INTERCONTINENTAL AND HEMISPHERIC POLLUTIONINTERCONTINENTAL AND HEMISPHERIC POLLUTION

Daniel J. JacobHarvard University

http://www-as.harvard.edu/chemistry/trop

PART 1: THE NORTHERN MID-LATITUDES PART 1: THE NORTHERN MID-LATITUDES POLLUTION BELTPOLLUTION BELT

SPATIAL SCALES OF AIR POLLUTION:SPATIAL SCALES OF AIR POLLUTION:A HISTORICAL PERSPECTIVEA HISTORICAL PERSPECTIVE

Urban(pre-1970s)

Regional (1970s)• acid rain• haze• ozone smog

Intercontinental (2000s)• ozone, PM• Mercury, POPs

Greenhouse and aerosol-driven climate change (1980s)

Stratospheric ozone depletion (1970s)

SOURCE CONTINENT RECEPTOR CONTINENTOCEAN

PART I: THE NORTHERN MID-LATITUDESPART I: THE NORTHERN MID-LATITUDESPOLLUTION BELTPOLLUTION BELT

THREE POLES OF ANTHROPOGENIC EMISSIONS:THREE POLES OF ANTHROPOGENIC EMISSIONS:NORTH AMERICA, EUROPE, EAST ASIANORTH AMERICA, EUROPE, EAST ASIA

109 atoms N cm-2 s-1

Anthropogenic2000 NOx emissions[IPCC, 2001]

20o

60o

40o

…define a northern midlatitudes pollution belt

Population

(billons)

NOx

(Tg N yr-1)

SOx

(Tg S yr-1)

Asia 2000

2020 (A2)

2020 (B2)

3.2

4.3

4.1

9

16

16

25

32

18

OECD90 2000

2020 (A2)

2020 (B2)

0.85

0.95

0.99

12

9

9

17

10

8

IPCC 2020projections(IMAGES model)

LARGE EPISODIC SOURCES FROM DUST AND FIRESLARGE EPISODIC SOURCES FROM DUST AND FIRES

Annual mean 2001dust emissions[Fairlie et al., 1004]…desertification could increase source

Annual meanfire emissionclimatology[Duncan et al., 2003]…climate warming,legacy of fire suppression could increase boreal forest fires

VERTICAL STRUCTURE OF THE ATMOSPHEREVERTICAL STRUCTURE OF THE ATMOSPHERE

Tropopause

Stratopause

Stratosphere

Troposphere

Ozonelayer

Mesosphere

Troposphere: 85% of atmospheric mass

Stratosphere: 15%

Mesosphere and above: less than 0.1%

Tropopause is at 8-18 km altitude depending on latitude and season

GLOBAL ATMOSPHERIC TRANSPORT:GLOBAL ATMOSPHERIC TRANSPORT:THE HADLEY CIRCULATION (1735)THE HADLEY CIRCULATION (1735)

HOT

COLD

COLD

Explains:• Intertropical Convergence Zone (ITCZ) with strong separation of northern and southern hemispheres

• Stronger winds in winter than in summer

Problem: does not account for Coriolis force. Meridional transport of air between Equator and poles would result in unstable longitudinal motion.

TROPICAL HADLEY CELLTROPICAL HADLEY CELL

• Easterly “trade winds” in the tropics at low altitudes• Westerlies at high altitudes• Subtropical anticyclones at about 30o latitude

Coriolis force in northern hemisphere pushes air to R of direction of motion;Hadley circulation initiated at Equator extends only to ~30o latitude

CLIMATOLOGICAL SURFACE WINDS AND PRESSURESCLIMATOLOGICAL SURFACE WINDS AND PRESSURES(January)(January)

stormtrack

CLIMATOLOGICAL SURFACE WINDS AND PRESSURESCLIMATOLOGICAL SURFACE WINDS AND PRESSURES(July)(July)

500 hPa (~6 km) CLIMATOLOGICAL WINDS IN JANUARY:500 hPa (~6 km) CLIMATOLOGICAL WINDS IN JANUARY:strong mid-latitude westerliesstrong mid-latitude westerlies

500 hPa (~5 km) CLIMATOLOGICAL WINDS IN JULY500 hPa (~5 km) CLIMATOLOGICAL WINDS IN JULYmid-latitude westerlies are weaker in summer than wintermid-latitude westerlies are weaker in summer than winter

TYPICAL TIME SCALES FOR HORIZONTAL TRANSPORTTYPICAL TIME SCALES FOR HORIZONTAL TRANSPORT

1 week1-2 months

1-2 months

1 year

LIFTING AND SUBSIDENCELIFTING AND SUBSIDENCE

Warm, wet or converging(low-pressure) area-

SMALL SCALE

Initiallifting

H2O condensationheats rising air, accelerates lifting (buoyancy)

1-3 km

BOUNDARYLAYER

FREETROPOSPHERE

Stable layer caps cloud

Outflow air cools and sinks - LARGE SCALE

SUBSIDENCE INVERSION

as air subsides it warms by compression

slow entrainmentIn boundary layer

CONTINENTAL VENTILATION CONTINENTAL VENTILATION AND INTERCONTINENTAL TRANSPORTAND INTERCONTINENTAL TRANSPORT

CONTINENTALBOUNDARY LAYER

mixing ~ 1 day

1-3 km

Source continent Ocean

weak winds

Fast removal of ozone, PM (deposition, chemistry)

FREE TROPOSPHERE

Receptor continent

      

frontsconvection

every ~ 5 days

strong winds

slow removal of ozone, PM

Subsidence

mixing~ 1 day

rain; scavengingof PM

Tropopause (8-18 km)

mixing~ weeks

STRATOSPHERE

Background

INTERCONTINENTAL TRANSPORT INTERCONTINENTAL TRANSPORT BETWEEN NORTHERN MIDLATITUDE CONTINENTSBETWEEN NORTHERN MIDLATITUDE CONTINENTS

Asia N. America Europe

Boundary layer

Free troposphere

lifting subsidence

boundary layer advection

Tropopause

HEMISPHERIC POLLUTION BACKGROUND

“Direct”intercontinental

transport

Mixing

• Direct intercontinental transport: fast (~1 week) transport from source to receptor continent; either by boundary layer advection or by lifting to lower free troposphere followed by subsidence

