SPARC logo (45 Ko)
S P A R C

Stratospheric Processes A nd their Role in Climate
A project of the World Climate Research Programme

Home Initiatives Organisation Publications Meetings Acronyms and Abbreviations Useful Links

 

Report on the Workshop on Molecular Spectroscopy for Atmospheric Sensing
San Diego, USA, 23-26 October, 2001

Bhaswar Sen, Jet Propulsion Laboratory, USA (bhaswar.sen@jpl.nasa.gov)
Ken Jucks, Smithsonian Astrophysical Observatory, USA (jucks@cfa.harvard.edu)
Mike Kurylo, NASA/NIST, USA (mkurylo@hq.nasa.gov)

Introduction

Remote sensing and in situ measurements play critical roles in developing an understanding of the chemistry and physics of the Earth’s atmosphere and of its susceptibility to change as a result of natural and anthropogenic forcings. Improvements in laboratory spectroscopic measurements are required to meet the needs of these endeavours, particularly in light of the increasing demands of current sensing technology and of the increased accuracy and precision required to address current atmospheric science issues. Nevertheless, the laboratory spectroscopy needs for atmospheric sensing have not been assessed since the early preparations for the UARS (Upper Atmosphere Research Satellite) observatory [1]. Accordingly a workshop on Molecular Spectroscopy for Atmospheric Sensing was organised in San Diego, CA during October 23-26, 2001 (http://atmoschem.jpl.nasa.gov/). The workshop brought together laboratory molecular spectroscopists and investigators who use spectroscopic techniques to probe atmospheric processes to stimulate discussion, to define the needs of the atmospheric sensing community, and coordinate these needs with the current capabilities of laboratory spectroscopy. The programme was balanced between invited oral presentations, posters, and discussions of the issues raised.

Even though there are many spectroscopic issues related to atmospheric sensing that are common to all wavelengths, the workshop was divided into sessions devoted essentially to specific wavelength regions with a separate session concentrating on aerosol spectroscopy. This was done primarily because most atmospheric sensing instruments individually cover one wavelength region due to instrumentation limitations. Each wavelength region had two sessions, one oral session with two or more invited overview presentations with concentrations on the current atmospheric sensing instruments and state of the art laboratory spectroscopy and a poster session with presentations of specific laboratory and atmospheric sensing projects.

The overview talks for each of the sessions were given by J. Sloan and R. Niedziela (aerosol session); J. Waters and F. Delucia (microwave session); D. Johnson and M. Birk (far infrared session); P. Bernath, Ch. Webster , J.-M. Flaud, and Ch. Benner (mid and near infrared session); and E. Hilsenrath and J. Orphal (visible and ultraviolet session). Overview presentations can be viewed at http://atmoschem.jpl.nasa.gov/

The workshop was attended by 70 researchers and was split evenly among the atmospheric sensing community and laboratory spectroscopists. The institutions represented included the Jet Propulsion Laboratory (JPL), NASA Headquarters, NASA Langley, NASA Goddard, NASA Ames, National Institute of Standards and Technology, University of Denver, UCLA, Harvard-Smithsonian Center for Astrophysics, Depaul University, William & Mary College, University of Massachusetts Lowell, Massachusetts Institute of Technology, State University of New York, Ohio State University, Computer Sciences Corporation, University of Waterloo, University of Paris, DLR (Germany), Université Libre de Bruxelles, Japanese Meteorological Research Institute, Communications Research Laboratory (Tokyo), Institut für Meteorologie und Klimaforschung (Karlsruhe, Germany), and National Institute of Environmental Studies (Japan).

