New absolute and relative line intensities of ozone fundamentals - a step towards the end of the ozone UV/MIR dilemma

Abstract

New ozone mid infrared measurements in the frame of an ESA project (SEOM-IAS) have been carried out at DLR and Reims. These measurements were combined with previous measurements taken in 1998 and published in 2002 (Wagner G., Birk M., Schreier F., Flaud J.-M., Spectroscopic Database for ozone in the fundamental spectral regions, J. Geophys. Res. 107 (D22), 4626, doi: 10.1029/2001JD000818 (2002)). The analysis was performed with a multispectrum fitting approach. It turned out that the treatment of the instrumental lineshape has a large influence on the line intensities in case of nearly Doppler-broadened measurements. Furthermore, it was found that without considering speed dependence the line intensities for measurements even in the 1 mbar range were too small by about 1%. Self speed dependence was expressed in terms of a polynomial versus the Lorentzian self broadening. Alternatively, the approach by H. Tran et al. (H. Tran et al., Non-Voigt line-shape effects on retrievals of atmospheric ozone: Collisionally isolated lines, Journal of Quantitative Spectroscopy & Radiative Transfer 111 (2010) 2012–2020, doi:10.1016/j.jqsrt.2010.04.002) was used to calculate the speed dependence from the temperature exponent of the Lorentzian self broadening which was available since one of the measurements was recorded at 233 K. The absolute and relative line intensities were found to be significantly different from HITRAN, where the relative line intensities have remained unchanged for 26 years. The new line intensities in the nu3 region, which is most relevant for atmospheric applications, were larger by 2% for weak and up to 4% for strong lines. It should be noted that atmospheric ozone retrievals pointed out a discrepancy of about 4% when ozone was retrieved from infrared and UV measurements. An increase of MIR ozone line intensities would partly close this gap. A new quantum mechanical treatment of the new data will also be presented. The results given are based on DLR measurements alone. A small spectrum excerpt from a Reims measurement was used for validation.