• Hemispheric pollution: pollution mixes in free troposphere, affecting free tropospheric background, in turn affecting surface concentrations by subsidence

2 km

ILLUSTRATION: ILLUSTRATION: GLOBAL TRANSPORT OF CARBON MONOXIDE (CO) GLOBAL TRANSPORT OF CARBON MONOXIDE (CO)

Sources of CO: Incomplete combustion (fossil fuel, biofuel, biomass burning), oxidation of VOCs

Sink of CO: atmospheric oxidation by OH radical (lifetime ~ 2 months)

MOPITT satellite observations ofCO concentrations at 500 hPa (~6 km)

SIMULATION OF CO TRANSPORT SIMULATION OF CO TRANSPORT WITH WITH GEOS-CHEMGEOS-CHEM GLOBAL 3-D MODEL GLOBAL 3-D MODEL

OF ATMOSPHERIC TRANSPORT AND CHEMISTRYOF ATMOSPHERIC TRANSPORT AND CHEMISTRY

• Model developed by Harvard Atmospheric Chemistry Modeling Group, presently used by 16 research groups in North America and Europe; documented in ~70 research publications.

http://www-as.harvard.edu/chemistry/trop/geos

3-d gridstructure

• Driven by assimilated meteorological data (“real winds”) from NASA Global Modeling and Assimilation Office (GMAO)

• Applied to a wide range of global and regional atmospheric problems involving ozone, PM, greenhouse gases, etc.

• Includes coupled ozone-PM-Hg simulation capability nested with EPA CMAQ regional model

• Most results presented today use a horizontal resolution of 2ox2.5o (~200 km); some use 1ox1o or 4ox5o

PART 1 (TRANSPORT): IMPORTANT POINTSPART 1 (TRANSPORT): IMPORTANT POINTS

• N. America, Europe, and Asia define a “northern mid-latitudes pollution belt” with fast westerlies driving circumpolar transport on time scale of a few weeks

• Lifting out of the continental boundary layer by convection and fronts is important for the intercontinental transport of ozone and PM (faster winds, longer lifetimes)

• This lifting and subsequent mixing in the free troposphere produces a “hemispheric pollution background” that contributes to surface pollution by subsidence. For ozone at least, this hemispheric pollution is more important than direct intercontinental transport

PART 2: OZONEPART 2: OZONE

ENVIRONMENTAL IMPACTS OF ATMOSPHERIC OZONEENVIRONMENTAL IMPACTS OF ATMOSPHERIC OZONE

NOx = NO + NO2: nitrogen oxide radicalsVOC (volatile organic compounds) = light hydrocarbons and substituted organic compounds

UV shield

Greenhousegas

Primary sourceof OH radicals

Smog

HEMISPHERIC OZONE POLLUTION:HEMISPHERIC OZONE POLLUTION:IMPLICATIONS OF ENHANCED BACKGROUND IMPLICATIONS OF ENHANCED BACKGROUND

FOR MEETING AIR QUALITY STANDARDS (AQS)FOR MEETING AIR QUALITY STANDARDS (AQS)

0 20 40 60 80 100 120 ppbv

Europe AQS(seasonal)

U.S. AQS(8-h avg.)

U.S. AQS(1-h avg.)

Preindustrialozone

background

Present-day ozone background at

northern midlatitudes

Europe AQS (8-h avg.)

QUANTIFYING THE OZONE BACKGROUND QUANTIFYING THE OZONE BACKGROUND BY CORRELATION WITH POLLUTION TRACERSBY CORRELATION WITH POLLUTION TRACERS

Summer afternoon data at eastern U.S. sites [Trainer et al., 1993]

Alternatives are to use back-trajectories, remote upwind sites; allIndicate background ozone concentrations in surface air of 20-45 ppbv[Altshuller and Lefohn, 1996]

OCCURRENCES OF VERY LOW OZONE (< 10 ppbv)OCCURRENCES OF VERY LOW OZONE (< 10 ppbv)AT U.S./EUROPEAN CONTINENTAL SITES AT U.S./EUROPEAN CONTINENTAL SITES

REFLECT LOCAL DEPLETION, NOT BACKGROUNDREFLECT LOCAL DEPLETION, NOT BACKGROUND

Harvard Forest, Massachusetts [Munger et al., 1996]

Local depletion is due to:• deposition (esp. at night when surface atmosphere is stratified)• chemical titration in fresh pollution plumes (esp. in winter)

GLOBAL MODEL BUDGET OF TROPOSPHERIC OZONEGLOBAL MODEL BUDGET OF TROPOSPHERIC OZONE

O3

O2 h

O3

OH HO2

h, H2O

Deposition

NO

H2O2

CO, VOC

NO2

h

STRATOSPHERE

TROPOSPHERE

8-18 km

Chem prod in troposphere

4330

1620

Chem loss in troposphere

3960

1650

Transport from stratosphere

390

390

Deposition 760

360

Tg O3 yr-1 present natural

NOx, CO, methane, nonmethane VOC (NMVOC) emissions

Ozone lifetime: ~1 wk in boundary layer~1 mo in free troposphere

Inventory (Tg):360 230

[Mickley et al., 1999]

Limiting ozone precursors: NOx and methane

SENSITIVITY OF GLOBAL TROPOSPHERIC OZONE SENSITIVITY OF GLOBAL TROPOSPHERIC OZONE INVENTORY (Tg) TO 50% GLOBAL REDUCTIONS IN INVENTORY (Tg) TO 50% GLOBAL REDUCTIONS IN

ANTHROPOGENIC EMISSIONSANTHROPOGENIC EMISSIONS

240

250

260

270

280

290

300

310

320

3301995 base case

50% methane

50% NOx

50% NMVOCs

50%NOx+NMVOCs

50% CO

50% all

naturalGEOS-CHEM model [Fiore et al., [2002]

NOx and methane have the greatest impacts

RISING METHANE OVER 20RISING METHANE OVER 20thth CENTURY CENTURY

Historical methane trend

Recent methane trend

The rise in methane over the 20th century accounts for about 50% of the concurrent rise in global tropospheric ozone according to models

