Asian and stratospheric influences on western U.S. ozone over the past 35 years

1. Asian and stratospheric influences on western U.S. ozone over the past 35 years Meiyun Lin Princeton University & NOAA GFDL Acknowledgements: O. Cooper (ESRL), P. Dolwick (EPA), A. Fiore (Columbia), R. Fine, M. Gustin (U. of Nevada), L. Horowitz (GFDL), A. Landford (ESRL), S. Oltmans (ESRL), J. Pinto (EPA), G. Tonnesen (EPA Regional 8), C. Wiedinmyer (NCAR) Western Regional Air Quality Partnership (7/10/2013)

2. U.S. background surface O3 in spring Mean Maximum (at grids with monitors) bias-corrected [ppbv] 30 35 40 45 50 55 60 65 LARGE SPATIAL VARIATION -Strongest in the West (~50 ppb on average, up to 60-75 ppb during events) -The monitoring gaps in the intermountain regions MAJOR CHALLENGES: -Wildfires[e.g. Pfister et al., 2008; Jaffe & Wigder 2012] -Rising Asian emissions[e.g. Jacob et al., 1999; Richter et al., 2005; Cooper et al., 2010] -Frequency of stratospheric intrusion events[e.g. Langford et al., 2009; Lin et al., 2012b]

3. Seasonality of Western U.S. background O3 High-altitude sites in U.S. West, 1988-2012 Observed Model Background max 75th median 25th min Need for process-level understanding from daily to multi-decadal time scales  Screening of exceptional events  Developing forecasting tools(with lead time of days to seasons)

4. The GFDL AM3 global chemistry-climate model Key model features: • Fullstrat & trop chem [Donner et al., 2011, Naik et al., 2013] as opposed to using a climatological stratosphere in prior CTMs • Stratospheric O3 tracer (O3S) defined relative to a dynamically-varying tropopause [Lin et al., 2012b; Prather et al., 2011] • Nudged to reanalysis U and V [Lin et al., 2012a] • Cubed-sphere grid at ~50 km or ~200 km resolution GFDL AM3 Satellites • Today’s talk: • Process-oriented analysis for NOAA CalNex 2010 • (~50 km AM3)[Lin et al., 2012a, 2012b, JGR] • Interannual variability and long-term trends (1978-2012) • (~200 km AM3) [Lin et al., 2013, in prep] In situ Surface networks

5. Asian influence (Events & Long-term trends)

6. Asian contribution to Western U.S. high-O3events Observed AM3 Model Model Asian June 21 2010 20% of the total June 22 2010 Current NAAQS [ppb] MDA8 O3 [ppb] Future NAAQS? • Asian influence may confound efforts to attain tighter standards [Lin et al., 2012a, JGR… Featured in Science Now, Nature News, AGU Editors’ Highlight]

7. Trends in NOx emissions from China versus U.S. (KNMI) Year Emission data in AM3: Lamarque et al., 2010 for 1980-2000, RCP 8.5 for 2005-2010

8. Did regional NOx reductions work? Observed GFDL AM3 See also Cooper et al., 2012 95% MAM 1988-2012 50% MDA8 O3 trends (ppb yr-1) Larger symbols indicate statistically significant trends - Ozone increases over WUS despite reductions in local emissions - AM3 captures key features of observed surface O3 trends 8

9. Stratospheric influence(Events & Interannual variability)

10. 25 35 45 55 65 75 Stratospheric intrusion events in spring 2010 NAAQS AQS/CASTNet, April 13 Background (~70%) April 12-16 NAAQS Model MDA8 O3[ppb] O3-Strat MDA8 O3 [ppbv] Observed MDA8 O3 [ppb] Not-so-rare events (see Table 1 of Lin et al., 2012b, JGR): • Thirteen intrusions identified for April-June 2010 • Contributed to a total of 27 exceedances, including the southwest states Discussed with EPA R8 managers to flag events

11. An unusually large tropopause fold over Southern California (May 23, 2010) The GFDL AM3 model captures key features observed by an ozone lidar [Langford et al, 2012] Altitude (km) Lidar O3S GFDL AM3 Aqua AIRS Ascending, 400 hPa Flight track, S. California • Useful for daily forecasts (Lin presented at Aura, Aqua, AGU meetings) O3 [ppb] 30 60 90 120 150 Lin, et al., 2012b, JGR