Aims

The main aims of the workshop were to:

Discussions and findings

A number of topics were common to the sessions of all the wavelength domains and are presented first to emphasize their importance. The topics include:
There was extensive discussion on the need for the formation of a recommendation panel for data that is included in molecular spectroscopy databases (e.g., HITRAN). An advisory group will comprehensively review and provide detailed specific recommendations for generating complete sets of parameters (e.g., line intensities, air-broadening coefficients, absorption cross sections as a function of temperature, sulphate aerosol optical constants as a function of temperature, etc.) for all atmospheric species with sufficient absorption to effect atmospheric absorption and emission observations and atmospheric radiation calculations. This panel will assign sub-panels assigned to specific molecules and/or wavelength region. The panel also will set guidelines for data submission, level of scientific review, validation procedures, quantification of uncertainties, and format of data. This has been performed in the past in an informal fashion for databases like HITRAN and the JPL sub-millimetre wave database. While the exact structure and procedures of this panel has yet to be determined, it was envisioned to operate in a fashion similar to that used previously for recommendations of spectroscopy in the visible and UV for the current ESA satellite instruments [2], the NASA Chemical Kinetics and Photochemistry panels (http://jpldataeval.jpl.nasa.gov/) and the IUPAC Gas Kinetics data evaluation panel (http://iupac.gas-kinetic.ch.cam.ac.uk/). NASA will sponsor this structure in the form of travel support for the panels to meet on a regular basis.

Uncertainties in pressure broadening coefficients can have a significant impact on the retrievals of molecules for instruments with high spectral resolution and for transitions that saturate over observational path lengths. Pressure broadening uncertainties also affect the use of line shape for the inference of altitude in nadir and zenith observations. Many of the laboratory observations of pressure broadening are taken at room temperature and not typical atmospheric temperatures. There is an obvious need for future pressure broadening line shape information, especially for water vapour, at all wavelengths.

Water vapour absorbs strongly from the microwave through the visible and is a significant contributor to the radiative budget of the atmosphere, convective transport, and atmospheric photochemistry. As a result, accurate spectroscopic parameters for water are especially important. However, many complications have prevented the spectroscopy of water vapour to be represented with sufficient accuracy to address the scientific questions being asked about it. These problems include the difficulty in determining the quantity of the water in the gas cell, the low vapour pressures at the lower atmospheric temperatures, the high centrifugal distortion values which confuse the calculations for line strengths, and the non-Lorenz line shapes that give a continuum opacity over a broad spectrum.

Potential solutions for solving these problems include improved rotational/vibrational coupled non-rigid-rotor models, performing the laboratory spectroscopy with multiple wavelength regions, and using atmospheric spectra directly in order to obtain longer pathlengths and colder temperatures.

1. Aerosol Spectroscopy

J. Sloan (University of Waterloo) and R. Niedziela (DePaul University) addressed the instrument characteristics and aerosol indices needed for remote measurement of Earth's atmosphere and the identification of atmospheric aerosols and the many aspects of laboratory studies of aerosol particles, respectively.

Although the UV/Visible region has good signal-to-noise characteristics for sizing aerosols, there is little spectral information about particle composition in this wavelength regime. The infrared region contains information about the chemical composition of observed particles, but sizing is more difficult. An instrument that can measure continuously over both wavelength regimes (e.g., ACE) offers distinct advantages over past mid-infrared measurements (e.g., ATMOS, ILAS). Retrievals of polar stratospheric clouds (NAT and H2SO4/HNO3 /H2O) over Antarctica using low-resolution ILAS spectra were shown. The largest uncertainty in the retrieval was the incomplete fitting of gas absorptions in a low-resolution spectra. High-resolution and high signal-to-noise ratio spectral measurements would alleviate many existing problems in aerosol retrieval. The presentation and discussion on laboratory measurements of the refractive indices of materials known or presumed to exist in particles of the upper troposphere and lower stratosphere focused on sulphates, organics, and biomaterials. An algorithm was presented for determining refractive indices from laboratory aerosol extinction measurements combined with series of spectra with no scattering components and with Mie scattering only.

From the discussions following the oral and poster presentations and a Thursday afternoon break-out session the spectroscopic requirements that are deemed important for atmospheric sensing are: (1) self-consistent sets of refractive indices across the UV to infrared spectral regions, (2) ternary solution refractive indices at polar stratospheric temperatures, (3) far-infrared refractive indices of sulphate, ice, and water at stratospheric temperatures, and (4) refractive indices of organics, mixtures, dusts, and urban pollution materials to characterise future satellite instruments that will probe into the troposphere.

