New spectroscopic data for Tropomi/S5P in the 2.3 mu region for CH4 and H2O and a new UV database for ozone

The ESA project „SEOM-Improved Atmospheric Spectroscopy Databases (IAS)“ will improve the spectroscopic database for retrieval of the data products CO, CH4, O3 and SO2 column amounts measured by the TROPOMI instrument (TROPOspheric Monitoring Instrument) aboard the Sentinel-5 Precursor. The project was launched in February 2014 with 3 years duration extended to 2018. New spectroscopic data have been obtained for CO, CH4 and O3 in the 2.3 µm region. Measurements were mainly taken with a high resolution Fourier Transform spectrometer combined with a coolable multi reflection cell. Cavity ring down measurements served for validation. While water analysis with multispectrum fitting turned out to be straight forward the analysis of methane turned out to be very complicated and time consuming. The input database HITRAN2012 was found to be not suitable as initial guess and a theoretical database provided by V. Tyuterev was used instead. Still about 2900 misplaced/new lines (1300 lines removed) were found which required pure methane measurements at low and ambient temperature in order to fit the lower state energy of the unassigned lines. In order to fit the spectrum down to the noise level speed dependence, Rosenkranz line mixing and temperature dependence of line mixing had to be considered. The data quality of the new database was validated by solar occultation measurements where residuals showed solar lines only, but no spectroscopic data base error. Furthermore, atmospheric CH4 profiles in agreement with models could be obtained using the new spectroscopic data, for the first time. The significant impact of the new spectroscopic data with respect to HITRAN2012 for CH4 and CO columns for TROPOMI observations was derived by retrieval simulations. New UV measurements for a new temperature dependent absorption cross section database have been carried out. Large effort has been taken to ensure a good baseline and to obtain highly accurate number densities in the cell. The new database shows good agreement (better than 1%) with other databases for the 253 nm mercury line, but 1-2% lower values above 330 nm in the region relevant for atmospheric remote sensing. Together with our new MIR results (see presentation “New absolute and relative line intensities of ozone fundamentals - a step towards the end of the ozone UV/MIR dilemma”) these results also could close the 4% gap between UV and infrared atmospheric observations.