The discrepancy between stratospheric ozone profiles from balloon soundings and from other techniques: A possible explanation

Regular balloon ozone soundings with electrochemical sondes have been performed at Uccle since 1969. More than 450 ozone soundings between 1985 and 1989 were used to calculate the altitudes Zs from the VIZ radiosonde data and the altitudes Zr deduced from the tracking of the balloon train with a primary wind-finding radar. The values of Zs at fixed times appeared to be systematically too low as compared to Zr. The differences Zr-Zs increase with altitude; at 30 km the annual mean values of Zr-Zs (plus or minus standard deviation) vary between 590 plus or minus 910 m and 1410 plus or minus 1160 m, according to the pressure calibration of different manufacturing series of radiosondes. From these results it is found that around the 30 km level the ozone concentrations calculated from soundings with VIZ sondes are too low by 7.5 to 14 percent, depending upon the manufacturing series of radiosondes. At least part of the discrepancy which has often been found between ozone profiles from balloon soundings and from other techniques such as rocket observations or Umkehr measurements may be explained by this effect. An altitude correction would have important consequences as to the climatology of ozone in the middle stratosphere as adopted at the moment. About half of the day-to-day variability of ozone observed from soundings with VIZ radiosondes above the 30 km level, is induced by the variability of Zr-Zs. The agreement between altitudes calculated from radar data and Vaisala radiosondes is much better; from 34 comparative soundings a mean difference (plus or minus standard deviation) of about -300 plus/minus 180 m was found at 30 km.