Understanding lowermost stratosphere (LMS) ozone variability is an important topic in the trends and climate assessment communities because of feedbacks among changing temperature, dynamics and ozone. LMS evaluations are usually based on satellite observations. Free tropospheric (FT) ozone assessments typically rely on profiles from commercial aircraft. Ozonesonde measurements constitute an independent dataset encompassing both LMS and FT. We used Southern Hemisphere Additional Ozonesondes (SHADOZ) data (5.8°N to 14°S) from 1998-2019 in the Goddard Multiple Linear Regression model to analyze monthly mean FT and LMS ozone changes across five well-distributed tropical sites. Our findings: (1) both FT (5-15 km) and LMS (15-20 km) ozone trends show marked seasonal variability. (2) All stations exhibit FT ozone increases in February-May (up to 15%/decade) when the frequency of convectively-driven waves have changed. (3) After May, monthly ozone changes are both positive and negative, leading to mean trends of +(1-4)%/decade, depending on station. (4) LMS ozone losses reach (4-9)%/decade mid-year, correlating with an increase in TH as derived from SHADOZ radiosonde data. (5) When the upper FT and LMS are defined by tropopause-relative coordinates, the LMS ozone trends all become insignificant. Thus, the 20-year decline in tropical LMS ozone reported in recent satellite-based studies likely signifies a perturbed tropopause rather than chemical depletion. The SHADOZ-derived ozone changes highlight regional and seasonal variability across the tropics and define a new reference for evaluating changes derived from models and satellite products over the 1998 to 2019 period.

An international effort to improve ozonesonde data quality and to reevaluate historical records has made significant improvements in the accuracy of global network data. However, between 2014 and 2016, ozonesonde total column ozone (TCO; O3) at 14 of 37 regularly reporting stations exhibited a sudden drop-off relative to satellite measurements. The ozonesonde TCO drop is 3-7 % compared to satellite and ground-based TCO, and 5-10 % or more compared to satellite stratospheric O3 profiles, compromising the use of recent data for trends, although they remain reliable for other uses. Hardware changes in the ozonesonde instrument are likely a major factor in the O3 drop-off, but no single property of the ozonesonde explains the findings. The bias remains in recent data. Research to understand the drop-off is in progress; this letter is intended as a caution to users of the data. Our findings underscore the importance of regular ozonesonde data evaluation.

The ozonesonde observations in Hanoi, Vietnam, over fourteen years since 2004 have confirmed the enhancement in lower tropospheric ozone concentration at about 3 km altitude in the spring season. We investigated the evolution of the ozone enhancement from analysis of meteorological data, backward trajectories, and model sensitivity experiments. In spring, air masses over Hanoi exhibit strong height dependence. At 3km, the high-ozone air masses originate from the land area to the west of Hanoi, while low-ozone air masses below about 1.5 km are from the oceanic area to the east. Above 4 km, it is traced back to the farther west area. The chemical transport model simulations revealed that precursor emissions from biomass burning in the inland Indochina Peninsula have the largest contribution to the lower tropospheric ozone enhancement, which is transported upward and eastward and overhangs the clean air intrusion from the ocean to the east of Hanoi. At this height level, the polluted air has the horizontal extent of about 20 degrees in longitude and latitude. The polluted air observed in Hanoi is transported further east and widely spread over the northern Pacific Ocean.

Quantifying variability in the lowermost stratosphere (LMS) is important because of feedbacks among changing temperature, dynamics and species like ozone. We used reprocessed Southern Hemisphere Additional Ozonesondes data from 1998-2018 in a Multiple Linear Regression (MLR) model to analyze variability and trends in free tropospheric (FT) and LMS ozoneacross five well-distributed tropical regions. The MLR also computed trends in a proxy for convection as determined from laminae in each ozonesonde-radiosonde pair. Only the equatorial Americas exhibits statistically significant annual trends in FT or LMS ozone. At the other sites, ozonetrends occur in isolated layers during months when convection has changed, February-April or July-November. Our results imply that large FT ozone increases reported for populated tropical areas may be caused by growing pollution overlying smaller changes caused by perturbed dynamics. They also provide regional data for evaluating LMS ozonetrends based on zonal averages of often sparse satellite measurements