Present-day sources (Tg yr-1) [IPCC, 2001]:• Natural: wetlands (180), termites (25), biomass burning (20)• Anthropogenic: livestock (90), rice (85), natural gas (60), landfills (50), coal (40)

Sink: oxidation by OH (lifetime 10 years)

NONOxx EMISSIONS (Tg N yr EMISSIONS (Tg N yr-1-1))

FOSSIL FUEL 23.1

AIRCRAFT 0.5

BIOFUEL 2.2

BIOMASSBURNING 5.2

SOILS 5.1

LIGHTNING 5.8

STRATOSPHERE 0.2

Limiting precursor for ozone production both regionally (smog) and globally

TROPOSPHERIC NOTROPOSPHERIC NO22

FROM THE GOME SATELLITE INSTRUMENT (July 1996)FROM THE GOME SATELLITE INSTRUMENT (July 1996)

Martin et al. [2002]

Maps the distribution of surface NOx emissions and has so far largely confirmed the validity of emission inventories

NONLINEAR DEPENDENCE OF OZONE PRODUCTION ON NONONLINEAR DEPENDENCE OF OZONE PRODUCTION ON NOxx

IS RESPONSIBLE FOR (1) HIGH OZONE BACKGROUND,IS RESPONSIBLE FOR (1) HIGH OZONE BACKGROUND,(2) DAMPED SENSITIVITY TO NO(2) DAMPED SENSITIVITY TO NOxx EMISSION REDUCTIONS EMISSION REDUCTIONS

NOx

Ozone production efficiency (OPE): number of ozone molecules producedper molecule of NOx consumed

Nitric acid (HNO3)

Peroxyacetylnitrate(PAN)

CONTINENTALBOUNDARY LAYER

High NOx:

OPE ~5-10

FREETROPOSPHEREfronts

convection

PAN NOx

HNO3

Lightning

Low NOx:

OPE ~50-100

NOx lifetime ~ hours

10-20% of emitted NOx is exported, mainly as PAN

O3

O3

1-3 km

BACKGROUND OZONE CONCENTRATIONS INCREASE BACKGROUND OZONE CONCENTRATIONS INCREASE WITH ALTITUDEWITH ALTITUDE

stratosphere

Latitude over NW PacificLongitude

Chinacoast

Californiacoast

[Browell et al., 2003]

Mean aircraft lidar observations over N Pacific (spring 2001)

…because of long lifetime, high-altitude production, transport from stratosphere

Climatology of observed ozone at 400 hPa (7 km altitude) in July from ozonesondes and MOZAIC aircraft

GEOS-CHEM model tropospheric ozone columns for July 1997.

NORTHERN HEMISPHERIC ENHANCEMENT OF OZONENORTHERN HEMISPHERIC ENHANCEMENT OF OZONE

Li et al. [2001]

LATE SPRING MAXIMUM OF OZONE BACKGROUND LATE SPRING MAXIMUM OF OZONE BACKGROUND AT NORTHERN MID-LATITUDESAT NORTHERN MID-LATITUDES

The origin of this spring maximum is complicated and reflects contributions from

• photochemical production• relatively long lifetime (thick stratospheric ozone column)• efficient lifting and fast westerly transport• maximum in stratospheric influence

Naja et al. [2003]

3-5 km

0-3 km

marine sites

MONTHLY MEAN AFTERNOON OZONE CONCENTRATIONS MONTHLY MEAN AFTERNOON OZONE CONCENTRATIONS AT NON-URBAN U.S. SITES (CASTNet NETWORK) IN 2001AT NON-URBAN U.S. SITES (CASTNet NETWORK) IN 2001

+

Natural ozone: 15-25 ppbv

Hemispheric pollution enhancement: 5-15 ppbv, highest in spring

* Observations

Background (no anthrop. emissions in N. America, present methane)

Natural (no anthrop. emissions globally, preindustrial methane)

Model: Base (2001)

Stratospheric influence

Fiore et al. [2002]

TRANSATLANTIC TRANSPORT OF N. AMERICAN OZONE TRANSATLANTIC TRANSPORT OF N. AMERICAN OZONE (GEOS-CHEM model results for 1997)(GEOS-CHEM model results for 1997)

Li et al. [2002]

APRIL

JULY

LL

HH

HH

LL

Iillustrates the stronger intercontinental influence in spring

Mace Head site

OZONE DATA AT MACE HEAD, IRELAND (MAR-AUG 1997)OZONE DATA AT MACE HEAD, IRELAND (MAR-AUG 1997)

Observed[Simmonds]

GEOS-CHEMmodel

N.America pollutionevents in model

Li et al. [2002]

Intercontinental pollution influence in surface air is not detectable from observations alone

QUANTIFYING INTERCONTINENTAL POLLUTION INFLUENCEQUANTIFYING INTERCONTINENTAL POLLUTION INFLUENCEREQUIRES MODEL SIMULATIONS WITH MODIFIED EMISSIONSREQUIRES MODEL SIMULATIONS WITH MODIFIED EMISSIONS

• Calculating a partial derivative of ozone relative to NOx emisions and extrapolating [Wild and Akimoto, 2001] iis unsatisfactory because of nonlinearity •Tagging model ozone by its location of origin (e.g., ozone produced over North America and transported to Europe) [Derwent et al., 2003] is unsatisfactory both because of nonlinearity and because it does not separate natural from anthropogenic production

•Statistics of observed ozone enhancements when trajectories point to an upwind continental origin [Weiss-Pezias et al., 2004] are unsatisfactory because of confounding effects from latitudinal and vertical gradients in background ozone

• Observations of intercontinental transport of pollution plumes in the free troposphere [Stohl and Trickl, 1999] are not necessarily relevant to enhancements in surface air

MEAN SURFACE OZONE ENHANCEMENTS FROM ANTHROPOGENIC MEAN SURFACE OZONE ENHANCEMENTS FROM ANTHROPOGENIC

NONOxx AND NMVOC EMISSIONS BY DIFFERENT CONTINENTS AND NMVOC EMISSIONS BY DIFFERENT CONTINENTS

GEOS-CHEMmodel, July 1997

North America

Europe

Asia

Li et al. [2002]

as determined from sensitivity simulations with these sources shut off

FORECASTING TRANSATLANTIC TRANSPORT OF FORECASTING TRANSATLANTIC TRANSPORT OF NORTH AMERICAN POLLUTION TO EUROPE WITH THENORTH AMERICAN POLLUTION TO EUROPE WITH THE

