1. EFFECTS OF TRANSBOUNDARY POLLUTION AND CLIMATE CHANGE ON U.S. AIR QUALITY Daniel J. Jacob with Rokjin J. Park, Lin Zhang, Huiqun Wang, Philippe LeSager, Tzung-May Fu, Noelle E. Selin, Loretta J. Mickley, Shiliang Wu, Eric M. Leibensperger and funding from EPRI, EPA, DOE, NSF, NASA

2. EPRI STARTED IT ALL! Early work spurred major subsequent investment by EPA, NASA, NOAA Cited 127 times as of today (Thompson Science Citation Index)

3. IMPLICATION OF RISING OZONE BACKGROUND FOR MEETING AIR QUALITY STANDARDS 0 20 40 60 80 100 120 ppb Europe AQS (seasonal) U.S. AQS (8-h avg.) U.S. AQS (1-h avg.) Preindustrial ozone background Present-day ozone background at northern midlatitudes Europe AQS (8-h avg.) EPA policy-relevant background (PRB) : U.S. surface ozone concentrations that would be present in absence of North American anthropogenic emissions Increase driven by global NO x, methane PRB

4. RISING ASIAN EMISSIONS AS SEEN FROM SPACE 1996-2002 trend in tropospheric NO 2 columns from GOME [Richter et al., 2005] 2006 observations of NO 2 columns from OMI show doubling of Chinese emissions since 2000 [Zhang et al., in prep.]

5. CONSEQUENCES FOR U.S. SURFACE OZONE GEOS-Chem Asian pollution:9 ± 2 Observed: 54 ± 10 GEOS-Chem: 53 ± 8 NASA INTEX-B aircraft/satellite campaign (April-May 2006) Observations (Dan Jaffe, UW) and GEOS-Chem at Mt. Batchelor, Oregon (2,700 m) 2000-2006 rise in Asian emissions has increased U.S. surface ozone by 1-2 ppbv in the West (3 ppbv at Mt. Batchelor) Zhang et al., in prep. Mean Asian surface pollution enhancement (GEOS-Chem) 4/17 4/23 4/29 5/5 5/11 ppb

6. U.S. SURFACE OZONE ENHANCEMENTS FROM CANADIAN AND MEXICAN POLLUTION Jun-Aug mean max-8h-avg values from a GEOS-Chem simulation (1 o x1 o resolution) with zero U.S. anthropogenic emissions and with PRB subtracted Wang et al., in prep. Regions of maximum influence: Northeast for Canada Southwest for Mexico Maximum influence in summer for both PRELIMINARY DATA (mask did not remove U.S. offshore emissions)

7. EPISODIC NATURE OF CANADIAN POLLUTION INFLUENCE Jun-Aug 01 time series of max-8h-avg ozone at Unionville, MI Wang et al., in prep. Observations Model (standard) Transboundary PRB Canada + Mexico Unlike PRB, Canadian influence shows large episodes (up to 35 ppb); model is testable with observations for these episodes ** PRELIMINARY DATA **

8. CANADIAN POLLUTION INFLUENCE: ENSEMBLE VIEW 16 U.S. sites with max Canadian Influence: model vs. obs correlation coefficients are 0.67-0.82 for individual sites Model, ppb Observed, ppb daily max -8h-avg ozone, Jun-Aug 2001 Transboundary pollution Wang et al., in prep. median max ** PRELIMINARY DATA ** maximum Total Canada

9. ELEVATED ORGANIC CARBON (OC) PM IN ASIAN OUTFLOW ACE-Asia aircraft observations offshore from China (spring 2001) Sulfate is scavenged from outflow but not OC Observations show a 1-3  g m -3 OC background missing from model Heald et al. [2005] Contribution to intercontinental pollution? Observed GEOS-Chem

10. NEW MECHANISM FOR OC FORMATION FROM DICARBONYLS Acetylene Aromatics Isoprene Terpenes Acetone Anthropogenic VegetationBiomass burning OH dissolution hydration oxidation oligomerization glyoxal methylglyoxal evaporation secondary organic aerosol (SOA) Lab studies show irreversible uptake of glyoxal and methylglyoxal by aqueous particles and apparent oligomerization This mechanism has now been implemented in GEOS-Chem including detailed gas-phase chemistry of dicarbonyl formation Fu et al., in prep.

11. SIMULATION OF OC PM OVER EASTERN U.S. Water-soluble OC (WSOC) PM observations by Rodney Weber (Georgia Tech) during ICARTT aircraft campaign out of Portsmouth, NH (Jul-Aug 04) Observed Model with dicarbonyl SOA Model without dicarbonyl SOA Boundary layer data (< 2 km) Observed model with dicarbonyl SOA model without dicarbonyl SOA Fu et al., in prep.

