Abstract: We present validation between total ozone from satellite and ground-based observations of the Dobson and Brewer spectrometers and ozone radiosonde at Zhongshan and Syowa Antarctic research stations, for September 2004 to March 2009. Results show that mean bias error between Zhongshan (Syowa) and Ozone Monitor Instrument Total Ozone Mapping Spectrometer (OMI-TOMS) data are −0.06%±3.32% (−0.44%±2.41%); between it and OMI Multi Axis Differential Optical Absorption Spectroscopy (OMI-DOAS) data, the error is −0.34%±4.99% (−0.22%±4.85%). Mean absolute bias error values of OMI-TOMS data are less than those of OMI-DOAS. This means that total ozone of OMI-TOMS is closer to ground-based observation than that of OMI-DOAS. Comparison between direct observational total ozone of ground-based and integrated ozone from the ozone profile measured by ozone radiosonde shows that ozone amount calculated with the Solar Backscatter Ultraviolet (SBUV) method above balloon burst height is similar to corresponding Microwave Limb Sounder (MLS) data. Therefore, MLS data can be substituted with SBUV data to estimate ozone amount above that level. Mean bias error of the MLS ozone column is 2% compared with the ozonesonde column, with standard deviation within 9.5%. Comparison of different layers from ozone profiler and MLS data indicates that at the 215 hPa layer, the MLS ozone value is high, with relative deviation more than 20%. At the 100 hPa and 68 hPa layers, the MLS ozone value is also high. This deviation is mainly in spring, during Antarctic ozone hole appearance. In this period, at the height of severe ozone loss, relative deviation of MLS ozone values is especially large.

Key words: validation, Antarctic ozone, satellite, ground-based, deviation