Ozone From the Stratosphere Helped Trigger Mexico City’s Worst Smog Episode

In March 2016, monitors across Mexico City recorded ozone concentrations climbing to roughly 210 parts per billion, among the most severe pollution episodes ever documented in the megacity. The usual suspects were present: traffic emissions, intense sunshine, and a stable mountain basin that traps pollutants. But a new study published in the journal Weather and Climate Dynamics argues that a surprising accomplice was at work far above the city, in the stratosphere itself. According to Clemente Lopez-Bravo of the University of New South Wales and his colleagues, a large-scale atmospheric disturbance originating over the North Pacific delivered ozone-rich air from the upper atmosphere down into the lower troposphere over central Mexico, quietly inflating the city’s background ozone levels a full day before the pollution emergency was declared.

The mechanism at the heart of the study is stratosphere–troposphere exchange, or STE, the process by which air crosses the tropopause, the boundary that separates the stratified, ozone-rich stratosphere from the well-mixed troposphere below. Stratospheric ozone constitutes a primary natural source of background tropospheric ozone, and when intrusions of stratospheric air descend deep enough, they can measurably raise ozone concentrations in the lower atmosphere. Most research on this process has focused on the mid-latitudes, where jet streams and storm systems routinely fold the tropopause. What makes the new analysis striking is that it documents an unusually deep intrusion at tropical latitudes, reaching the lower troposphere above a city that sits 2,300 meters above sea level in a mountain-ringed basin.

The story begins with Rossby waves, the vast undulations of the jet stream that encircle the planet. Between 6 and 14 March 2016, an upper-level trough associated with a mid-latitude cyclone amplified over the eastern North Pacific and propagated equatorward toward central Mexico. As the wave grew, the potential vorticity contours that trace the dynamical tropopause began to overturn, a signature of anticyclonic Rossby wave breaking. A filament of high-PV stratospheric air, known as a PV streamer, stretched southward below 20 degrees north latitude. Part of this filament pinched off from the main stratospheric reservoir, forming a cut-off low over Mexico, a closed circulation that kept the disturbance anchored over the region for days.

With the cut-off low in place, the tropopause itself began to fold. The researchers, using the ERA5 reanalysis from the European Centre for Medium-Range Weather Forecasts, tracked the 2 PVU surface, the conventional dynamical marker of the tropopause, as it plunged below 400 hectopascals, well into the middle and lower troposphere. Vertical cross-sections revealed sharp potential vorticity gradients and locally reduced static stability within the intrusion, hallmarks of stratospheric air descending into tropospheric surroundings. The fold persisted for roughly three days, from 9 to 11 March, and was intensified by the subtropical jet, whose winds exceeded 50 meters per second, and by the elevated terrain of central Mexico, which helped steer the descending air toward the surface.

To confirm that air parcels actually crossed the tropopause rather than merely deforming it, the team computed backward trajectories with the Lagrangian Analysis Tool LAGRANTO, initialized over Mexico City between 700 and 100 hectopascals. The results were unambiguous: parcels arriving near 700 hectopascals over the city between 11 and 13 March had originated in the stratosphere above 250 hectopascals, within the fold region. Two distinct descent pathways emerged. One group of parcels plunged rapidly, losing more than 200 hectopascals of pressure in 48 hours, but arrived with diluted potential vorticity, suggesting heavy mixing along the way. A second group descended more gradually, preserving potential vorticity values near 2 PVU aloft and carrying ozone tracer concentrations of roughly 8 times 10 to the minus 8 kilograms per kilogram down to the 750 to 700 hectopascals layer, directly above the city.

Isentropic analysis added a third line of evidence. By interpolating winds, potential vorticity, and ozone onto constant potential-temperature surfaces between 320 and 340 Kelvin, the researchers quantified quasi-horizontal transport along sloping isentropic surfaces, the efficient conveyor that carried ozone-rich air from mid-latitude source regions toward central Mexico. Between 8 and 10 March, ozone-rich air remained confined to the 340 to 360 Kelvin range, above the dynamical tropopause. After 10 March, the signal descended progressively, and by 12 and 13 March air with stratospheric characteristics appeared on the tropospheric side of the tropopause near 315 to 320 Kelvin. The timing matched the surface observations almost perfectly: maximum downward transport occurred approximately one day before the peak in surface ozone mixing ratios.

