Climate Audit

By 1999, systematic errors in the HITRAN-96 database for NIR water vapor absorption (which oppose water vapor feedback!) had been identified and widely reported by the most eminent authorities. For example,

Systematic errors have been found and corrected in the HITRAN (High Resolution Transmission Molecular Absorption Database) water vapor line absorption intensities in the visible and near-infrared spectral regions. The HITRAN data base has been used extensively in the calculation of atmospheric absorption of solar radiation. The most important corrections found were a 14.4% increase of the intensity of the 940 nm band and an 8.7% increase of the intensity of the 820 nm band. These systematic errors in the HITRAN tabulations were due to errors in the unit conversion from the measurements published in cm-1/(cm-atm) to HITRAN’s common units cm-1/(molecule/cm2). Since the absorption of water vapor in these important regions is greater than has been used in model calculations for the earth’s atmospheric absorption, there is a diminished necessity for an hypothesized "continuum absorption" in the atmosphere.

These errors impacted all the GCMs used in IPCC TAR, but the potential problem was not disclosed by IPCC as a factor of potential concern.

I’ll provide a few references.

P. Varanasi, Q. Zou, and C. Sun, V. Nemtchinov, L. P. Giver and C. Chackerian, Jr., S. Mathur, B. Ranganayakamma, T. Lippa, P. Kabro, and C. R. Prasad, Laboratory Measurement of the Absorption Spectra of Water Vapor and Carbon Dioxide Relevant to the ARM Program, Ninth ARM Science Team Meeting Proceedings, San Antonio, Texas, March 22-26, 1999 here

Important new concerns have arisen recently during the process of applying the currently available molecular spectroscopic data bases such as HITRAN (HIgh-Resolution TRANmission; Rothman et al. 1996) and GEISA (Gestion et Etude des Informations Spectroscopiques Atmospheriques; Jacquinet-Husson et al. 1999). These range from a) the lack of accurate data on the spectral lines of CO2 in the hot bands and in the n2-fundamental band in the 15-mm region, the spectral lines of water vapor in the n2-fundamental band at 6.3 mm and in the thermal infrared (8 mm to 12 mm) region; b) the spectral lines in all of the shortwave (500 nm to 1100 nm) bands of water vapor; to c) the difference, if any, in the absorption characteristics of water vapor near saturation, especially in the shortwave region. The concerns b) and c) arose on account of the recently identified enhanced absorption of solar radiation by the atmosphere.

L. P. Giver, C. Chackerian, Jr., and D. W. Schwenke, R. S. Freedman and M. D. DiRosa, P. Varanasi, R. L. Sams, Water Vapor Line Intensity Corrections and Rovibrational Assignment Updates of the Shortwave HITRAN and GEISA Databases Tenth ARM Science Team Meeting Proceedings, San Antonio, Texas, March 13-17, 2000 here

While measuring weak water vapor lines in the wings of the 0.94-mm band on spectra obtained with the 25-m base-path White cell and Bomem DA8 Fourier-transform spectrometer (FTS), we compared some of our measured intensity data to those listed in the HIgh resolution TRANsmission molecular absorption (HITRAN-96), Rothman et al. (1996), database, and also to the prior measurements of Chevillard et al. (1988). Our measurements were, on the average, about 20% higher than the entries in HITRAN, but generally compatible with the intensities reported by Chevillard et al. (1988). This was strange, since HITRAN refers to the Chevillard publication as the source data for most of the lines listed in the 9500-cm-1 to 11500-cm-1 region. We therefore selected over 50 of the best-measured lines from the tables of Chevillard et al. (1988) to compare with the HITRAN values. About half of these lines were previously measured by Giver et al. (1982); each of these prior line measurements agreed with the corresponding Chevillard et al. (1988) measurement within 6%, which was their uncertainty estimate. After making this comparison and finding that the HITRAN intensities did not agree with Chevillard’s published values, we made similar comparisons of the experimental data from four other articles reporting the line intensities in the visible and near-infrared that formed the basis for HITRAN lists in the respecive spectral regions. The four articles are by Camy-Peyret et al. (1986), Mandin et al. (1986), Mandin et al. (1988), and Toth (1994b). The measured intensity data are all in the units of cm-1/(cm-atm) at room temperature. All the measurements described in the five mentioned articles are Tenth ARM Science Team Meeting Proceedings, San Antonio, Texas, March 13-17, 2000 based on FTS spectra obtained with the 6-m base-path White cell at the Kitt Peak solar telescope and they all contain reports of intensities in the units of cm-2 atm-1. The unit used in the HITRAN database for the line intensity is cm/molecule at 296. The corrections we refer to here point to several cases of oversight by the creators of the database during the process of converting the measured intensity data reported in the five source papers into the units adapted by the creators of the database. The corrections needed to bring the HITRAN intensities into agreement with the published measurements have been described by Giver et al. (2000). These corrections only apply to assigned lines of the main isotopomer of water reported in the above cited five publications of experimental data, but these lines account for over 97% of the absolute intensity in each of the wavenumber intervals of interest. The corrections are described in detail by Giver et al. (2000). The list of parameters for lines between 8000-cm-1 to 25200-cm-1, with the proper intensity revisions, has been posted on the “database updates” page of the HITRAN Internet site: http://www.hitran.com. These corrections affect the Gestion et Etude des Informations Spectroscopiques Atmospheriques (GEISA), Jacquinet-Husson et al. (1999), database as well.

Giver, L. P., C. Chackerian, Jr., and P. Varanasi, 2000: Visible and near-infrared H2 16O line intensity corrections for HITRAN-96. J.Q.S.R.T., 66, 101-105.

I’ll provide a few other references on developments, when infrared specialists went back and checked NIR absorption, and on GCM impacts.