2008-07-01: Assuring quality for long-term ozone trend studies

The record of total ozone derived from the TOMS instrument series since November 1978 and continued by the Ozone Monitoring Instrument (OMI) since July 2004, is the longest to date and essential to study the recovery of the ozone layer. To verify the quality of the OMI-TOMS ozone data and its potential for trend analysis, Kroon et al. [2008] compare two ozone retrieval methods applied for the OMI instrument.
They find that ozone retrieved with the traditional, 4-wavelength TOMS version 8 algorithm and with the Differential Optical Absorption Spectroscopy (DOAS) method generally agrees within 9 DU (better than 3%). The largest differences occur over the polar regions, as illustrated below.
Total ozone columns retrieved with the OMI TOMS algorithm (Left), OMI DOAS (Middle), and their difference (Right) over the Northern Hemisphere on 21 March 2006. At a 70o solar zenith angle, OMI TOMS shows a 40 DU discontinuity that is absent in OMI DOAS. At this angle, the TOMS version 8 algorithm suffers from errors caused by stray-light in the 313 nm channel, whereas the DOAS algorithm is much less sensitive to such errors owing to the use of the complete 331-335 nm spectrum.