NORTH ATLANTIC OSCILLATION (NAO) INDEXNORTH ATLANTIC OSCILLATION (NAO) INDEX

NAO IndexNorth American ozone pollution enhancementAt Mace Head, Ireland (GEOS-CHEM model)

r = 0.57

NAO index = normalized surface P anomaly between Iceland and Azores

Li et al.[2001]

Greenhouse warming NAO index shift change in transatlantic transport of pollution

Fiore et al. [2001]and Li et al. [2002]

tropical air

Subsidence of Asian pollution+ local production

stagnation

RANGE OF RANGE OF INTERCONTINENTAL INTERCONTINENTAL OZONE POLLUTION OZONE POLLUTION

ENHANCEMENTSENHANCEMENTSAT SURFACE SITES AT SURFACE SITES (GEOS-CHEM model)(GEOS-CHEM model)

over U.S.

over Europe

• variability is small

• effect is maximum for ozone concentrations in mid-range (40-70 ppbv)

EFFECT OF NORTH AMERICAN SOURCESEFFECT OF NORTH AMERICAN SOURCESON EXCEEDANCES OF EUROPEAN AIR QUALITY ON EXCEEDANCES OF EUROPEAN AIR QUALITY

STANDARD (55 ppbv, 8-h average)STANDARD (55 ppbv, 8-h average)GEOS-CHEM modelresults, summer 1997

Number of exceedance days(out of 92)

# of exceedance days thatwould not have beenin absence of N.American emissions

Li et al. [2002]

OZONE POLLUTION ENHANCEMENTS ARE LARGER IN OZONE POLLUTION ENHANCEMENTS ARE LARGER IN FREE TROPOSPHERE THAN AT SURFACEFREE TROPOSPHERE THAN AT SURFACE

Mean GEOS-CHEM ozone enhancements at 45oN in summer from anthropogenic emissions of NOx and NMVOCs in different continents

N America Europe Asia

N. Americanemissions

Europeanemissions

Asianemissions

Li et al. [2002]

NOAA/ITCT-2K2 AIRCRAFT CAMPAIGN IN APRIL-MAY 2002 NOAA/ITCT-2K2 AIRCRAFT CAMPAIGN IN APRIL-MAY 2002 Monterey, CAMonterey, CA

High-ozone Asian pollution plumes observed in lower free troposphere but not at surface (Trinidad Head)

COO3

PAN

HNO3

May 5 plume at 6 km:High CO and PAN,no O3 enhancement

May 17 subsidingplume at 2.5 km:High CO and O3,PAN NOxHNO3

Hudman et al. [2004]

Observations by D. Parrish, J. Roberts, T. Ryerson (NOAA/AL)

CONCEPTUAL PICTURE OF OZONE PRODUCTIONCONCEPTUAL PICTURE OF OZONE PRODUCTIONIN TRANSPACIFIC ASIAN POLLUTION PLUMESIN TRANSPACIFIC ASIAN POLLUTION PLUMES

NOx

HNO3

PANAsianboundarylayer(OPE ~ 5)

PAN, weak O3

Warm conveyor belt; 5-10% export of NOy mainly as PAN

strong O3

Subsidence Over E Pacific

OPE 60-80PAN NOxHNO3

U.S.boundarylayer very weak O3

10x dilution(Asian dust data)

E. Asia Pacific United States

Hudman et al. [2004]

Stratosphericdownwelling

GEOS-CHEM

CALIFORNIA MOUNTAIN SITES ARE PARTICULARLY CALIFORNIA MOUNTAIN SITES ARE PARTICULARLY SENSITIVE TO ASIAN OZONE POLLUTIONSENSITIVE TO ASIAN OZONE POLLUTION

…because there is no dilution in the boundary layer…because there is no dilution in the boundary layer

Observed 8-h ozone at Sequoia National Park (1800 m) in May 2002vs. corresponding simulated (GEOS-CHEM) Asian pollution ozone enhancement

Asian enhancements are 6-10 ppbvduring NAAQS exceedances;unlike at surface sites, Asian pollution influence is not minimum under high-ozone conditions!

May 17 obs. Asian plume event in red

Hudman et al. [2004]

IPCC [2001] PROJECTION OF FUTURE METHANE EMISSIONSIPCC [2001] PROJECTION OF FUTURE METHANE EMISSIONS

Methane is the second most important anthropogenic greenouse gas after CO2

…and also a limiting precursor for global production of tropospheric ozone

Combined effects of future anthropogenic emission trendsCombined effects of future anthropogenic emission trendson U.S. ozone air quality and on global climateon U.S. ozone air quality and on global climate

50%

NM

VO

C

1995

(bas

e)

50%

CH 4

50%

NO

x20

30 A

120

30 B

1

50%

NM

VO

C50

% C

H 4

50%

NO

x20

30 A

120

30 B

1

IPCC scenario

Fossil fuel NOx emissions

(2020 vs. present)

Global U.S.

