Articles | Volume 10, issue 1
Atmos. Meas. Tech., 10, 231–246, 2017
Atmos. Meas. Tech., 10, 231–246, 2017

Research article 20 Jan 2017

Research article | 20 Jan 2017

Improved GOMOS/Envisat ozone retrievals in the upper troposphere and the lower stratosphere

Viktoria F. Sofieva1, Iolanda Ialongo1, Janne Hakkarainen1, Erkki Kyrölä1, Johanna Tamminen1, Marko Laine1, Daan Hubert2, Alain Hauchecorne3, Francis Dalaudier3, Jean-Loup Bertaux3, Didier Fussen2, Laurent Blanot4, Gilbert Barrot4, and Angelika Dehn5 Viktoria F. Sofieva et al.
  • 1Finnish Meteorological Institute, Helsinki, 00101, Finland
  • 2Royal Belgian Institute for Space Aeronomy (BIRA-IASB), Brussels, Belgium
  • 3Université Versailles St-Quentin, UPMC University Paris 06, CNRS/INSU, LATMOS-IPSL, 78280 Guyancourt, France
  • 4ACRI-ST, Sophia Antipolis, France
  • 5ESA/ESRIN, Frascati, Italy

Abstract. Global Ozone Monitoring by Occultation of Stars (GOMOS) on board Envisat has performed about 440 000 nighttime occultations during 2002–2012. Self-calibrating measurement principle, good vertical resolution, excellent pointing accuracy, and the wide vertical range from the troposphere up to the lower thermosphere make GOMOS profiles interesting for different analyses.

The GOMOS ozone data are of high quality in the stratosphere and the mesosphere, but the current operational retrieval algorithm (IPF v6) is not optimized for retrievals in the upper troposphere–lower stratosphere (UTLS). In particular, validation of GOMOS profiles against ozonesonde data has revealed a substantial positive bias (up to 100 %) in the UTLS region. The retrievals in the UTLS are challenging because of low signal-to-noise ratio and the presence of clouds.

In this work, we discuss the reasons for the systematic uncertainties in the UTLS with the IPF v6 algorithm or its modifications based on simultaneous retrievals of several constituents using the full visible wavelength range. The main reason is high sensitivity of the UTLS retrieval algorithms to an assumed aerosol extinction model.

We have developed a new ozone profile inversion algorithm for GOMOS data (ALGOM2s version 1.0), which is optimized in the UTLS and uses IPF v6 advantages in the middle atmosphere. The ozone retrievals in the whole altitude range from the troposphere to the lower thermosphere are performed in two steps, as in the operational algorithm: spectral inversion followed by the vertical inversion. The spectral inversion is enhanced by using a DOAS-type method at visible wavelengths for the UTLS region. This method uses minimal assumptions about the atmospheric profiles. The vertical inversion is performed as in IPF v6 with the Tikhonov-type regularization according to the target resolution.

The validation of new retrieved ozone profiles with ozonesondes shows a dramatic reduction of GOMOS ozone biases in the UTLS. The new GOMOS ozone profiles are also in a very good agreement with measurements by MIPAS, ACE-FTS, and OSIRIS satellite instruments in the UTLS. It is also shown that the known geophysical phenomena in the UTLS ozone are well reproduced with the new GOMOS data.

Short summary
This paper presents a new ozone profile inversion algorithm for GOMOS/Envisat satellite data. This algorithm is enhanced with a DOAS-type method at visible wavelengths in the upper troposphere and the lower stratosphere. The new GOMOS ozone profiles have a significantly improved data quality in the UTLS compared to the official IPF V6 ozone profiles. The paper describes the inversion algorithm and present inter-comparisons with ozonesonde and satellite measurements.