Why Heat Makes Ozone Worse in Cities but Not the Countryside

When a heatwave settles over eastern China, the air does not simply feel hotter — it becomes chemically more dangerous. Surface ozone, the corrosive pollutant responsible for summer smog, climbs steadily as temperatures rise, and a new study of the Yangtze River Delta reveals that the reasons behind that climb are strikingly different in a megacity compared with the forested countryside nearby. The research, published in Atmospheric Chemistry and Physics, combined a full summer of ground-level measurements with a suite of controlled chemistry-model experiments to disentangle exactly which processes turn hot weather into bad air, and why the answer changes depending on where you stand.

The team, led by Chang Su of the University of Edinburgh together with colleagues at Reading and Edinburgh, focused on the summer of 2022, when the Yangtze River Delta endured severe and persistent extreme heat. They compared urban Shanghai, a densely built megacity with heavy traffic and industry, against Lishui, a region in southwestern Zhejiang where more than 80 percent of the land is forest and nitrogen oxide levels sit far below the urban average. Across 250 air-quality monitoring stations in the delta, the selected Shanghai sites ranked near the top of the regional nitrogen dioxide distribution, while the Lishui sites sat near the bottom — a natural experiment in how the same regional heat plays out in two contrasting chemical worlds.

The observational headline is unambiguous: daily maximum ozone rises with daily maximum temperature at both locations, but roughly twice as fast in the city. Averaged across eight Shanghai monitoring sites, ozone increased by 7.27 micrograms per cubic metre for every additional degree Celsius, whereas the three Lishui sites recorded a sensitivity of 3.36 micrograms per cubic metre per degree. Both relationships were statistically robust across the whole summer. During heatwave days the picture grew more complicated — the rural slope actually steepened slightly, while in Shanghai the heatwave-day relationship became statistically uncertain, hinting that at the most extreme temperatures additional factors such as stagnation, humidity swings and shifting precursor chemistry begin to blur the simple temperature-ozone link.

Humidity emerged as a crucial co-conspirator. Mapping ozone across a two-dimensional grid of temperature and relative humidity showed that the pollutant favours hot, dry air: concentrations rise with temperature and fall with humidity at both sites. Relative humidity drops by roughly 2.4 to 3.2 percentage points for every degree of warming in the delta, meaning that part of what looks like a pure temperature effect is actually mediated by drying. This matters chemically because water vapour provides a major destruction pathway for ozone — when energetic oxygen atoms produced by ozone photolysis collide with water molecules, the ozone is effectively dismantled. Drier air weakens that sink, letting ozone accumulate.

To move beyond correlation, the researchers deployed the UKCA Box Model, a simplified single-parcel version of the chemistry scheme used in the UK Earth System Model. The box treats each location as a well-mixed volume of air with prescribed temperature, humidity and emissions, deliberately excluding transport, clouds and boundary-layer dynamics. That limitation is also its strength: any ozone change produced inside the box must come from local chemistry or temperature-sensitive emissions, allowing the team to run factorial experiments that switch individual processes on and off across a temperature range from 25 to 40 degrees Celsius. Sixteen configurations tested combinations of diurnal temperature cycles, temperature-dependent isoprene emissions and two humidity scenarios.

The results split cleanly along chemical-regime lines. In Shanghai, which the model confirmed sits firmly in a VOC-limited regime — where ozone responds strongly to volatile organic compounds but can even be suppressed by extra nitrogen oxides — temperature-dependent isoprene emissions explained about 69 percent of the observed ozone increase between 30 and 40 degrees Celsius. Isoprene, the five-carbon molecule that trees and plants exhale in prodigious quantities, is emitted exponentially faster as leaf temperatures climb. In the urban box, that surge of biogenic reactivity feeds directly into ozone production, lifting simulated daily maxima by 20 to 40 micrograms per cubic metre in the mid-thirties. The diurnal temperature cycle added a further 14 percent, while pure thermal chemistry contributed under 20 percent.