Methane concentration

(2020 vs. present)

A1 +80% -30% +35%

B1 +10% -60% +20%

GEOS-CHEM model simulations [Fiore et al.,2002]

Ozone pollution

LENGTHENING OF OZONE POLLUTION SEASON IN LENGTHENING OF OZONE POLLUTION SEASON IN UNITED STATES IN 2030 A1 SCENARIOUNITED STATES IN 2030 A1 SCENARIO

2030 A1

1995 Base Case

Fiore et al.[2002]

Rising background from methane and Asian NOx emissions has most effectIn spring

Fiore et al. [2002]

INCREASE IN FREE TROPOSPHERIC BACKGROUND OZONE INCREASE IN FREE TROPOSPHERIC BACKGROUND OZONE OVER EUROPE IN THE PAST CENTURYOVER EUROPE IN THE PAST CENTURY

Observations at mountain sites [Marenco et al., 1994]

Preindustrialmodel ranges

Are natural ozone sources (esp. lightning NOx) overestimated in models?…but the old observations also have uncertain calibrations

1970-200 TREND IN BACKGROUND OZONE IN EUROPE1970-200 TREND IN BACKGROUND OZONE IN EUROPEHohenpeissenberg and Payerne data

NOx emission trends

3-5 km

0.5-3 kmpolluted

background

Increase until the mid-1980s and then leveling off; would seem consistent with emission trends

Naja et al. [2003]

TRENDS IN OZONESONDE DATA TRENDS IN OZONESONDE DATA AT NORTHERN MID-LATITUDES, 1970-1995AT NORTHERN MID-LATITUDES, 1970-1995

observed

GEOS-CHEMmodel

Some indication of positive trend but models cannot reproduce• contrast between N. America vs. Europe and Asia• seasonal variation in the trend

Fusco and Logan [2003]

8-h daily maximum ozone frequency distribution at rural U.S. sites[Lin et al., 2000]

BACKGROUND OZONE IN SURFACE AIR OVER U.S.BACKGROUND OZONE IN SURFACE AIR OVER U.S.APPEARS TO HAVE INCREASED BY ~3 ppbv OVER THE PAST 20 YEARSAPPEARS TO HAVE INCREASED BY ~3 ppbv OVER THE PAST 20 YEARS

1980-1984 1994-1998

1980-1984

1994-1998

This increase in background and compression of the frequency distribution has also been observed in Switzerland [Bronnimann et al., 2002] and is consistent with models [Fiore et al., 2002]

OBSERVED TREND IN OZONE BACKGROUND OVER OBSERVED TREND IN OZONE BACKGROUND OVER CALIFORNIA IN SPRING SUGGESTS 10-15 ppbv INCREASECALIFORNIA IN SPRING SUGGESTS 10-15 ppbv INCREASE

OVER PAST 20 YEARSOVER PAST 20 YEARS

Trend: 0.5-0.8 ppbv yr-1

Jaffe et al. [2003]

Such a large increase is not consistent with models

EFFECT OF INCREASING SIBERIAN FOREST FIRES ON EFFECT OF INCREASING SIBERIAN FOREST FIRES ON SUMMER SURFACE OZONE IN PACIFIC NORTHWESTSUMMER SURFACE OZONE IN PACIFIC NORTHWEST

Mean summer 2003 enhancementof 5-9 ppbv (9-17 ppbv in events)

Jaffe et al. [2004]

Observations GEOS-CHEM ozone enhancements

Siberian fires

CO

Ozone

BACKGROUND OZONE IS NOT ONLY IMPORTANTBACKGROUND OZONE IS NOT ONLY IMPORTANTFOR EXCEEDANCE OF AIR QUALITY STANDARDS,FOR EXCEEDANCE OF AIR QUALITY STANDARDS,

IT IS ALSO IMPORTANT FOR SETTING THE STANDARDSIT IS ALSO IMPORTANT FOR SETTING THE STANDARDS

Environmental risk

PollutantconcentrationBackground AQS

Acceptableadded

risk

EPA defines as “Policy-Relevant Background” those concentrations that would be present in the absence of North American)anthropogenic emissions: presently 40 ppbv is assumed

Lefohn et al. [2001] challenge to 84 ppbv NAAQS…under current revision

20 40 60 80 100

presently usedby EPA (underreview)

O3 (ppbv)

Bacgkround range considered by EPA in last revision

of ozone standard

84 ppbv:current NAAQS

Frequent observationsat remote U.S. sites attributed to natural background

[Lefohn et al., “Present-day Variability of background ozone in the lower troposphere”,Journal of Geophysical Research, 106, 9945-9958, 2001]

Ozone time series at Ozone time series at CASTNet stations used by CASTNet stations used by

Lefohn et al. [2001]Lefohn et al. [2001]

CASTNet sitesModelBackgroundNatural O3 levelStratospheric

+

*

Hemisphericpollution

Regionalpollution}

}

Model reproduces structure; regional pollution is #1 factor,hemispheric pollution alsosignificant. Natural background has little variability

Fiore et al., JGR 2003

1-5 pm daily data

Daily afternoon (1-5 p.m.) surface ozoneDaily afternoon (1-5 p.m.) surface ozoneMarch-October 2001 March-October 2001

Background 15-35 ppbv; Natural 10-25 ppbv; Stratosphere < 20 ppbv

Pro

bab

ility

pp

bv-1

Typical ozone values in U.S. surface air:

Compiling model results from all CASTNet sites…

Observations at CASTNet sites

Model (base)

Stratospheric

NaturalBackground

DEPLETION OF OZONE BACKGROUND DEPLETION OF OZONE BACKGROUND DURING REGIONAL POLLUTION EPISODESDURING REGIONAL POLLUTION EPISODES

Background(clean conditions)

O3 vs. (NOy-NOx) At Harvard Forest,

Massachusetts

Background(pollution episodes)

Observed (J.W. Munger)model (GEOS-CHEM)model background

Pollution coordinateFiore et al. [2002]

Cumulative probability distributions Cumulative probability distributions for daily mean afternoon Ofor daily mean afternoon O3 3 at at

CASTNET sites, spring-summer 2001CASTNET sites, spring-summer 2001

CASTNet sitesModelBackgroundNatural O3 levelStratospheric

+

*

Fiore et al., JGR 2003

Apr-May Jul-Aug

11 remote sites (western U.S.)

34 polluted sites (eastern U.S.)

An improved specification of the “policy-relevant ozone background”should recognize decrease from spring to summer and under pollutedconditions; background should in any case be lower than 40 ppbv,Implying that current NAAQS is too lax

Regional pollution

PART 2 (OZONE): SUMMARYPART 2 (OZONE): SUMMARY

• The natural ozone concentration in surface air over northern mid-latitudes continents is 10-25 ppbv with little variability according to models; observations suggest that it could be even lower.

• The present-day background ozone concentration in surface air over norhern mid-latitudes continents (zeroing anthropogenic emissions from that continent) is 15-35 ppbv according to both models and observations.

• Observed trends in background over the past decades are inconsistent, and the ability of models to reproduce them is unclear.

