The HITRAN2024 methane update

TL;DR

HITRAN2024 advances methane spectroscopy by integrating multiple high-quality line lists (ExoMol, TheoReTS, HITEMP, MeCaSDa) and validating them against high-resolution FTS and CRDS data. The update replaces tens of thousands of transition wavenumbers and intensities, adds thousands of new lines, and extends coverage to 14,000 cm$^{-1}$ across 12CH4, 13CH4, and 12CH3D, with a focus on pentad, octad, and tetradecad regions. It also introduces speed-dependent line-shape parameters and first-order line mixing for select regions, leveraging Padé-approximants to model pressure-broadening across broad spectral ranges. The consolidated results yield a HITRAN2024 methane line list of about 471,000 lines, substantially reducing residuals in key spectral windows and enabling more accurate atmospheric and exoplanetary radiative transfer simulations, while outlining paths for future enhancements in line mixing, higher-quanta assignments, and planetary-broadening data.

Abstract

Spectroscopic parameters of methane from many different studies were gathered to improve the HITRAN database towards its 2024 version. After a validation process using high-resolution FTS and CRDS spectra, about 80,000 lines of the four most abundant isotopologues were replaced from the dyad to the triacontad regions. These changes amount to 51,000 transition wavenumbers, 18,000 line intensities, 33,000 pressure-broadening half-widths, and 3300 assignments. 44,000 new lines were added with 16,000 old lines removed, extending the database from 12,000 cm$^{-1}$ up to 14,000 cm$^{-1}$, and covering some gaps. A greater focus was brought on the pentad, octad, and tetradecad regions, targeted by several remote sensing instruments. In these regions, comparisons of spectral fits from multiple line lists were performed, taking only the parameters that provide best fit for each line. In the $ν_3$ band, in addition to replacing the previous values, speed-independent pressure broadening parameters of $^{12}$CH$_4$ were gathered and used to fit Padé-approximant functions. These functions then replaced any outdated experimental data in $ν_3$, missing data in the new lines, as well as the values that were determined to be outside their physical boundaries. The CH$_3$D broadening parameters were replaced in the same manner, for missing and low or high values, using a semi-empirical formula instead.

Figures (16)

• Figure 1: Residuals obtained at 296 K between experiment from $S2$ Bertin & Vander Auwera bertin_co2_2024 and Voigt calculations from the line lists used for the HITRAN2024 update
• Figure 2: Comparison of synthetic spectra generated with HITRAN2020 (in pink) and our updated composite line list (in red) with high-resolution Kitt Peak FTS spectrum at 291 K (Left, in black) and JPL-FTS spectrum at 80 K (Right, in blue). This plots show better performance and broader spectral coverage by the new composite line list. Note that the synthetic spectra were computed for the same conditions as the corresponding laboratory spectra. Any H2O (not presented) and 12CH4 features (in green) were identified during the line-by-line matching, as presented for 12CH4 features.
• Figure 3: Residuals obtained at 296 K and 0.17143 hPa between the experiment $S3$ from Ref. birk_esa_2017 and Voigt calculations from the line lists used for the HITRAN2024 update.
• Figure 4: Residuals obtained at 198 K and 0.16941 hPa between the experiment $S3$ from Ref. birk_esa_2017 and Voigt calculations from the line lists used for the HITRAN2024 update.
• Figure 5: Residuals obtained between experiment $S3$ from Birk and Voigt calculations from the past and new HITRAN versions