflow

1. Convert Input Data

Convert the rte-examples NetCDF data to ARTS XML format:

cd scripts
python convert_rte-examples2arts2.py

This reads source files from data/ and writes ARTS-compatible XML files.

2. Run Benchmarks

Single-profile mode:

python rte_benchmarks.py

Batch mode:

python rte_benchmarks_batch.py

Results are written to results/<setup>/LW/ and results/<setup>/SW/ as NetCDF files.

3. Generate Analysis Plots

Generate comprehensive visualization plots showing flux distributions and statistics:

python rte_benchmarks_overview.py

This script:

• Loads computed reference flux data from NetCDF files
• Computes mean and standard deviation profiles for all flux variables
• Creates altitude-stratified flux distributions via 2D histograms
• Generates a 2x3 subplot figure per setup and radiation type showing:
  • Mean upwelling/downwelling flux profiles
  • Flux variability (standard deviation) with altitude
  • 2D histograms of flux vs. altitude distributions
• Saves high-resolution PDF plots to plots/overview_<setup>_<radtype>.pdf

Author

Manfred Brath