2. Microwave Spectroscopy

J. Waters (Jet Propulsion Laboratory) and F. DeLucia (Ohio State University) addressed the spectral database and laboratory measurements necessary for remote sensing in the microwave, respectively. Currently ground-based, air-, balloon-, and space-borne sub-millimetre radiometers (e.g., MLS, ASUR, SLS, ODIN, SMILES) measure a long list of tropospheric and stratospheric molecules (H2O, O3, HCl, CO, N2O, HNO3, HCN, H2CO, CH3CN, H2O2, HNO4, ClO, OH, etc.). The principal spectroscopic deficiencies in microwave spectral parameters are linewidths and non-resonant absorption continuum (i.e., H2O). For both, the laboratory data are sparse and theoretical models imprecise. Although number of institutions (JPL, OSU, Ibaraki University, DLR) are performing highly precise rotational linewidth measurements, statistically relevant discrepancies persist. Inter-comparison of microwave and infrared spectral lineshape measurements could be made to resolve this issue. Similarly, combination of microwave and far- and mid-infrared measurements could assist in understanding air and H2O continua as no physical models exist to explain the phenomena.
From the discussions following the oral and poster presentations and a Thursday afternoon break-out session the spectroscopic requirements deemed important for atmospheric sensing are: (1) estimate uncertainties in linewidths, lineshifts, and continuum parameters, (2) a combined microwave and infrared investigation of spectral parameters for important but ''challenging'' gases (e.g., H2O, O3, HNO3) to ascertain the ability of current data set to reproduce spectrally resolved atmospheric measurements, (3) precise dipole moment measurements for important isotopomers (18O, HDO, BrO, acetone), and (4) frequency predictions for tropospheric and stratospheric species beyond 1 THz.

3. Far-, Mid-, and Near-Infrared Spectroscopy

D Johnson (NASA, Langley Research Center), C. Webster (Jet Propulsion Laboratory) and P. Bernath (University of Waterloo) addressed the far-, mid-, and near-infrared spectroscopic parameters needed for in situ and remote measurement of Earth's atmosphere. In the context of the meeting the region covers 20-10,000 cm-1 (or 1–500 mm). C. Webster noted that in situ measurements using narrow-band (0.005 cm-1) tunable diode and quantum cascade lasers operating in the 1-3 mm region achieve sensitivity currently not obtainable by remote sensing instruments. Measurements of linewidths and their temperature dependence were identified as a critical need for the in situ community. Spectral line parameters for H2O and CO2 isotopomers with 0.1 ‰ accuracy are needed to understand transport of water vapour across the tropical tropopause and the sources and sinks of atmospheric carbon, respectively. The remaining presentations and discussions in infrared sessions focused on the immediate need for consistent and precise O2 and CO2 spectroscopy (needed by remote sensing instruments for pressure and temperature retrievals), O3 line strengths and widths, H2O continuum, and HNO3 band strengths (hot bands in particular). M. Birk (DLR), J.-M. Flaud (University of Paris), and C. Benner (William and Mary College) addressed the many aspects of current laboratory studies and future needs for infrared remote sensing. These presentations were less of a review and more in form of requirements on the quantifiable quality of laboratory measurements. The metrology required for highly accurate spectroscopic measurements was thoroughly discussed including need for precise characterisation of instrument line shape, detector performance (i.e., non-linearity, baseline, and channelling), calibrations (i.e., wavenumber, pressure, and temperature), gas cell geometry, and isotopomer mixing ratio of measured gas. There was general agreement that high correlation between the measured spectral parameters (strengths, widths, and their temperature dependence) often results in unrecognised systematic errors. However, simultaneously analysis of spectra recorded over a wide range of conditions (i.e., pressure, temperature, resolution) could reduce systematic uncertainties. An important discussion focused on the SNR characteristics of atmospheric measurements necessary to discern errors in spectroscopic parameters. It was illustrated that while a SNR of 1000:1 was sufficient to measure a 1% error in halfwidths, a SNR of 50000:1 was needed to detect a 1 % error in zero level.
From the discussions following the oral and poster presentations and a Thursday afternoon break-out session the spectroscopic requirements deemed important for atmospheric sensing again are: (1) high precision near-IR measurements of H2O and CO2 (and their isotopomers) line intensities, widths, and their temperature dependence to support in situ tunable diode laser measurement programs aiming for sub-parts per billion accuracy, (2) accurate and consistent sets of O2 and CO2 spectroscopy (position and intensity measurements) in the mid- and far-IR, (3) O3 intensities and air-broadening coefficients in the 10-µm region since current measurements differ by about 4%, and (4) absolute band intensities of HNO3 between 390–3600 cm-1 for precise retrieval of nitric acid abundance.