12. How typical were conditions in spring 2010? CalNex 2010 High-altitude sites 2012 2008 SIP O3S (ppb) 75th 50th 25th 1990 Apr-May • Using AM3 at ~2x2º resolutionwith constant ozone precursor emissions • Surface O3 in spring 2010 are not exceptionally higher as compared to other years • Stratospheric influence is the major driver of interannual variability 12

13. MDA8 O3 MDA8 O3 April-June 2008 April-June 2008 Stratospheric intrusion events in 2008 OBS ModelO3Strat • 2º AM3 model is useful for screening of “SI exceptional events” • But limitations, due to resolution of meso-scale meteorology 13

14. Stratospheric intrusion events over Nevada in 2012 Nevada Rural Ozone Initiative (c/o Mae Gustin & Rebekka Fine) OBS ModelO3Strat Counties with monitors 4/15 5/1 6/1 7/1 7/31 4/15 5/1 6/1 7/1 7/31 April-July 2012 April-July 2012 14

15. Climate drivers of strat-to-trop ozone transport Interannual climate variability: -El Niño Southern Oscillation (ENSO) -Volcanic eruption (e.g. Pinatubo) -Arctic Oscillation (AO) Fingerprints in surface O3 measured at WUS sites?! Pinatubo (1991) Major signatures: -Changes in mean vs. high-O3 events -Changes in lower stratospheric O3 -Changes in dynamics and associated transport events AO+ 15

16. Climate variability modulates western O3through stratospheric influence High-altitude sites O3S (ppb) 75th 50th 25th 1990 Apr-May • Increase most following La Niña conditions • Reach a minimum following the Mt. Pinatubo volcanic eruption 16

17. Changes in mean vs. high-O3 events NAAQS 1999 (La Niña) Frequent STT events in spring Langford et al., 2009 NAAQS 1992 (Pinatubo) Weaker events, lower mean values 17

18. Changes in surface O3 distribution OBS El Niño (Little overall change) La Niña(Increases in the mid-to-high distribution) O3-Strat Pinatubo (Significant decreases in the high tail)

19. Changes in lower stratospheric ozone r2 (OBS, AM3)= 0.81 r2 (OBS, AM3)= 0.82 12-month running mean • UT/LS O3 is observed to increase following El Niño conditions (see also Langford et al.,1998; Bronnimann et al., 2004; Manzini, 2009; Randel et al., 2009) • Little change in UT/LS O3during La Niña despite increased O3S in surface air 19

20. Changes in mid-trop O3 associated with transport (500hPa, Apr-May) AM3 model O3S Background O3 El Niño composite Strengthening mid-latitude jet Greater Asian pollution Jet anomaly La Niña composite Southward extension of the polar jet  Greater stratospheric influence Pinatubo (1992) AO+ [Stenchikov et al., 2002] The polar jet blows consistently from west to east  Lock cold Arctic air in the polar region [ppb] [ppb]  Future work: satellite constraints(new OMI/TES O3 retrievals; CO/O3 correlations?) 20

21. Take home messages • Rising Asian emissions contribute most to O3 increases observed at western U.S. sites during the past 20 years  Offsets ozone decreases from regional NOx reductions • Climate variability modulates western U.S. O3 through stratospheric influence GFDL AM3 Satellites - Enhanced lower stratospheric ozone for El Niño - Enhanced transport to the lower troposphere for La Niña  Implications for daily to seasonal forecasts  Implications for screening of exceptional events  Implications for realistic BCs for regional modeling In situ Surface networks (Comments and questions: Meiyun.Lin@noaa.gov)

22. Additional slides (for questions)

23. The GFDL AM3 model hindcasts: 1978- 2012 Model setup similar to Lin et al. [2012a, b], but at 2ºx2º resolution Designed to isolate the role of changes inmeteorology, background, STE, wildfires, methane, andAsian vs.North American anthropogenic emissions

24. Observed GFDL AM3 95% 50% MDA8 O3 (ppb yr-1) Larger symbols indicate statistically significant trends U.S. rural ozone trends in summer JJA 1988-2012 The model is filtered to remove local influence -- more representative of observed conditions at remote WUS sites 24

25. Changes in lower-strat O3 and the mid-trop response Weekly samples as in OBS AM3 monthly data 12-month running mean O3S Volcanic aerosols (Pinatubo) enhance ClOx-induced stratosphericO3 depletion [see also Prather, 1992; Fahey et al., 1993; McCormick et al., 1995]