Questions: Line-by-Line Radiative Transfer Calculations

Question 1

Why are line-by-line (LBL) radiative transfer models not used as the radiation scheme in general circulation models (GCMs) that simulate global climate?

• A. LBL models are less accurate than the band models currently used in GCMs
• B. LBL models require spectroscopic databases that are not publicly available
• C. An LBL calculation must evaluate absorption at millions of spectral points per atmospheric column, making it far too slow for the thousands of columns and time steps in a GCM
• D. LBL models cannot handle the temperature and pressure variations in the atmosphere

Answer: (C)

A single LBL calculation evaluates absorption at over a million spectral points across 50–100 atmospheric layers — a computationally enormous task. A GCM runs thousands of atmospheric columns at thousands of time steps over months of simulation time. Multiplying these numbers makes direct LBL integration impossible even on modern supercomputers. GCMs therefore use faster approximations (correlated-k, band models) that group absorption lines. LBL models are actually more accurate than band models — that is precisely why they serve as the benchmark for validating the faster schemes.

Question 2

What is the primary role of line-by-line calculations in modern climate science, given that they are too slow for GCMs?

• A. They are used to discover new greenhouse gases whose spectral lines have not yet been measured
• B. They serve as benchmarks to validate the faster radiation parameterization schemes used in GCMs
• C. They are used to run short-duration climate simulations when the highest possible accuracy is needed
• D. They replace satellite observations when direct measurements are unavailable

Answer: (B)

LBL calculations serve as the gold standard for accuracy. When a GCM's radiation scheme (using band models or correlated-k methods) computes the radiative forcing from doubled CO₂, that number is trustworthy because it has been validated against LBL results for standardized atmospheric profiles. The benchmark role is essential: without it, climate modelers would not know whether their faster approximations introduce significant errors. LBL models are also used directly in remote sensing retrievals, where spectral precision determines retrieval accuracy.

Question 3

Line-by-line radiative transfer models are the most computationally demanding option, but their accuracy is ultimately limited by approximations in the radiative transfer equations they solve.

• T. True
• F. False

Answer: False

The accuracy of LBL models is limited primarily by the completeness and precision of spectroscopic databases (like HITRAN) and by how well the atmospheric temperature and composition profiles are known — not by approximations in the radiative transfer method. The LBL approach itself makes essentially no approximation in the physics: it resolves individual spectral lines and applies Beer-Lambert law and Planck emission layer by layer. The uncertainty lies in the input data (line positions, intensities, broadening parameters) rather than in the mathematical method.

Question 4

A line-by-line model evaluates the absorption coefficient at hundreds of thousands to millions of individual spectral points because each atmospheric gas absorbs radiation at many discrete, narrow wavelengths rather than smoothly across broad bands.

• T. True
• F. False

Answer: True

This is exactly right. Each gas molecule (CO₂, H₂O, O₃, CH₄, etc.) transitions between quantized rotational and vibrational energy states at specific, discrete frequencies. These absorption lines are narrow — widths of ~0.1 cm⁻¹ due to pressure broadening — and there are millions of them across the thermal infrared spectrum. A band model that divides the spectrum into broad bins misses the fine structure within each bin. LBL calculations resolve every line individually, requiring spectral points spaced at ~0.001 cm⁻¹ or finer, which is why the number of evaluation points is so enormous.

Question 5

Why must an LBL radiative transfer model resolve individual spectral lines at very high spectral resolution rather than grouping absorption into broad wavelength bands?

Model answer: Atmospheric gases absorb radiation at specific discrete frequencies determined by their quantum energy levels, not uniformly across broad bands. Within any broad band, there are regions of very strong absorption (at line centers) and very weak absorption (between lines). Radiation at frequencies near line centers is absorbed near the surface; radiation between lines escapes to space. A broad-band average smooths out this structure and systematically misrepresents the actual flux reaching each altitude. Resolving individual lines captures the real physical behavior — particularly important for optically thick lines of CO₂ and H₂O that dominate the greenhouse effect.

The practical consequence of band averaging is that it introduces systematic errors in computed fluxes that are acceptable for fast GCM runs but unacceptable for a benchmark. For example, the radiative forcing from doubling CO₂ depends on the detailed shape of CO₂ absorption lines in the 15 μm band — specifically, whether the line wings (where absorption is weaker) are correctly represented. Band models parameterize this using the correlated-k method, which works well but introduces approximation errors of a few percent. LBL models eliminate these errors by resolving every line, making them the definitive reference.