SAOZ UV-Vis remote measurements

Long series of ozone profiles from 6 to 28 km at the Southern Tropics have been obtained from solar occultation measurements at twilight using a SAOZ spectrometer borne by long duration balloons. Two flights have been performed from Bauru in Brazil both in the summer, in February and March 2001 and 2003, from where the balloons 5 are moving westward at almost constant latitude (20 • ±5 • S). The flight in 2001 passed right over Reunion Island where the SAOZ measurements could be compared to those of both tropospheric and stratospheric ozone lidars. In the stratosphere, compared to that of SAOZ, the measurements of the SHADOZ ozonesondes network, and the HALOE and SAGE II instruments in orbit are found a little noisier (2–3% for SAGE II, 3– 10 4% for HALOE, 4–5% for the sondes), and of insignificant or small high biased (SAGE II). No differences in altitude could be found between SAOZ and SAGE II at all levels, and HALOE above 22 km. But the ozonesondes appear to be systematically displaced upwards by some 400–700 m at all levels, and the HALOE profiles do show an increasing altitude bias at decreasing altitude below 22 km. In the upper troposphere, the 15 SAOZ measurements are found consistent with those of the sondes at Reunion Island and high biased on average over the Western Pacific, at American Samoa and Fiji. Compared to SAOZ, SAGE II shows a 50–60% low bias similar to that already found with the ozonesondes, and a 2–2.5 times larger zonal variability, suggesting a degradation of its precision below the tropopause. Finally, the unrealistic large offsets and 20 variability in the HALOE data compared to all others suggest that its measurements are no reliable in the tropical troposphere below 17 km.