4. Ultraviolet and Visible Spectroscopy

J. Orphal and E. Hilsenrath gave the visible and ultraviolet overviews of the current states and needs for laboratory and atmospheric spectroscopy, respectively. A number of general outlines for laboratory data were given to ensure consistency and usefulness of the data. These include:
  • instrumental lineshape to be reported for all spectra,
  • simultaneous measurements in other spectral regions where possible,
  • wavelength standards should be reported where possible (e.g. I2, hollow cathode lamps),
  • validation of laboratory spectra from atmospheric observations useful in many cases.
The topic of a well calibrated (both in frequency and intensity) high resolution solar spectrum was given considerable discussion. Such a spectrum would be invaluable for atmospheric observations from the near-infrared through the UV. To be most useful, the spectra should cover the range from 115 nm to 2.5 mm with a spectral resolution of 0.05 cm-1. The spatial resolution on the disk should be 1/10 and include a full disk average and be taken over a full solar cycle. Such a project might require collaboration with solar physics projects. For specific molecules, the main focus were on O2, O3, NO2, and H2O, with some discussion on molecules like BrO, ClO, and organic compounds. For some, the current laboratory is sufficient or nearly sufficient for the current and upcoming atmospheric observations. This includes O2, which needs some minor improvements in the Wulf bands, the Herzberg continuum, and the B-band, ClO, and OClO. O3 requires more observations of the temperature dependence of the cross sections throughout the visible and UV down to 180 K, preferably with high resolution FTS observations. H2O has significant problems. Most of the transitions cannot currently be assigned. As a result, the temperature dependence of the strengths cannot be characterised. Temperature dependence of the pressure broadening coefficients and shifts are also needed. Other molecules that require some work are O2 complexes, NO2, NO3, IO, BrO, and many organic compounds.

Acknowledgements

We wish to thank the participants of the workshop for their enthusiasm during the workshop. We wish to thank the rapporteurs: Dr. A. Eldering (JPL) and Dr. L. Iraci (NASA Ames) in aerosol, Drs. B. Drouin and W. Read (JPL) in microwave, Dr. M. Sirota (NASA GSFC) and Dr. B. Winnewisser (Ohio State University) in far-infrared, Dr. C. Rinsland (NASA Langley) and Dr. L. Rothman (Harvard/CfA) in mid- and near-infrared, and Dr. K. Pfeilsticker (University of Heidelberg) and Dr. S. Sander (JPL) in ultraviolet and visible for their assistance in preparation of the workshop summary. We wish to thank Dr. L. Brown (JPL), Ms. K. Thompson (CSC), and Ms. R. Kendall (CSC) for their invaluable assistance in organising the workshop. We also thank Dr. M. Kurylo (NASA/NIST) and the Upper Atmosphere Research Program for asking Drs. K. Jucks and B. Sen to organise the workshop, assisting in its organisation, and for funding both the workshop and numerous speakers and rapporteurs to attend.

References

[1] NASA Conference Publication 2396, edited by Mary Ann H. Smith, Langley Research Center, 1985.

[2] J. Callies and J. Orphal, A critical review of O3 and NO2 reference data for atmospheric remote-sensing in the ultraviolet, visible, and near-infrared, Earth Observation Quarterly, ESA, in press.