Water line intensities in the near-infrared and visible

https://doi.org/10.1016/S0022-4073(03)00151-1Get rights and content

Abstract

Water is the single most important molecule for models of the earth's atmosphere but line parameters for water, particularly at shorter wavelengths, are difficult to measure reliably. We suggest that the most reliable way of generating water line parameters is to combine data obtained from a variety of sources, thereby separating line parameter determination into results for strong lines, for weak lines and for isotopically substituted water. Theoretical considerations which are addressed include line assignments and labeling of energy levels and the prospects of a full theoretical solution to the water vapor problem. Particular attention is paid to strong line absorption intensities in the near-infrared where recent studies have given significantly different results. The experimental data used to construct the ESA-WVR linelist (J. Mol. Spectrosc. 208 (2001) 32) is re-analyzed with a focus on effects due to pressure determination in the cell, subtraction of the baseline and parameterization of the line profiles. A preliminary re-analysis suggests that the line intensities given by the ESA-WVR study should be closer to those of Brown et al. (J. Mol. Spectrosc. 212 (2002) 57) used in the HITRAN. This shows the vital importance of validating the data for water by independent means.

Introduction

Water is molecule number one in the list of atmospheric species in the HITRAN database [1], [2]. Its number one status is a reflection of the dominant influence water has in both the absorption of sunlight as it passes through our atmosphere and its role as the predominant greenhouse gas. The rotation-vibration spectrum of water is not only complicated but also extensive. It is therefore unsurprising that there have been many detailed studies of the water spectrum, see [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18] for examples at the near-infrared and visible wavelengths which concern us here. Despite these studies, even basic line parameters. such the intensity of the strongest lines in the near-infrared and optical region, remain a subject of debate [19].

In particular the influence of relatively small changes (∼10%) in line intensities for water have been shown to have significant influence on radiation transfer in the earth's atmosphere. Indeed changes of this magnitude could account for much of the missing absorption in the atmosphere [20], [21]. Recently Belmiloud et al. [19] suggested that the data in the 1996 edition of HITRAN [1], and those corrected by Giver et al. [22], systematically underestimate the effects of strong water lines in the near-infrared and red region of the spectrum, among the most critical regions for these absorptions. The details of the experiment giving rise to this assertion and the associated parameters were subsequently published by Schermaul et al. [14], [23] and made available electronically as the ESA-WVR linelist [24]. The ESA-WVR linelist covered the range 8592–15000cm−1. In a parallel study Brown et al. [16] analyzed in detail the line parameters of water in the range 1015011190cm−1. Their complete data, which covered the extended range 9650–11190cm−1, is included in the 2000 edition of HITRAN. Brown et al.'s data also leads to an increase in the overall absorption by water but only by about half that given by ESA-WVR. Brown et al.'s analysis also emphasized the difficulty of getting reliable data for absorption by isotopically substituted water from spectra recorded for water in natural abundance. Coheur et al. [17] have also recently reported a long pathlength study of water vapor and water–air spectra which covers the 1300026000cm−1 region. This work again finds increased absorption by water vapor, by about 5%, compared to HITRAN but, in the region where they overlap, somewhat lower intensities than given by ESA-WVR [24].

In this paper we attempt to develop a strategy for determining reliable line intensities for water. This strategy addresses strong lines, weak lines, isotopically substituted water and spectral assignments. For the strong line data we re-analyze the data recorded as part of the ESA-WVR study to try and pinpoint the source of the discrepancy between that work and the study of Brown et al. Finally the need for independent means of testing the reliability of water vapor line parameters is emphasized.

Section snippets

Data sources: an overview

The large dynamic range of water absorption intensities required for atmospheric modeling means that line parameters are difficult, if not impossible, to determine from a single experiment. To accumulate the best possible data for each case, the most reliable way of generating water line parameters would therefore appear to be to combine data obtained from a variety of sources.