12. CONTRIBUTIONS TO N. AMERICAN MERCURY DEPOSITION FROM GLOBAL POOL vs. REGIONAL POLLUTION POOL Hg(0) Hg(II) N. American boundary layer Hg(0) emission (53%) Hg(II) Global pool (lifetime ~ 1 y) Regional pollution pool Hg(II) emission (RGM 33%. Hg(P) 14%) reduction External anthropogenic Oceans Land N. America accounts for only 7% of global anthro. emission (2000) NORTH AMERICA cycling and re-emission

13. GEOS-Chem GLOBAL 3-D ATM-LAND-OCEAN MODEL FOR MERCURY Inventories in Mg Rates in Mg y -1 Selin et al. [submitted] Accounts for coupling between atmospheric and surface reservoirs on short-term to centurial time scales

14. SIMULATION OF U.S. MERCURY DEPOSITION NETWORK (MDN) DATA Seasonal variation vs. latitude Observations Model Model (N.Amer. sources) Annual mean wet deposition fluxes, 2004-2005: observations = circles, model = background Selin et al., in prep. Major contribution to mercury deposition from subsidence and scavenging of Hg(II) from global pool

15. SOURCE ATTRIBUTION FOR U.S. MERCURY DEPOSITION % contribution of N. American sources to annual total mercury deposition Natural (32%) North American anthropogenic (20%) Rest of world anthropogenic (31%) – half cycled through ocean on annual time scale Legacy anthropogenic re-emitted from soil on centurial time scale (16%) Selin et al., in prep.

16. OBSERVED DEPENDENCE OF AIR QUALITY ON WEATHER WARNS OF POTENTIALLY LARGE EFFECT OF CLIMATE CHANGE Interannual variability of exceedances of ozone NAAQS in the Northeast # summer days with 8-hour ozone > 84 ppbv, average for 257 northeast U.S. AIRS sites 1988, hottest of record Ozone is strongly correlated with temperature in observations (R 2 ~ 0.5); reflects (1) chemistry, (2) biogenic VOC emissions, (3) association with stagnation Lin et al. [2001] 1992, coldest on record (Pinatubo)

17. PROJECTING EFFECT OF CLIMATE CHANGE ON OZONE AIR QUALITY USING OBSERVED OZONE-TEMPERATURE CORRELATIONS Probability of max 8-h O 3 > 84 ppbv vs. daily max. temperature Projected T change for northeast U.S. in 2000-2100 simulated with ensemble of downscaled GCMs for different scenarios [IPCC, 2007]  T = 3K Probability of ozone AQS exceedance Temperature, K Probability of exceedance doubles By 2025,  T = 1-3 K depending on model and scenario; use statistical approach at right to infer increased probability of ozone exceedance for a given region or city assuming nothing else changes. Effect is large! Lin et al. [2001] Northeast Los Angeles Southeast Lin et al., in prep.

18. COMPREHENSIVE MODEL APPROACH FOR INVESTIGATING EFFECT OF CLIMATE CHANGE ON AIR QUALITY Global climate model (GCM) Global chemical transport model (CTM) for ozone-PM Regional climate model (RCM) Regional CTM for ozone-PM boundary conditions input meteorology input meteorology boundary conditions IPCC future emission scenario greenhouse gases ozone-PM precursors internal EPA project (CIRAQ) and several EPA-STAR projects

19. EPA-STAR GLOBAL CHANGE AND AIR POLLUTION (GCAP) PROJECT GISS GCM 3 GEOS-Chem global CTM MM-5 RCM CMAQ regional CTM boundary conditions input meteorology input meteorology boundary conditions IPCC emission scenario greenhouse gases ozone-PM precursors Applied to 2000-2050 global change simulations with IPCC SRES A1 scenario; compare 2050 climate (GCM 2049-2051, 3-y averages) to 2000 (1999-2001) D.J. Jacob and L.J. Mickley (Harvard), J.H. Seinfeld (Caltech), D. Rind (NASA/GISS), D.G. Streets (ANL), J. Fu (U. Tenn.), D. Byun (U. Houston)

