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

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.