Characterizing Global Ozonesonde Profile Variability From Surface to the UT/LS With a Clustering Technique and MERRA-2 Reanalysis

Our previous studies employing the self-organizing map (SOM) clustering technique to ozonesonde data have found significant links among meteorological and chemical regimes, and the shape of the ozone (O3) profile from the troposphere to the lower stratosphere. Those studies, which focused on specific northern hemisphere midlatitude geographical regions, demonstrated the advantages of SOM clustering by quantifying O3 profile variability and the O3/meteorological correspondence. We expand SOM to a global set of ozonesonde profiles spanning 1980 to present from 30 sites to summarize the connections among O3 profiles, meteorology, and chemistry, using the Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2) reanalysis and other ancillary data. Four clusters of O3 mixing ratio profiles from the surface to the upper troposphere/lower stratosphere (UT/LS) are generated for each site, which show dominant profile shapes and typical seasonality (or lack thereof) that generally correspond to latitude (i.e., tropical, subtropical, midlatitude, and polar). Examination of MERRA-2 output reveals a clear relationship among SOM clusters and covarying meteorological fields (geopotential height, potential vorticity, and tropopause height) for polar and midlatitude sites. However, these relationships break down within ±30° latitude. Carbon monoxide satellite data, along with velocity potential, a proxy for convection, calculated from MERRA-2 wind fields assist characterization of the tropical and subtropical sites, where biomass burning and convective transport linked to the Madden-Julian oscillation (MJO) dominate O3 variability. In addition to geophysical characterization of O3 profile variability, these results can be used to evaluate chemical transport model output and satellite measurements of O3 profiles.