Satellite observations corroborated the picture from orbit. Total column ozone measurements from the Earth Polychromatic Imaging Camera on the Deep Space Climate Observatory showed an enhancement across Mexico, while tropospheric ozone columns increased over eastern Mexico on 10 and 11 March, coinciding with the development of the fold. Independent measurements from the Ozone Mapping and Profiler Suite on the Suomi National Polar-orbiting Partnership satellite displayed similar large-scale features. The tropospheric signal over central Mexico itself was weaker, which the authors attribute to the high elevation of the terrain and the reduced sensitivity of satellite instruments in the upper troposphere and lower stratosphere.

Crucially, the study does not claim that stratospheric air alone caused the exceedances. The authors are careful to frame the intrusion as a preconditioning mechanism: the descending stratospheric air elevated the background ozone over the region, and local photochemical production from traffic and industry then pushed concentrations over regulatory thresholds. This distinction matters for attribution. In a polluted tropical megacity, stratospheric transport interacts with local precursor emissions, boundary-layer dynamics, and basin circulations, making it extraordinarily difficult to separate the large-scale contribution from local production. The reanalysis-based diagnostics cannot quantify the exact fraction of surface ozone that came from the stratosphere, and the authors acknowledge that ERA5 may smooth out the sharpest small-scale features of folds and mixing layers.

The implications extend well beyond a single event. When the team examined four other major high-ozone episodes in Mexico City, on 25 December 2015, 14 April 2007, 10 April 2019, and 22 February 2024, they found the same dynamical fingerprint: tropopause folds preceding the surface enhancements by a few days, with both shallow and deep intrusions acting as effective pathways for stratospheric ozone. Elevated ozone mixing ratios during the March 2016 event were also detected over Guadalajara, Monterrey, San Luis Potosí, and Torreón, showing that a single breaking Rossby wave can synchronize pollution events across distant metropolitan areas. Cities located in elevated tropical basins appear especially vulnerable, because the altitude brings the lower troposphere closer to the descending stratospheric air, while lower-elevation regions experience weaker surface impacts.

The findings carry a practical message for air-quality forecasting. Rossby wave amplification, PV streamer formation, cut-off low development, and depression of the dynamical tropopause are all large-scale phenomena that numerical weather prediction models can identify days before the associated intrusion arrives. In the March 2016 case, ozone-rich air reached the lower troposphere about a day before the environmental contingency began, suggesting that monitoring potential vorticity on isentropic surfaces and tracking the 1.5 to 2 PVU tropopause could provide genuine early warning of elevated ozone risk during boreal winter and early spring. The authors caution that a single reanalysis-based case cannot establish operational skill, and that multi-event hindcast experiments are needed to quantify achievable lead times and false-alarm rates. But the core insight stands: weather disturbances thousands of kilometers away, in the mid-latitude jet stream, can reach into the tropics and shape the air that millions of city dwellers breathe.

Subject of Research: Stratosphere–troposphere exchange and its influence on surface ozone pollution over tropical Mexico

Article Title: Stratosphere–troposphere exchange and surface ozone pollution over tropical regions: a case study of Rossby wave breaking and tropopause folding

Article References: Lopez-Bravo, C., Caetano, E., & Franco-Díaz, A. (2026). Stratosphere–troposphere exchange and surface ozone pollution over tropical regions: a case study of Rossby wave breaking and tropopause folding. Weather and Climate Dynamics, 7(4), 1995-2013. https://doi.org/10.5194/wcd-7-1995-2026

Image Credits: AI Generated

DOI: 10.5194/wcd-7-1995-2026

Keywords: stratosphere-troposphere exchange, ozone pollution, Mexico City, Rossby wave breaking, tropopause fold, potential vorticity, cut-off low, ERA5 reanalysis, Lagrangian trajectories, isentropic transport, air quality forecasting, subtropical jet