Tropospheric ozone (O₃) is a powerful greenhouse gas and air pollutant that negatively impacts human health and ecosystems. Routine observations to better understand its vertical distribution are rare. We derive vertical profiles of tropospheric O₃ from TROPOMI total columns using cloud-slicing designed and optimised with synthetic experiments. Cloud-slicing produces multiyear (June 2018 to May 2022) seasonal means of O₃ mixing ratios at 1° resolution for five discrete layers throughout the troposphere: two upper (320-180 hPa, 450-320 hPa), two middle (600-450 hPa, 800-600 hPa) and one in the boundary layer (below 800 hPa). This vertical resolution is superior to existing satellite-derived datasets that offer at most two independent vertical layers. Cloud-sliced data coverage is near-global in the middle layers and limited mostly to the tropics in the top and remote marine boundary layers. In the southern hemisphere and tropics, cloud-sliced O₃ are typically within 15% of ozonesondes from NOAA, SHADOZ and WOUDC networks, but exhibit negative biases of up to 35% in the northern hemisphere free troposphere. We use our vertically-resolved cloud-sliced O₃ to assess the state of knowledge as simulated with the GEOS-Chem model. Our comparison provides support for a persistent model overestimate in upper tropospheric O₃ year-round and in tropical free tropospheric O₃ in March-May and September-November, as well as a model underestimate in the southern high latitudes. Further cloud-sliced O₃ development is needed to resolve a northern hemisphere free troposphere bias to exploit the potential to examine hourly variability in vertically resolved O₃ from geostationary sensors.
Horner et al. (Thu,) studied this question.