A program to measure with the highest accuracy the line parameters of high resolution Fourier transform spectra recorded in the laboratory has been written. Tests of the accuracy of the calculated parameters have been performed by comparing the results of the analysis of the 2-0 band of CO and of the v3 + v4 band of CO2. The line position accuracy has been found to be 9 by 10-6, and the line intensity accuracy is of the order of 1 percent or better.
This work reports on the analysis of the near-UV and visible absorption spectrum of water vapor. Obtained by combining a high- resolution FT spectrometer and a long path White multiple- reflection cell. A large number of water vapor lines belonging to highly excited vibrational levels are identified. Most of these lines have not been observed previously and do not appear in the spectroscopic databases HITRAN and GEISA, widely used for atmospheric calculations. All identified lines are fitted with a Voigt profile using the WSPECTRA program and their cross section and self-broadening parameters at 291 K are determined. A particular attention is give to the integrate cross section over the total spectral range investigated, in order to estimate the contribution of the weak UV-visible water vapor absorption lines to the earth's radiation balance. Preliminary measurements of cross section in the 20000-16000 cm-1 spectral range are also presented.
Concentration measurements of trace gases in the atmosphere require the use of highly sensitive and precise techniques. The Fourier transform absorption spectroscopy technique is one of them heavily used for atmospheric measurements. It is currently used in all spectral regions, from the far-IR to the near-UV. The spectra recorded will be either fully resolved, showing well separated lines, or will only show absorption bands consisting of a great number of overlapping lines. The algorithm used to retrieve the concentrations of the atmospherically important molecules depends on this, as well as the resolution used to record the spectra. In both cases however, the instrumental function will modify the spectrum shape and has to be taken into account. In order to retrieve concentrations with the best possible accuracy, a thorough understanding of how the instrumental function affects line profiles or absorbances is essential.
In order to better understand the chemistry and the transport mechanisms in the lower troposphere, a new original technique has been developed and tested. The experiment consists in recording high resolution infrared solar absorption spectra containing signatures of important atmospheric constituents, simultaneously from the International Scientific Station of the Jungfraujoch in Switzerland [ISSJ, 3580 m a.s.l., 46.5 degrees N, 8 degrees E, Bruker 120 HR Fourier transform spectrometer (FTS)] and from a nearby valley (Grindelwald, 1070 m a.s.l., Bruker 120 M FTS). Analysis of individual spectra allows to determine vertical column abundances: differences between measurements at ISSJ and at Grindelwald enable us to retrieve the constituents' concentrations between 1070 m and 3580 m, assuming a constant volume mixing ratio in this layer. A first measurement campaign has been organized during the months of May and June 1998. After an initial period of instrument intercomparison at ISSJ, the mobile instrument was moved down in the valley and installed for one month in Grindelwald. When operated side by side at the Jungfraujoch, measurements made by both instruments showed a very good agreement (maximum bias of 1.5%). Analysis of spectra recorded synchronously at the Jungfraujoch and at Grindelwald gave average boundary layer concentrations for a selected set of tropospheric molecules, i.e. methane, nitrous oxide, carbon monoxide and ethane. Comparison with other results and with carbon monoxide in-situ measurements made at ISSJ showed a good agreement.
Concentration measurements of trace gases in the atmosphere require the use of highly sensitive and precise techniques. The UV-visible DOAS technique is one of them heavily used for tropospheric measurements. In order to assess the advantages and drawbacks of using a Fourier transform spectrometer, we have built a DOAS optical setup based on a Bruker IFS 120M spectrometer. The characteristics and capabilities of this setup have been studied and compared to those of the more conventional grating based instruments during several intercomparison campaigns. The main advantages of the FTS are: (i) the existence of a reproducible and precise wavenumber scale, which greatly simplifies the algorithms used to analyze the atmospheric spectra; and (ii) the possibility to record large spectral regions at relatively high resolution, enabling the simultaneous detection of numerous chemical species with better discriminating properties. The main drawback, on the other hand, is due to the fact that an FTS records high frequency signals and does not have the signal integration capabilities of the CCD based grating spectrographs. The FTS therefore needs fairly large amounts of light and is limited to short to medium absorption pathlengths.
NO2 absorption cross-sections have been obtained at 220 K and 294 K at a resolution of 2 cm-2 from a series of spectra recorded with pure NO2 at pressures from 0.007 to 2 torr. The N2O4 absorption cross-section has been obtained at 220 K. The uncertainty in the NO2 cross-sections is estimated to be less than 3 percent for the spectral region below 40000 cm-1 at 294 K, 3 percent below 30000 cm-1 at 220 K. Temperature and pressure effects have been observed. Comparison with literature data show good agreement between 37500 and 20000 cm-1. O2 absorption spectra have been recorded at high and low resolution in the UV. A re-analysis of the three Herzberg band systems has been performed, extending the rotational assignment up to N equals 31 and identifying several new vibrational bands. Determination of the integrated intensities and oscillator strengths of the bands has been performed. These values have been compared with data of the literature. Spectra recorded with increasing pressures of N2 or Ar show that the structure d continuum overlapping the Herzberg bands cannot be explained by O2-O2 absorption, but comes from collision induced absorption. Low resolution spectra have also been recorded in the visible region in order to determine the (O2)2 absorption cross-section.
Concentrations of SO2, NO2, H2CO, and O3 have been measured regularly since October 1990 at the urban site of the Campus of the Universite Libre de Bruxelles, using the differential optical absorption spectroscopy (DOAS) technique associated with a Fourier Transform Spectrometer. The experimental set up has already been described elsewhere (Vandaele et al., 1992). It consists of a source (either a high pressure xenon lamp or a tungsten filament) and an 800 m long path system. The spectra are recorded in the 26,000 - 38,000 cm-1 and 14,000 - 30,000 cm-1 spectral regions, at the dispersion of 7.7 cm-1. The analytical method of the DOAS technique is based on the fact that in atmospheric measurements, it is impossible to obtain an experimental blank spectrum. Therefore, the Beer-Lambert law has to be rewritten as: I equals I'oen(Delta (sigma) d) where I is the measured intensity, Io the measured intensity from which all absorption structures have been removed, n the concentration, d the optical path length, and (Delta) (sigma) the differential absorption cross section of the molecule. Numerous methods for determining I'o exist. Fourier transform filtering has been used in this work. This method defines I'o as the inverse Fourier transform of the lower frequencies portion of the power spectrum of the experimental data. A least squares procedure is then applied in order to determine the concentration of the desired molecules.