Cavity ring-down spectroscopy of CO2 near λ = 2.06 µm: Accurate transition intensities for the Orbiting Carbon Observatory-2 (OCO-2) "strong band". (September 2020)

• Record Type: Journal Article
• Authors: Fleurbaey, Hélène; Yi, Hongming; Adkins, Erin M.; Fleisher, Adam J.; Hodges, Joseph T.
• Is Part Of: Journal of quantitative spectroscopy & radiative transfer. Volume 252(2020)
• Publication Date: 2020-09
• Subjects: Precision molecular spectroscopy -- Line lists -- Carbon dioxide -- Cavity ring-down spectroscopy -- Dipole moment surfaces -- Herman-Wallis expansion; Spectrum analysis -- Periodicals; Radiation -- Periodicals; Analyse spectrale -- Périodiques; Rayonnement -- Périodiques; Radiation; Spectrum analysis; Periodicals; 543.0858
• DOI: 10.1016/j.jqsrt.2020.107104
• Languages: English
• ISSNs: 0022-4073

Abstract

Highlights: Accurate CO2 transition intensities were measured near λ = 2.06 µm by FS-CRDS. The relative combined standard uncertainty for all measured intensities is 0.08 %. Comparisons with existing databases revealed systematic biases as large as 1 %. We scaled the ab initio DMS of Zak et al. to yield S band 626 = 7.183 × 10 −21 cm molecule −1 . Results constrained the Herman-Wallis expansion, with an RMSD of σ ^ = 0.18 %. Abstract: The λ = 2.06 µm absorption band of CO2 is widely used for the remote sensing of atmospheric carbon dioxide, making it relevant to many important top-down measurements of carbon flux. The forward models used in the retrieval algorithms employed in these measurements require increasingly accurate line intensity and line shape data from which absorption cross-sections can be computed. To overcome accuracy limitations of existing line lists, we used frequency-stabilized cavity ring-down spectroscopy to measure 39 transitions in the 12 C 16 O2 absorption band. The line intensities were measured with an estimated relative combined standard uncertainty of u r = 0.08 %. We predicted the J -dependence of the measured intensities using two theoretical models: a one-dimensional spectroscopic model with Herman-Wallis rotation-vibration corrections, and a line-by-line ab initio dipole moment surface model [Zak et al. JQSRT 2016;177:31-42]. For the second approach, we fit only a single factor to rescale the theoretical integrated band intensity to be consistent with the measured intensities. We find that the latter approach yields an equally adequate representation of the fitted J -dependent intensity data and provides the most physically general representation of the results. Our recommended value for the integrated band intensity equal to 7.183 × 10 −21 cm molecule −1 ± 6 × 10 −24 cm molecule −1 is based on the rescaled ab initio model and corresponds to a fitted scale factor of 1.0069 ± 0.0002. Comparisons of literature intensity values to our results reveal systematic deviations ranging from −1.16 % to +0.33 %.