Validation of AIRS v4 ozone profiles in the UTLS using ozonesondes from Lauder, NZ and Boulder, USA

Ozonesonde observations from Lauder (45.0 degrees S, 169.7 degrees E) and Boulder (39.9 degrees N, 105.3 degrees W) are used to examine the quality of the Atmospheric Infrared Sounder (AIRS) v4 vertical ozone profile product in the upper troposphere lower stratosphere (UTLS). At lower altitudes (similar to 700-200 hPa pressure range), AIRS ozone mixing ratios are larger than ozonesonde measurements, and at higher altitudes (similar to 100-30 hPa pressure range), AIRS ozone mixing ratios are smaller. Compared to the ozonesondes, AIRS retrieval results at Lauder have a median bias of 80% in the region 700-200 hPa, and 0 to -20% in the region 100-30 hPa. For Boulder these values are 40% and 0 to 5%, respectively. Using a tropopause adjusted vertical coordinate system, Lauder has median biases of +90 to +120% in the troposphere and 0 to +25% in the stratosphere whereas Boulder shows median biases of +45 to +70% in the troposphere and 0 to +35% in the stratosphere. Despite the bias, AIRS retrieval in the UTLS region shows a statistically significant positive correlation with the ozonesonde data, indicating that while the absolute values have a large uncertainty, the retrieval captures the variability of ozone in the UTLS region. Hence AIRS ozone is suitable for studies where the change in ozone is important rather than the absolute ozone mixing ratio. Examinations of the training data set show that the retrieval biases are likely influenced by the deficiency of the training data to represent ozone distribution during the regression step of the retrieval. Furthermore the physical retrieval adds little additional information to the final result.