• Simulated mean surface ozone enhancements from anthropogenic NOx and NMVOC emissions in other continents are typically 2-5 ppbv, highest when ozone concentrations are in midrange (40-70 ppbv).

• Anthropogenic methane adds another 4- 6 ppbv enhancement to surface ozone according to models.

• Background influence is typically low under regional pollution episodes, implying that air quality standards based on risk increments above background are currently too lax.

PART 2 (OZONE): FUTURE DIRECTIONSPART 2 (OZONE): FUTURE DIRECTIONS

• ICARTT aircraft campaign (joint U.S-E.U., summer 2004) will observe export from North America and transatlantic transport to Europe.

• Satellite observations are providing ability to map emissions of NOx and NMVOCs; the TES instrument (launched on Aura this month) will provide the first global mapping of tropospheric ozone.

• Nested regional-global models are being developed to refine source-receptor relationships.

• There is continued interest in statistical analysis of ozone background and trends at remote sites; but a synthetic approach is needed.

• Work to better quantify methane sources has been on the back-burner over the past few years, as the climate change community is fixated on CO2; interest from the air quality community would make a difference.

• Work to improve lightning source of NOx (currently uncertain by an order of magnitude!) is at a standstill and needs to be revived.

ICARTT: COORDINATED ATMOSPHERIC CHEMISTRY CAMPAIGN OVER ICARTT: COORDINATED ATMOSPHERIC CHEMISTRY CAMPAIGN OVER EASTERN NORTH AMERICA AND NORTH ATLANTIC IN SUMMER 2004EASTERN NORTH AMERICA AND NORTH ATLANTIC IN SUMMER 2004

SCIENTIFIC OBJECTIVES

• Regional Air Quality: characterize sources and transport of pollution in northeastern North America

• Continental Outflow: quantify North American outflow of environmentally important gases and aerosols, relate to sources

• Transatlantic Pollution: understand transport and chemical evolution of North American pollution across the Atlantic

• Aerosol Radiative Forcing: characterize direct/indirect effects of aerosols over northeastern North America and western North Atlantic

International, multi-agencycollaboration

ICARTT: THE PLAYERSICARTT: THE PLAYERS

NASA/INTEX

NOAA/ITCT

UK/ITOP

DLR, CNRS

Caltech/ONR

DOE/ASP

MSCNSF/COBRA

Zoom overnortheastern North America

NOAA/NEAQS

SATELLITE NEAR-REAL-TIME DATA DURING ICARTTSATELLITE NEAR-REAL-TIME DATA DURING ICARTT

(July 2000) AOD (July 2000)

CO from MOPITT, AIRS AODs (and fires) from MODIS

NO2 from SCIAMACHY HCHO, NO2 from GOME

GOMEHCHO(July 1996)

Monthly means from previous years

TRANSATLANTIC LAGRANGIAN EXPERIMENTTRANSATLANTIC LAGRANGIAN EXPERIMENTwill involve coordination of NOAA, NASA, UK, DLR aircraftwill involve coordination of NOAA, NASA, UK, DLR aircraft

NESTING REGIONAL AND GLOBAL MODELS:NESTING REGIONAL AND GLOBAL MODELS:necessary for fine source-receptor attributionnecessary for fine source-receptor attribution

(e.g., individual countries or states)(e.g., individual countries or states)

GLOBAL MODEL(GEOS-CHEM)

200 km resolution

Regional model (CMAQ)

36 km

12 km

4 kmBoundary conditions(“1-way nesting”)

Ongoing EPA/ICAP project: C. Jang (EPA), D. Byun (UH), D. Jacob (Harvard)

• Has been applied by D. Byun to examine Mexican pollution influences over Texas; application to transpacific pollution influence on continental U.S. is under way• Still very much in infancy; needs support from policy community• Another similar effort underway in U.S. involves U. Iowa (G. Carmichael) with GFDL (Larry Horowitz)

PART 2 (OZONE): WHO IS DOING THE WORK?PART 2 (OZONE): WHO IS DOING THE WORK?

• North America– Harvard (D. Jacob): ICARTT campaign– EPA (C. Jang) and U. Houston (D. Byun): nested regional-global modeling– Princeton/GFDL (A. Fiore): projections with future emissions– U. Washington (D. Jaffe, L. Jaegle): field studies and modeling of transpacific trasnport– NOAA/AL (D. Parrish, A. Stohl): ICARTT campaign, transport pathways

• Europe– U. East Anglia (S. Penkett), DLR (H. Schlager): ICARTT campaign– UK Met. Office (R. Derwent): observations and modeling of intercontinental influence

on Europe– EPFL (I. Bey): global modeling of intercontinental influence on Europe– ETH (J. Staehelin): ozone trend statistics

• E. Asia– FRSGC (H. Akimoto): ozone trend statistics, global modeling

PART 3: AEROSOLSPART 3: AEROSOLS

DUST STORMS PROVIDE THE VISIBLE EVIDENCEDUST STORMS PROVIDE THE VISIBLE EVIDENCEOF INTERCONTINENTAL TRANSPORT! OF INTERCONTINENTAL TRANSPORT!

GlenCanyon, AZ

Clear day April 16, 2001: Asian dust!

Mean April 2001PM concentrationsmeasured by MODIS

ASIAN AND SAHARAN DUST CLOUDS CAN CAUSE ASIAN AND SAHARAN DUST CLOUDS CAN CAUSE EXCEEDANCES OF PM AIR QUALITY STANDARDS IN U.S.EXCEEDANCES OF PM AIR QUALITY STANDARDS IN U.S.

