Emerging Divergence of Observed and Modeled Ozone in the Tropical Stratosphere and Troposphere

Total column ozone (TCO) is essential for protection of the biosphere from harmful solar ultraviolet radiation. TCO is the most frequently used metric to estimate recovery times (WMO, 2022). However, TCO sums stratospheric and tropospheric contributions; these have largely separate sources and sinks, and such independence complicates recovery estimates, as ozone depleting substances decline due to implementation of the Montreal Protocol. Here, we provide observational evidence indicating that separating tropospheric and stratospheric ozone is critical to determining the effectiveness of the Montreal Protocol even at low latitudes, in agreement with previous modeling work. While observed changes in tropical TCO fall within the spread of changes simulated by models participating to CCMI-1, we show this is fortuitous only: observed tropospheric ozone increases exceed models since the year 2000. We also revisit modeled and observed ozone trends in the tropical upper stratosphere, a region where the expected recovery due to declining halogenated ODSs is detectable since the turn of the century. We find that the timing of ozone recovery onset is uncertain, with a general tendency of earlier times in models than in satellite data. As a result, recovery estimates are sensitive to the turnaround date chosen for linear trend analyses. Due to this sensitivity, we find that the agreement between observations and models in the upper stratospheric ozone recovery rates is less obvious than in previous studies. Lastly, we show that models displaying tropical stratospheric column ozone changes that are more consistent with observations also project the most negative stratospheric column ozone changes by the end-of-century (up to -8%), although this is not well constrained. These results call for more rigorous statistical analyses of stratospheric ozone trends in models and observations. At the same time, reasons for disagreement between CCMs and observations in certain regions need to be identified, in order to build confidence in future projections of the ozone layer.