Four years of ground-based total ozone measurements by visible spectrometry in Antarctica

Visible spectrometers SAOZ have been developed at Service d'Aeronomie for permanent ground-based ozone monitoring at all latitudes up to the polar circle in winter. Observations are made by looking at the sunlight scattered at zenith in the visible range, twice a day, at sunrise and sunset. Compared to ozone observations in the UV generally in use, visible observations in the small Chappuis bands at twilight have the advantages of being independent of stratospheric temperature, little contaminated by tropospheric ozone and multiple scattering, and of permitting observations even in winter at the polar circle. SAOZ instruments have been installed since 1988 at several stations in the Antarctic and the Arctic. More than four years data at Dumont d'Urville in Terre Adelie (67 deg S) are now available. The station is generally located at the edge of the vortex in spring and therefore the ozone hole is seen there only occasionally. The lowest values (140 DU) were reported in early October 1991. According to these first regular observations throughout the whole winter ozone seems to increase in late autumn and winter. Its decay does not start before the end of August. Although of smaller amplitude than with the previous version five data, the ratio between the groundbased and satellite/TOMS measurements displays a systematic seasonal variation correlated partly to the sun zenith angle of observations from orbit and partly to the temperature of the stratosphere. Since ground-based measurements are always made at 90 deg SZA, the SZA dependence must come from the satellite data interpretation (TOMS observations are between 43 to 88 deg SZA). The temperature dependence could be partly due to variations of ozone absorption cross-sections in the ultraviolet used by the satellite spectrometer, and partly to a systematic seasonal cycle of the air mass factor use in the interpretation of the ground based observations. However, the last contribution appears to be too small to compensate the ozone increase in winter reported by SAOZ, which is then real.