April 1998 dust event [Husar et al., 2001]

Some of that dust could be anthropogenic (soil erosion)

LIFE CYCLE OF THE ATMOSPHERIC AEROSOLLIFE CYCLE OF THE ATMOSPHERIC AEROSOL

Soil dustSea salt

Aerosol: dispersed condensed matter suspended in a gasSize range: 0.001 m (molecular cluster) to 100 m (small raindrop)

SO2, NOx,NH3, VOCs

Most important components:-Sulfate- nitrate-ammonium-Organic carbon (OC), elemental carbon (EC)-Soil dust-Sea salt

SOURCES OF SULFATE-NITRATE-AMMONIUM SOURCES OF SULFATE-NITRATE-AMMONIUM AEROSOLS (2001)AEROSOLS (2001)

GLOBAL UNITED STATES

Sulfur,Tg S yr-1

Ammonia,Tg N yr-1

NOx,Tg N yr-1

78 8.3

55 2.8

43 7.4

SOURCES OF CARBONACEOUS AEROSOLS (1998)SOURCES OF CARBONACEOUS AEROSOLS (1998)

ORGANIC CARBON (OC)

ELEMENTAL CARBON (EC)

GLOBAL UNITED STATES

130 Tg yr-1

22 Tg yr-1

2.7 Tg yr-1

0.66 Tg yr-1

ANNUAL MEAN PARTICULATE MATTER (PM) ANNUAL MEAN PARTICULATE MATTER (PM) CONCENTRATIONS AT U.S. SITES, 1995-2000CONCENTRATIONS AT U.S. SITES, 1995-2000

NARSTO PM Assessment , 2003NARSTO PM Assessment , 2003

PM10 (particles < 10 m), g m-3 PM2.5 (particles < 2.5 m), g m-3

Red circles indicate violations of national air quality standard:50 g m-3 for PM10 15 g m-3 for PM2.5

> 5033-50< 33

> 1510-15< 10

Toronto (1997-99)Egbert (1994-99)

Abbotsford (1994-95)

Quaker City OH (1999)

Arendstville PA (1999)

Atlanta (1999)Yorkville (1999)Mexico City - Pedregal (1997)

Los Angeles (1995-96)

Fresno (1988-89)

Kern Wildlife Refuge (1988-89)

Sulfate

Nitrate

Ammonium

Black carbon

Organic carbon

Soil

Other

12.3 ug m-38.9 ug m-3

7.8 ug m-3

12.4 ug m-3

10.4 ug m-3

19.2 ug m-314.7 ug m-3

55.4 ug m-3

30.3 ug m-3

23.3 ug m-3

39.2 ug m-3

Washington DC (1996-99)

14.5 ug m-3

Colorado Plateau (1996-99)3.0 ug m-3

Mexico City - Netzahualcoyotl (1997)

24.6 ug m-3

Esther (1995-99)

St. Andrews (1994-97)5.3 ug m-3

4.6 ug m-3

PM2.5 COMPOSITION IN THE UNITED STATESPM2.5 COMPOSITION IN THE UNITED STATESAnnual meansAnnual means

SULFATE, NITRATE, AMMONIUM SULFATE, NITRATE, AMMONIUM AEROSOL CONCENTRATIONS AEROSOL CONCENTRATIONS

IN EUROPEIN EUROPE

Annual means [EMEP, 2003]

Sulfate

NitrateAmmonium

GLOBAL AEROSOL DISTRIBUTION OBSERVED BY GLOBAL AEROSOL DISTRIBUTION OBSERVED BY MODIS SATELLITE INSTRUMENTMODIS SATELLITE INSTRUMENT

• Dust and fires are larger global influences than pollution;• Contrast with remote background is much stronger than for ozone because aerosols are scavenged efficiently by precipitation

In contrast to ozone, direct intercontinental transport is more importantthan hemispheric pollution enhancement.

AIRCRAFT OBSERVATIONS IN ASIAN OUTFLOW AIRCRAFT OBSERVATIONS IN ASIAN OUTFLOW ILLUSTRATE THE DEPLETION OF AEROSOLS ILLUSTRATE THE DEPLETION OF AEROSOLS

DURING LIFTING TO FREE TROPOSPHEREDURING LIFTING TO FREE TROPOSPHERE

Longitude Browell et al. [2003]

INTERCONTINENTAL TRANSPORT OF ASIAN AND INTERCONTINENTAL TRANSPORT OF ASIAN AND NORTH AMERICAN ANTHROPOGENIC SULFATENORTH AMERICAN ANTHROPOGENIC SULFATE

As determined from GEOS-CHEM 2001 sensitivity simulations with these sources shut off

Enhancements are insignificant for health-based air quality standards; difference with ozone reflects (1) scavenging of aerosols during export; (2) larger increment from background to standard

ASIAN POLLUTION INFLUENCES ON SULFATE,ASIAN POLLUTION INFLUENCES ON SULFATE,NITRATE, AMMONIUMNITRATE, AMMONIUM

Annual means as determined from a GEOS-CHEM 2001 sensitivity simulation with Asian anthropogenic sources shut off

Sulfate

Nitrate

Ammonium

g m-3

Park et al.[2004]

INTERCONTINENTAL SULFATE DEPOSITIONINTERCONTINENTAL SULFATE DEPOSITION

N. American and Asian anthropogenic sources each contribute 2-4% to mean sulfate deposition over Europe; mainly driven by events [Tarrason and Iversen, 1998]

Intercontinental influence is weaker for surface concentrations than for deposition because of the additional dilution from subsidence

TRANSPACIFIC POLLUTION TRANSPORT EVENTTRANSPACIFIC POLLUTION TRANSPORT EVENT

Asian pollution plume sampledat Mt. Rainier, Mt. Lassen, and Crater Lake on 28 April 1993: 2.2 g m-3 ammonium sulfate 1.8 g m-3 dust 1.1 g m-3 organic carbon 0.22 g m-3 nitrate 0.16 g m-3 elemental carbon

Compare to 24-h NAAQS of 85 g m-3

Jaffe et al. [2003]

INTERCONTINENTAL FIRE INFLUENCEINTERCONTINENTAL FIRE INFLUENCE

Canadian fire plume sampled at Mace Head, Ireland [Forster et al., 2001]

CO enhancementsfrom Canadian firesBlack carbon

enhancement

Assuming a 7/1 organic/elemental aerosol carbon mass ratio from forest fires implies a PM enhancement less than 0.8 g m-3 on Aug. 13 event – not much!

