An Examination of the Recent Stability of Ozonesonde Global Network Data

The recent Assessment of Standard Operating Procedures for Ozonesondes 2.0 (WMO/GAW Report #268) addressed questions of homogeneity and long-term stability in global electrochemical concentration cell (ECC) ozone sounding network time series. Among its recommendations was adoption of a standard for evaluating data quality in ozonesonde time series. Total column ozone (TCO) derived from the sondes compared to TCO from Aura's Ozone Monitoring Instrument (OMI) is a primary quality indicator. Comparisons of sonde ozone with Aura's Microwave Limb Sounder (MLS) are used to assess the stability of stratospheric ozone. This paper provides a comprehensive examination of global ozonesonde network data stability and accuracy since 2004 in light of the sudden post-2013 TCO "dropoff" of similar to 3%-4% that was reported previously at select stations (Stauffer et al., 2020, ). Comparisons with Aura OMI TCO averaged across the network of 60 stations are stable within about +/- 2% over the past 18 years. Sonde TCO has similar stability compared to three other TCO satellite instruments, and the stratospheric ozone measurements average to within +/- 5% of MLS from 50 to 10 hPa. Thus, sonde data are reliable for trends, but with a caveat applied for a subset of dropoff stations in the tropics and subtropics. The dropoff is associated with only one of two major ECC instrument types. A detailed examination of ECC serial numbers pinpoints the timing of the dropoff. However, we find that overall, ozonesonde data are stable and accurate compared to independent measurements over the past two decades. Plain Language Summary Ozonesondes provide accurate ozone measurements from the surface to similar to 30-km altitude and are used as a reference for studies of satellite data, trends, pollution, and climate. Updated guidelines for sonde preparation and adoption of sonde total column ozone (TCO) comparisons with satellite TCO as a "data quality" reference were published in 2021 by the Assessment of Standard Operating Procedures for Ozonesondes (ASOPOS) 2.0 panel in WMO/GAW Report 268. We report the first application of the ASOPOS 2.0 protocol to TCO evaluation from the 60-station global ozonesonde network (42,042 profiles total). With Aura OMI TCO as the satellite reference (October 2004 to mid-2021), we find that TCO readings from the global ozonesonde network are remarkably stable, consistently within +/- 2% of the satellite. An exception occurs at only a small subset of tropical and subtropical locations that use one type of ozonesonde instrument. The latter result confirms our earlier report that a sudden TCO drop occurs at selected sites after 2013. The timing and magnitude of the dropoff are revisited. The hypothesis that ozonesonde production changes are a contributor remains, with station-specific factors possibly affecting the magnitude of the bias. Overall, global ozonesonde network data are of high quality and stability.