Air Quality Applied Sciences Team (AQAST) EARTH SCIENCE SERVING AIR QUALITY MANAGEMENT NEEDS Earth science resources Air Quality Management Needs Pollution monitoring Exposure assessment AQ forecasting Source attribution of events Quantifying emissions Assessment of natural and international influences Understanding of transport, chemistry, aerosol processes Understanding of climate-AQ interactions satellites AQAST suborbital platforms models AQAST

AQAST Membership: PIs and Co-Is Daniel Jacob (leader), Loretta Mickley (Harvard) Greg Carmichael (U. Iowa) Dan Cohan (Rice U.) Russ Dickerson (U. Maryland) Bryan Duncan, Yasuko Yoshida, Melanie Follette-Cook (NASA/GSFC); Jennifer Olson (NASA/LaRC) David Edwards (NCAR) Arlene Fiore (Columbia); Meiyun Lin (Princeton) Jack Fishman, Ben de Foy (Saint Louis U.) Daven Henze, Jana Milford (U. Colorado) Tracey Holloway, Steve Ackerman (U. Wisconsin); Bart Sponseller (Wisconsin DRC) Edward Hyer, Jeff Reid, Doug Westphal, Kim Richardson (NRL) Pius Lee, Tianfeng Chai (NOAA/NESDIS) Yang Liu, Matthew Strickland (Emory U.), Bin Yu (UC Berkeley) Richard McNider, Arastoo Biazar (U. Alabama – Huntsville) Brad Pierce (NOAA/NESDIS) Ted Russell, Yongtao Hu, Talat Odman (Georgia Tech); Lorraine Remer (NASA/GSFC) David Streets (Argonne) Jim Szykman (EPA/ORD/NERL) Anne Thompson, William Ryan, Suellen Haupt (Penn State U.)

AQAST organization AQAST supports two types of projects: Investigator Projects (IPs). These involve typically a single AQAST member working with an air quality management partner; Tiger Team Projects (TTPs). These involve a collaboration of several AQAST members pooling their expertise to address urgent needs from one or more air quality management partners. AQAST projects must bridge Earth Science and air quality management: Focus on use of Earth Science resources Clear air quality management outcomes 1-year deliverables of value for air quality agencies AQAST has great flexibility in how it allocates its resources Members are encouraged to adjust their IPs to the evolving needs of air quality management The TTPs are redefined yearly to address the most pressing needs The team is self-organizing and can respond quickly to demands

Scope of current AQAST projects (IPs and TTPs) Partner agency IC/BC for AQ models Future satellites AQ processes SIP Modeling Background AQ-Climate Forecasting Monitoring Emissions Local: RAQC, BAAQD State: TCEQ, MDE, Wisconsin DNR, CARB, Iowa DNR, GAEPD, GFC Interstate: EPA Region 8, LADCO National: NPS, NOAA, EPA Theme Satellites: MODIS, MISR, MOPITT, AIRS, OMI, TES, GOES Suborbital: ARCTAS, DISCOVER-AQ, ozonesondes, PANDORA Models: MOZART, CAM AM-3, GEOS-Chem, RAQMS, STEM, GISS, IPCC Earth Science resource

AQAST bimonthly newsletter subscribe through web site or send me email

GEOS-Chem model simulation of background ozone over the US in support of the current ozone ISA and REA 2x2.5 0.5x0.667 Color scale Indicates topography (surface pressure) 0.5o x0.67o resolution for North America, 2o x2.5o global, 48 vertical layers Driven by NASA GEOS-5 assimilated meteorological data Detailed mechanistic representation of ozone-NOx -VOC-aerosol chemistry 2006 anthropogenic emissions: NEI05 for US, CAC for Canada, BRAVO (scaled) for Mexico, EMEP for Europe, Streets for East Asia, EDGAR (scaled) for rest of world Natural sources: lightning (OTD/LIS), fires (GFED2 monthly), soils (Yienger), stratosphere (Linoz) 2006-2008 3-year simulation; focus on 2006 for evaluation Zhang et al. [Atmos.Environ., 2011] – DOWNLOAD PDF HERE

Four GEOS-Chem simulations Standard – as described above US background – no US anthro emissions NA background - no N.American anthro emissions Natural – no anthro emissions worldwide 2006 MDA8 ozone at CASTNet sites- with mean (4th highest) inset For discussion of US background see Wang et al. [2009] (DOWNLOAD PDF)

Ozone statistics for the US (CASTNet sites) Natural: 18 ± 6 PRB: 27 ± 8 Model: 52 ± 12 Obs: 50 ± 13 + > 1.5 km Frequency distribution of MDA8 ozone for March-August 2006 27 ± 6 GEOS-Chem is overall unbiased PRB is 4 ppb higher than in previous GEOS-Chem versions Model shows little interannual variability for 2006-2008, neither do observations 40 ± 7 57 ± 10 58 ± 9 Zhang et al. [2011]

Model “4th highest” MDA8 ozone in 2006 Annual 4th highest ozone 4th highest PRB value PRB for annual 4th highest ozone Background may limit or confuse ability of West to achieve a 60 ppb NAAQS; no such issue in East Zhang et al. [2011]