VISIB ILITY REDUCTION BY AEROSOLSVISIB ILITY REDUCTION BY AEROSOLS

Scattering by particles is most efficient when radiation wavelength = particle radius

Visible light is in 0.4 – 0.7 m range, so fine aerosols (0.1-1 m) are efficient scatterers

EPA REGIONAL HAZE RULE:EPA REGIONAL HAZE RULE:

Federal class I areas in the U.S. (including national parks and other large wilderness areas) must return to “natural visibility” conditions by 2064

•

Acadia National Park

clean day moderately polluted day

http://www.hazecam.net/

…will require essentially total elimination of anthropogenic aerosols!

PHASE I IMPLEMENTATION OF REGIONAL HAZE RULEPHASE I IMPLEMENTATION OF REGIONAL HAZE RULE

State Implementation Plans (SIPs) must be submitted by 2007 for linear improvement in visibility over the 2004-2018 period toward the 2064 natural visibility endpoint

Because visibility is a logarithmic (sluggish) function of PM concentration,The 2004-2018 phase I implementation requires ~50% reduction in emissions, highly sensitive to specification of 2064 endpoint

visibility(deciviews)

from EPA [2001]

Anthropogenicemissions(illustrative)

““DEFAULT ESTIMATED NATURAL PM CONCENTRATIONS” DEFAULT ESTIMATED NATURAL PM CONCENTRATIONS” FOR APPLICATION OF THE REGIONAL HAZE RULEFOR APPLICATION OF THE REGIONAL HAZE RULE

PM mass concentration (g m-3) Extinction coefficient (Mm-1)

Intercontinental transport of pollution could prevent attainabilityof these natural PM targets through domestic emsision reductions only; becomes issue of “natural” vs. “background”

SURFACE PM IN EASTERN AND WESTERN U.S.:SURFACE PM IN EASTERN AND WESTERN U.S.:contributions from natural and transboundary pollutioncontributions from natural and transboundary pollution

• EPA default natural estimates are OK except for OC in west (forest fires)• Transboundary pollution of SO4

2- and NO3- makes natural visibility objective

unachievable without international controls• Transboundary sulfate pollution influence from Asia is comparable in magnitude to that from Canada + Mexico

Annual regional means from GEOS-CHEM standard and sensitivity simulations

(NH4)2SO4

(g m-3)

West East

NH4NO3

(g m-3 )

West East

OC

(g m-3 as OMC)

West East

Baseline (2001) 1.52 4.11 1.53 3.26 2.0 3.2

Natural (no global anthrop.) 0.11 0.11 0.03 0.03 1.2 1.1

Background (no U.S. anthrop.) 0.43 0.38 0.27 0.37 1.3 1.2

Transboundary pollution

Canada and Mexico

Asia

0.15

0.13

0.14

0.12

0.20

-0.02

0.25

-0.02

0.05

0.013

0.05

0.007

EPA default natural for RHR 0.11 0.23 0.1 0.1 0.5 1.4

Park et al. [2004]

IMPLICATIONS FOR 2004-2018 IMPLEMENTATION IMPLICATIONS FOR 2004-2018 IMPLEMENTATION OF REGIONAL HAZE RULEOF REGIONAL HAZE RULE

Illustrative calculation for mean western U.S. conditions, assuming linear relationship between emissions and PM concentrations, and assuming zero trend in anthropogenic sources from foreign countries

Desired trend in visibility

Required % decrease of U.S. anthropogenic emissions

Phase 1

30%

48%

Park et al. [2004]

PART 3 (AEROSOLS): SUMMARYPART 3 (AEROSOLS): SUMMARY

• Export of aerosols from continents is far less efficient than for ozone because of scavenging by precipitation

• Intercontinental transport of pollution aerosols is negligible with regard to meeting current PM2.5 and PM10 standards; intercontinental transport of dust is of more concern

• Intercontinental transport is more important for acid deposition than for PM air quality standards

• Intercontinental pollution transport enhances sulfate concentrations several-fold relative to natural sources – important for formulation of U.S. EPA Regional Haze Rule

PART 3 (AEROSOLS): WHO IS DOING THE WORK?PART 3 (AEROSOLS): WHO IS DOING THE WORK?

• North America

– Harvard (D. Jacob): ICARTT campaign, modeling aerosol background over U.S., MODIS data analysis

– EPA/IMPROVE (W. Malm) background aerosol measurements in U.S.

– NASA/GSFC (M. Chin): MODIS data analysis

– U. Washington (D. Jaffe): statistical analysis of transpacific transport

– NOAA/AL (A. Stohl): global transport pathways

– U.C. Irvine (C. Zender): dust sources and transport

• Europe

– U. Oslo (L. Tarrason): hemispheric modeling of intercontinental transport

– ISPRA (F. Dentener): global modeling of intercontinental transport

PART 4: MERCURYPART 4: MERCURY

PRESENT-DAY GLOBAL BUDGET OF MERCURY (Mg yrPRESENT-DAY GLOBAL BUDGET OF MERCURY (Mg yr-1-1))

Anthropogenic(mostly coal): 2200

Hg0 Hg2+

Natural (ore): 500Re-emission:1500

OCEAN MIXED LAYER (0-150 m)

DEEP OCEAN

ATMOSPHERE 5200Atmospheric lifetimeof mercury ~ 1 year

SOILS 1,200,000

98% 2%INVENTORIES in MgFluxes in Mg yr-1

11,000

216,000

Burial400

DepositionLand 2200Ocean 2000

Re-emission2000

SHIP DATA FOR MERCURY vs. LATITUDESHIP DATA FOR MERCURY vs. LATITUDEWeak latitudinal gradient is indicative of long lifetime

Lamborg et al. [2002]

MERCURY RECORD FROM ICE CORE (WYOMING)MERCURY RECORD FROM ICE CORE (WYOMING)

GEOS-CHEM SIMULATION OF ATMOSPHERIC MERCURYGEOS-CHEM SIMULATION OF ATMOSPHERIC MERCURY

+

…again illustrating the northern mid-latitudes pollution belt

Circles indicate long-term observations

SOURCE ATTRIBUTION SOURCE ATTRIBUTION OF DEPOSITED OF DEPOSITED

MERCURY IN U.S.MERCURY IN U.S.[Seigneur et al., 2004][Seigneur et al., 2004]

“Natural” includes reemitted mercury – legacy of past century of anthropogenic emissions!