Home > Research > Publications & Outputs > Global ozone and air quality : a multi-model as...
View graph of relations

Global ozone and air quality : a multi-model assessment of risks to human health and crops.

Research output: Contribution to journalJournal article


  • K. Ellingsen
  • M. Gauss
  • R. Van Dingenen
  • F. J. Dentener
  • L. Emberson
  • Arlene M. Fiore
  • M. G. Schultz
  • D. S. Stevenson
  • M. R. Ashmore
  • C. S. Atherton
  • D. J. Bergmann
  • I. Bey
  • T. Butler
  • J. Drevet
  • H. Eskes
  • D. A. Hauglustaine
  • I. S. A. Isaksen
  • L. W. Horowitz
  • M. Krol
  • J. F. Lamarque
  • M .G. Lawrence
  • T. Van Noije
  • J. Pyle
  • S. Rast
  • J. Rodriguez
  • N. Savage
  • S. Strahan
  • K. Sudo
  • S. Szopa
<mark>Journal publication date</mark>6/02/2008
<mark>Journal</mark>Atmospheric Chemistry and Physics Discussions
Number of pages61
Pages (from-to)2163-2223
<mark>Original language</mark>English


Within ACCENT, a European Network of Excellence, eighteen atmospheric models from the U.S., Europe, and Japan calculated present (2000) and future (2030) concentrations of ozone at the Earth's surface with hourly temporal resolution. Comparison of model results with surface ozone measurements in 14 world regions indicates that levels and seasonality of surface ozone in North America and Europe are characterized well by global models, with annual average biases typically within 5–10 nmol/mol. However, comparison with rather sparse observations over some regions suggest that most models overestimate annual ozone by 15–20 nmol/mol in some locations. Two scenarios from the International Institute for Applied Systems Analysis (IIASA) and one from the Intergovernmental Panel on Climate Change Special Report on Emissions Scenarios (IPCC SRES) have been implemented in the models. This study focuses on changes in near-surface ozone and their effects on human health and vegetation. Different indices and air quality standards are used to characterise air quality. We show that often the calculated changes in the different indices are closely inter-related. Indices using lower thresholds are more consistent between the models, and are recommended for global model analysis. Our analysis indicates that currently about two-thirds of the regions considered do not meet health air quality standards, whereas only 2–4 regions remain below the threshold. Calculated air quality exceedances show moderate deterioration by 2030 if current emissions legislation is followed and slight improvements if current emissions reduction technology is used optimally. For the "business as usual" scenario severe air quality problems are predicted. We show that model simulations of air quality indices are particularly sensitive to how well ozone is represented, and improved accuracy is needed for future projections. Additional measurements are needed to allow a more quantitative assessment of the risks to human health and vegetation from changing levels of surface ozone.