Model PRB statistics vs. observed ozone in Intermountain West Comparison of daily MDA8 ozone concentrations for the ensemble of elevated sites (>1.5 km; 11 sites) in Intermountain West for spring and summer 2006 1:1 line max min 75th 50th 25th r = 0.57 r = 0.33 Model underestimates high extrema in spring (> 75 ppbv) - stratosphere Correlation with measurements is poor in summer – lightning, wildfires Lin Zhang, Harvard

Model underestimate of stratospheric intrusions in spring Sample stratospheric intrusion at Pinedale (WY) : model captures timing but not magnitude. Obs. observations model PRB stratosphere Observed vs. simulated MDA8 ozone in spring 2006 at the 11 elevated CASTNet sites, colored by model stratospheric ozone concentration. Model has correct STE and vertical transport; we attribute problem to the general difficulty of Eulerian models in resolving fine structures. Lin Zhang, Harvard

Ozone enhancement from lightning in Intermountain West observations model PRB lightning Model lightning increases ozone concentrations by 7-19 ppbv in summer, comparable to enhancements from US anthropogenic emissions. This lightning influence in the model is biased high, even though lightning location is constrained by OTD/LIS satellite observations; focus future model improvement on NOx yield per lightning flash, vertical distribution Note the model overestimates in August. Lin Zhang, Harvard

Sensitivity of model ozone to wildfire emission inventory Maximum fire enhancements of MDA8 ozone In GEOS-Chem – August 2006 Sensitivity of model ozone to wildfire emission inventory fire reports monthly Model is highly sensitive to choice of inventory, spatial and temporal resolution Influence of fires is mainly limited to fire region; little influence at CASTNet sites in 2006 Need to examine other years, MOPITT and AIRS CO data for fires Lin Zhang, Harvard

Process-oriented analysis of multi-platform observations with a new, global high-resolution chemistry-climate model AQAST PI Arlene Fiore (Columbia U.) Problems: 1) Coinciding ozone maxima of stratospheric and anthropogenic origin [e.g. Stohl and Trickl 1999; Cooper et al 2004; Ambrose et al., 2011] 2) Limitations of tropospheric CTMs in capturing observed dynamic variability in ozonesondes [e.g. Jaegle et al., 2003; Liang et al., 2007; Jonson et al., 2010] 3) Prior multi-model evaluations on campaign /monthly mean basis [e.g. Dentener et al 2006; Stevenson et al., 2006; Fiore et al., 2009] GFDL AM3 Our Approach: 1) Global high-res (50 km) with fully coupled strat+trop chemistry [Donner et al., 2011]; Nudged to GFS 2) Analyze transport events on daily basis, leveraging intensive measurements from CalNex 2010 Sondes AQS/CASTNet

Trans-Pacific transport of Asian pollution plumes to WUS often coincides with O3 injected from stratosphere from Asia Primarily strat 1-min: Observed GFDL AM3 Zero Asian emissions O3-strat Complex mixing of Asian pollution + stratospheric air in continental inflow How much does Asian pollution contribute to surface high-O3 events? Lin et al., JGR, 2012 – DOWNLOAD PDF HERE

Tighter standard… harder to attain with domestic control Asian pollution contribution to high surface O3 events, confounding to attain tighter standard in WUS Obs (CASTNet/AQS) AM3/C180 total ozone AM3/C180 Asian ozone June 21 2010 June 22 2010 Max daily 8-h average Tighter standard… harder to attain with domestic control Lin et al., JGR, 2012

Transport of Asian pollution on the isentropic surfaces to the lower troposphere over the LA Basin Asian enhancements to trop column O3 on May 8, 2010 Vertical cross-section of Asian O3 along California coast [ppb] 1-min: θ Previously identified isentropic transport mechanism [Brown-Steiner and Hess, 2011] [K] Asian ozone available to be intercepted by the elevated terrain or entrained into the daytime boundary layer (~1-4 km in depth) Lin et al., JGR, 2012

The Asian enhancement increases for total O3 in the 70-80 ppb range over Southern California, Arizona 1-min Overall patter  increase -- explain points in the green and red trapezoid -- statistics -- Model bias does not have a direct relationship with Asian enhancements 25th Lin et al., JGR, 2012

Two models show different strengths in capturing distributions of base-case and N. American background O3 U.S. CASTNet sites Observed GEOS-Chem total GFDL AM3 total GEOS-Chem NAB GFDL AM3 NAB below 1.5 km + above 1.5 km Zhang et. al.,2011 2006 MAM intermountain west sites > 1.5 km 2006 JJA intermtn west sites > 1.5 km 0.06 0.04 0.02 0.00 0.06 0.04 0.02 0.00 Frequency per ppb Summer (JJA); sites < 1.5 km Frequency unit? 0 20 40 60 80 100 20 40 60 80 Surface MDA8 O3 [ppb] GC and AM3 bracket observed distribution but switch from spring (AM3 higher; strat. O3?) to summer (GC higher; lightning NOx?) GC narrower N American background distribution Value in multi-model approach to inform policy Oberman et al., in prep 19