Lishui told the opposite story. There, in a nitrogen-oxide-limited regime, the temperature response of ozone was governed almost entirely by radical-nitrogen oxide chemistry: faster reaction rates and stronger sunlight accounted for roughly 80 percent of the 30-to-40-degree increase, while isoprene’s temperature dependence chipped in only about 10 percent. Extra biogenic reactivity, in other words, has little to burn in a rural atmosphere starved of nitrogen oxides. Budget diagnostics showed the HO2-plus-NO reaction as the dominant production pathway at both sites, but in Lishui its role became shared among a broader family of peroxy-radical reactions as temperatures rose, and the hydroxyl-radical sink grew large enough to divert much of the radical pool away from ozone formation altogether.

The team also traced the fate of peroxyacyl nitrate, or PAN, a reservoir species that parks reactive nitrogen until heat cracks it open again. Thermal decomposition of PAN accelerated monotonically with temperature at both sites, peaking in the early afternoon. But the concentration response diverged: in Lishui, PAN levels fell steadily as warming outpaced its slow production, while in Shanghai concentrations actually rose from 25 to 35 degrees before collapsing at 40 degrees — evidence that the urban atmosphere can sustain PAN production until a thermal threshold is crossed, beyond which the reservoir dumps its nitrogen back into the ozone-forming pool. This PAN-mediated nitrogen recycling is a key reason hot afternoons in polluted cities become so productive for ozone.

Humidity experiments sharpened the policy implications further. Holding relative humidity fixed at 70 percent while raising temperature — a wetter atmosphere than reality — cut simulated peak ozone by about 22 micrograms per cubic metre at 40 degrees in Shanghai, because the extra water vapour strengthened the ozone-destroying photolysis pathway by roughly 42 percent. In Lishui, the humidity perturbation reduced ozone through both suppressed HO2-plus-NO production and enhanced destruction. The lesson is that temperature and humidity must be treated jointly: analyses that ignore drying over-attribute the warming penalty to reaction kinetics and understate the role of moisture in cleaning the air.

For policymakers, the regime-specific findings point toward heat-resilient control strategies. In Shanghai-like VOC-limited cities, curbing the ozone climate penalty demands simultaneous nitrogen oxide reductions, targeted management of volatile organic compounds — including temperature-sensitive evaporative sources such as solvents and gasoline, which the authors estimate could rise by more than 50 percent between 30 and 40 degrees — and careful selection of low-isoprene-emitting tree species for urban greening. In rural, nitrogen-oxide-limited regions like Lishui, nitrogen oxide cuts remain the most reliable lever, with VOC controls offering only marginal benefit. Although the quantitative sensitivities are specific to the exceptional summer of 2022 and the box model deliberately excludes transport and meteorological feedbacks, the underlying mechanisms — biogenic amplification in cities, radical-nitrogen chemistry in the countryside, and humidity’s damping hand in both — echo findings from other polluted mid-latitude regions, suggesting the diagnosis will travel well as the planet keeps warming.

Subject of Research: Temperature sensitivity of summertime surface ozone and its urban versus rural chemical mechanisms in the Yangtze River Delta, China

Article Title: Summertime ozone sensitivity to temperature in China: observational evidence and mechanistic attribution in urban and rural areas in the Yangtze River Delta region

Article References: Su, C., Bollasina, M., Weber, J., & Stevenson, D. (2026). Summertime ozone sensitivity to temperature in China: observational evidence and mechanistic attribution in urban and rural areas in the Yangtze River Delta region. Atmospheric Chemistry and Physics, 26(19), 14439-14469. https://doi.org/10.5194/acp-26-14439-2026

Image Credits: AI Generated

DOI: 10.5194/acp-26-14439-2026

Keywords: surface ozone, heatwaves, Yangtze River Delta, isoprene emissions, VOC-limited regime, NOx-limited regime, PAN thermal decomposition, relative humidity, UKCA box model, air quality, climate penalty, Shanghai