The wavelengths of strong line absorptions by water in the near-infrared and optical are well known [3], [4], [12], [13]

Re-analysis of the strong line intensities

Fig. 1 compares line intensities reported in the ESA-WVR [24] and HITRAN [2] databases. The figure shows clear systematic differences. A frequency-dependent analysis shows that the data reported by Brown et al. [16] largely gives a linear feature showing systematic differences at the less than 10% level, while larger systematic differences arise from longer wavelength spectra. Brown et al. made a particularly careful study and obtained results which differ from Schermaul et al.'s ESA-WVR work

Conclusions

Although the HITRAN data base is an invaluable source of data on water line parameters, there remains a number of outstanding problems with these parameters at near-infrared and visible wavelengths. Our strategy for tackling these issues is based on four separate approaches: analysis of water–air spectra to determine parameters for strong lines, analysis of long pathlength pure water spectra to obtain parameters for weaker water lines, analysis of isotopically enhanced spectra to obtain data on

Acknowledgements

We thank Jim Brault and Roland Schermaul for helpful conversations during the course of this work. We thank a referee for helpful comments on our original manuscript. This work has been supported by the UK Research Councils NERC and EPSRC, The Royal Society, the INTAS foundation and the Russian Fund for Fundamental Studies.

References (51)

  • P.F. Coheur et al.

    New water vapor line parameters in the 26 000–13000cm−1 region

    JQSRT

    (2002)
  • O. Naumenko et al.

    High-order resonance interactions in HDOanalysis of the absorption spectrum in the 14 980–15350cm−1 spectral region

    J Mol Spectrosc

    (2000)
  • L.P. Giver et al.

    Visible and near-infrared (H2O)-O-16 line intensity corrections for HITRAN-96

    JQSRT

    (2000)
  • R. Schermaul et al.

    The water vapour spectrum in the region 8600–15000cm−1: experimental and theoretical studies for a new spectral line database iiconstruction and validation

    J Mol Spectrosc

    (2001)
  • R.N. Tolchenov et al.

    Weak line water vapor spectrum in the 11,787–13,554cm−1 region

    J Mol Spectrosc

    (2002)
  • H. Naus et al.

    Cavity-ring-down spectroscopy on water vapor in the range 555–604nm

    J Mol Spectrosc

    (2001)
  • A. Bykov et al.

    The second decade of H218O—line positions and energy-levels

    J Mol Spectrosc

    (1995)
  • M. Tanaka et al.

    Absorption spectrum of H218O in the 12,400–14,520cm−1 range

    J Mol Spectrosc

    (2002)
  • D.W. Schwenke

    New H2O rovibrational line assignments

    J Mol Spectrosc

    (1998)
  • O.L. Polyansky et al.

    Water vapour line assignments in the near infrared

    J Mol Spectrosc

    (1998)
  • A. Jenouvrier et al.

    The visible and near ultraviolet overtone spectrum of HOD

    J Mol Spectrosc

    (2001)
  • Rothman LS, HITRAN 2000,...
  • J.P. Chevilliard et al.

    H218O—the (030), (110), and (011) interacting states—line positions and intensities for the 3v2, v1+v2 and v2+v3 bands

    Can J Phys

    (1985)
  • J.P. Chevilliard et al.

    H216Oline positions and intensities between 9500 and 11500cm−1. The interacting vibrational states (041), (220), (121), (022), (300), (201), (102) and (003)

    Can J Phys

    (1989)
  • R.A. Toth

    Extensive measurements of H216O line frequencies and strengths—5750–7965cm−1

    Appl Opt

    (1994)
  • Cited by (5)

    • IUPAC critical evaluation of the rotational-vibrational spectra of water vapor, Part III: Energy levels and transition wavenumbers for H<inf>2</inf><sup>16</sup>O

      2013, Journal of Quantitative Spectroscopy and Radiative Transfer
      Citation Excerpt :

      This criterion for the error is therefore more stringent than the usual standard deviation used to represent statistical error and will usually lead to the quoted MARVEL errors being systematically larger. For H216O, there exists an unusually large number of at least partially assigned experimental spectra [17,28,30,33–35,52–165]. The data from room-temperature spectra are augmented by data from a number of warm (400–700 K) and hot (up to 3000 K) H216O spectra [83,93,94,96–98,107,124,135,148,164,166–177.

    • The HITRAN molecular spectroscopic database: Edition of 2000 including updates through 2001

      2003, Journal of Quantitative Spectroscopy and Radiative Transfer
    • Water vapor absorption spectrum measurements and its application in concentration measurement

      2012, Proceedings of SPIE - The International Society for Optical Engineering
    • Application of wavelength scanning for measuring water vapour concentration by distributed laser diode

      2011, Proceedings of SPIE - The International Society for Optical Engineering
    1

    Permanent address: Institute of Applied Physics, Russian Academy of Science, Uljanov Street 46, Nizhnii Novgorod 603024, Russia.

    View full text