Changes in atmospheric composition discerned from long-term NDACC measurements: trends in direct greenhouse gases derived from infrared solar absorption spectra recorded at the Jungfraujoch station

摘要

The University of Liège (ULg) is operating -under clear sky conditions- two state-of-the-art Fourier Transform Infrared (FTIR) spectrometers at the high-altitude research station of the Jungfraujoch (Swiss Alps, 46.5ºN, 3580m asl), within the framework of the Network for the Detection of Atmospheric Composition Changes (NDACC). Routine FTIR operation started in 1984. Since then, it has been continued without disruption, allowing collecting more than 45000 high-resolution broadband IR solar absorption spectra, between 2 and 16 µm, using either HgCdTe or InSb detectors as well as a suite of optical filters. Typically, the spectral resolutions achieved lie in the 0.003 to 0.009 cm-1 interval while signal-to-noise ratios of 1000 and more are reached. Numerous narrow-band IR spectra essentially recorded from 1976 to 1989 with grating instruments are also available. Their analyses with modern tools have recently started [Bader et al., 2011] and will be pursued to consistently extend our datasets back in the 1970s. Geophysical parameters are deduced from the ULg observational database either with the SFIT-1, SFIT-2 or PROFFIT-9 algorithm, allowing producing total column time series of the target gases. In addition, information on their vertical distributions with altitude can generally be derived when using SFIT-2 or PROFFIT-9 which both implement the Optimal Estimation Method of Rodgers [1990]. Presently, more than two dozen atmospheric species are systematically retrieved from the Jungfraujoch observations, allowing the monitoring of key constituents of the Earth's atmosphere which play important roles in stratospheric ozone depletion and/or in global warming. This communication will focus on the direct and major greenhouse gases available from our database, namely water vapor, CO2, CH4, N2O, tropospheric ozone, CFC-11, CFC-12, HCFC-22, CCl4, SF6, as well as CF4 which has recently been added to our targets list [Duchatelet et al., 2011]. Trends and associated uncertainties characterizing the available -and often multi-decadal- time series have been derived or updated with a statistical bootstrap resampling tool [Gardiner et al., 2008], they will be presented and critically compared with data available from the literature.