20. 2000-2050 EMISSIONS OF OZONE PRECURSORS (A1) GlobalUnited States 2000 emissions % change, 2000-2050 2000 emissions% change, 2000-2050 NO x, Tg N y -1 Anthropogenic Lightning Soils (natural) 34 4.9 6.1 +71% +18% +8% 6.0 0.14 0.35 -39% +21% +11% NMVOCs, Tg C y -1 Anthropogenic Biogenic 46 610 +150% +23% 9.3 40 -52% +23% CO, Tg y -1 1020+25%87-47% Methane, ppbv17502400 (+37%) Wu et al. [in press] Global increase in anthropogenic emissions but 40-50% decreases in U.S. Climate-driven increases in natural NO x, NMVOC emissions

21. CHANGES IN SUMMER MEAN 8-h AVG. DAILY MAXMUM OZONE FROM 2000-2050 CHANGES IN CLIMATE AND GLOBAL EMISSIONS Models agree that 2000-2050 climate change will decrease background ozone but will increase surface ozone in U.S, with effects ranging from <1 ppb to up to 10 ppb Most but not all models find maximum effect during pollution episodes All models find significant effect in Northeast but disagree in other regions Differences in Southeast may partly reflect different mechanisms for isoprene chemistry Effect of changing climate Effect of changing emissions GCAP results [Wu et al., in press] ppb

22. CLIMATE CHANGE PENALTY: MEETING A GIVEN AIR QUALITY GOAL WILL REQUIRE GREATER EMISSION REDUCTIONS IN FUTURE CLIMATE 2000 conditions NOx emission - 40% (2000 climate) NOx emission - 40% (2050 climate) NOx emissions - 50% (2050 climate) In this example, 2000–2050 climate change implies an additional 25% reduction in NOx emissions (from 40% to 50%) to achieve the same ozone air quality. Wu et al. [in press]

23. NO x EMISSION REDUCTIONS DECREASE CLIMATE CHANGE PENALTY …and can even turn it into a climate benefit Change in summer 90 th percentile max-8h-avg surface ozone (ppb) from 2000-2050 climate change with 2000 emissions with 2050 emissions Reducing U.S. anthrop. emissions (1) increases relative influence of background, (2) decreases isoprene efficiency for making ozone Wu et al. [in press] ppb

24. CHANGES IN ANNUAL MEAN PM 2.5 FROM 2000-2050 CHANGES IN CLIMATE AND GLOBAL EMISSIONS Effect of changing climate Effect of changing emissions GCAP results [Shiliang Wu, Harvard]  g m -3 Effect of climate change on PM 2.5 is more complex than for ozone because different components have different sensitivities to meteorological variables (including sign) Most important meteorological variable is precipitation – difficult to predict in climate models Overall effect likely less important than for ozone  g m -3

25. INCREASING WILDFIRE FREQUENCY IN PAST DECADES Westerling et al. [2006] Temperature and drought index explain 50- 60% of interannual variability in fires Fires projected to increase by 50% by 2050, increasing PM 2.5 in western U.S. by 10% Spracklen et al. [in press] S. California fire plumes, Oct. 25 2004

26. 2000-2050 DECREASE OF CYCLONE FREQUENCY Wu et al. [in press] 1999-20012049-2051 Summertime cyclone frequency decreases by 17% in 2050 climate (GISS GCM A1 scenario) Cold fronts associated with mid-latitude cyclones tracking across Canada are the principal process ventilating Midwest and Northeast Clean air sweeps behind cold front

27. CORRELATION BETWEEN 40-50 o N CYCLONE FREQUENCY AND OZONE AIR QUALITY STANDARD EXCEEDANCES, 1980-1998 observed summer cyclone tracks (NCEP reanalysis) Correlation coefficient (r) between # ozone AQS exceedances per summer and # 40-50N cyclones Leibensperger et al., in prep.

28. TREND IN SUMMERTIME CYCLONE FREQUENCY, 1950-2006 1950 1960 1970 1980 1990 2000 2010 1980-2006 trend Observations (NCEP reanalysis) GISS GCM Leibensperger et al., in prep.

29. DECREASE IN CYCLONE FREQUENCY SINCE 1980 HAS IMPAIRED IMPROVEMENTS IN OZONE AIR QUALITY 1980 1985 1990 1995 2000 Observed 1980-1998 JJA trends in daily # 40-50N cyclones # 80 ppb O 3 AQS exceedances Interannual variability in the two is highly anticorrelated (r = - 0.64) # AQS exceedances # cyclones Black: observed AQS exceedances Red: AQS exceedances predicted from trend in cyclone frequency Green: AQS exceedances predicted in absence of trend in cyclone frequency # 80 ppb exceedance days in Northeast dropped from 38 in 1980 to 19 in 1998, but would have dropped to 5 in absence of cyclone trend! Leibensperger et al., in prep.