1. HTAP Multi-model Assessment of Ozone Source-receptor Relationships 3 rd GEOS-Chem Users’ Meeting, Harvard University April 13, 2007 Arlene M. Fiore (arlene.fiore@noaa.gov) Kees Cuvelier, Frank Dentener (JRC) and the TF HTAP Modeling Team including Rokjin Park (GEOS-Chem v7.0 2x2.5) and Marta Garcia Vivanco (GEOS-Chem 4x5) Draft interim report at www.htap.org

2. TF HTAP mission: Develop a fuller understanding of hemispheric transport of air pollution to inform future negotiations under CLRTAP Co-chairs: Terry Keating (U.S. EPA) and André Zuber (EC) TF HTAP Products: 2007 (Interim) and 2009 (Final) Assessment reports 49 parties in Europe, North America, and Central Asia Convention on Long-Range Transboundary Air Pollution (CLRTAP)

3. Literature Estimates of Surface O 3 Enhancements at Northern Mid-latitudes from Hemispheric Transport Difficult to conduct meaningful assessment due to differences in: 1) methods 2) regional definitions 3) reported metrics USA EU ASIA Estimates are from studies cited in current draft of TF HTAP interim report Ch5, updated from tables in Holloway et al., ES&T, 2003 and Fiore et al., EM, 2003 RECEPTOR REGIONS SOURCE REGIONS: N. AMERICA EUROPE ASIA BACKGROUND Solid bars: annual means Thick dashes: events Thin dashes: seasonal mean

4. HTAP Multi-model approach to estimate intercontinental source-receptor (S-R) relationships for ozone 20% decreases in:  anthrop. emis. in HTAP regions for NO x, CO, NMVOC individually  anthrop. emis. of all O 3 and aerosol precursors in HTAP regions  global CH 4 Conduct base simulations with 3-D models  horizontal resolution of 4°x5 ° or finer  2001 meteorology  each group’s best estimate for emissions in 2001  methane set to a uniform value of 1760 ppb Conduct sensitivity simulations (17 total) NA EU SA EA

5. Surface O 3 seasonal cycles in base simulations over the HTAP regions  Model range often spans ~15 ppbv Surface O 3 (ppbv) EUNA SA Month of 2001 17-model mean 17-model range GEOS-Chem 2x2.5 GEOS-Chem 4x5 EA Other HTAP O 3 models: CAMCHEM ECHAM5-HAMMOZ EMEP FRSGCUCI GEMAQ INCA LLNL-IMPACT-T5a MOZARTGFDL MOZECH OsloCTM2 STOC-HadAM3 STOCHEM TM5-JRC ULAQ UM-CAM

6. Recent ACCENT multi-model surface O 3 vs. observations in HTAP regions: NA and EU North America Ellingsen et al., in prep Europe and North Africa OBS Multi-model mean standard deviation range SW US: 11 sitesSE US: 12 sitesUS Great Lakes: 13 C. Mediterranean: 21C. Europe: 101 sitesMiddle East: 1 site 25 participating modelsYear 2000

7. Estimates of S-R relationships: Annual mean surface ozone decrease due to 20% global methane reduction  1 ppbv O 3 decrease over all NH receptor regions  Consistent with prior studies What causes the ~1 ppbv range across individual models? -- examine ozone production efficiency; NO x :VOC ratios Full range of 14 models EUROPEN. AMER.S. ASIAE. ASIA HTAP receptor region x x x x x x GEOS-Chem 2x2.5 x GEOS-Chem 4x5 x x x 2.0 1.5 1.0 0.5 0.0 ppbv

8. Estimates of S-R relationships: Sensitivity of surface O 3 to NO x sources during season of max domestic O 3 production Season:JJA JJA SON JJA Source region:  20% decreases in foreign NO x ) 20% decrease in domestic NO x = 10-30% Largest S-R pair: NA  EU Largest HTAP influence on SA Smallest HTAP influence on NA Similar impact from all 3 foreign source regions Decrease in surface O 3 (ppbv)  Examine springtime when intercontinental transport typically peaks  Compare role of NO x emissions with other O 3 precursors 14 model range Receptor region: x x x x x x x x x x x x x x x x x x x x x GEOS-Chem

9. EU and NA ≥ SA RECEPTOR = EAST ASIA   20% foreign emis.) -20% domestic emis. ≈ 85% ppbv Estimates of S-R relationships: Springtime surface O 3 decrease in HTAP receptor regions from 20% reductions of O 3 precursors EA/EU contribute similarly EU VOC at least as important as NO x RECEPTOR = NORTH AMERICA ppbv   20% foreign emis.) -20% domestic emis. ≈ 55% SOURCE REGIONS:  3 foreign regions Larger uncertainty in role of foreign VOC

10. Some remaining questions... (to be examined as part of the TF HTAP project) Which processes contribute most to uncertainty in S-R relationships?  Insights from idealized tracer experiments  Normalize model response to emission perturbations How can we best test models of hemispheric transport of air pollution?  Need to test processes (export, chemical evolution, mixing to surface) Can 20% emission decreases be scaled to reductions of other magnitudes?  Yes for CH 4 ; simulations underway for other O 3 precursors Role of hemispheric transport in polluted vs. background conditions?  Examine range of model responses across the O 3 distribution How will climate change affect hemispheric transport of air pollution? www.htap.org for more info and to comment on draft 2007 interim report Overall conclusion from previous analysis: Range of estimates narrows from that in the literature under HTAP setup