GLOBAL MODELS OF ATMOSPHERIC COMPOSITIONGLOBAL MODELS OF ATMOSPHERIC COMPOSITION

Daniel J. JacobDaniel J. JacobHarvard UniversityHarvard University

HOW TO MODEL ATMOSPHERIC COMPOSITION?HOW TO MODEL ATMOSPHERIC COMPOSITION?Solve continuity equation for chemical mixing ratios Ci(xx, t)

Fires Landbiosphere

Humanactivity

Lightning

Ocean Volcanoes

Transport

Eulerian form:

ii i i

CC P L

t

U

Lagrangian form:

ii i

dCP L

dt

U = wind vector

Pi = local source

of chemical i

Li = local sink

ChemistryAerosol microphysics

EULERIAN MODELS PARTITION ATMOSPHERIC DOMAIN EULERIAN MODELS PARTITION ATMOSPHERIC DOMAIN INTO GRIDBOXESINTO GRIDBOXES

Solve continuity equation for individual gridboxes

• Detailed chemical/aerosol models can presently afford -106 gridboxes

• In global models, this implies a horizontal resolution of ~ 1o (~100 km) in horizontal and ~ 1 km in vertical

This discretizes the continuity equation in space

• Chemical Transport Models (CTMs) use external meteorological data as input• General Circulation Models (GCMs) compute their own meteorological fields

OPERATOR SPLITTING IN EULERIAN MODELSOPERATOR SPLITTING IN EULERIAN MODELS

i i i

TRANSPORT LOCAL

C C dC

t t dt

… and integrate each process separately over discrete time steps:

( ) (Local)•(Transport) ( )i o i oC t t C t

• Split the continuity equation into contributions from transport and local terms:

Transport advection, convection:

Local chemistry, emission, deposition, aerosol processes:

(

ii

TRANSPORT

ii

LOCAL

dCC

dt

dCP

dt

U

) ( )iLC C

These operators can be split further:• split transport into 1-D advective and turbulent transport for x, y, z (usually necessary)• split local into chemistry, emissions, deposition (usually not necessary)

Reduces dimensionality of problem

SPLITTING THE TRANSPORT OPERATORSPLITTING THE TRANSPORT OPERATOR

• Wind velocity UU has turbulent fluctuations over time step t:( ) '( )t t U U U

Time-averagedcomponent(resolved)

Fluctuating component(stochastic)

1( )i i i

xx

C C Cu K

t x x x

• Further split transport in x, y, and z to reduce dimensionality. In x direction:

( , , )u v wU

• Split transport into advection (mean wind) and turbulent components:

1ii i

CC C

t

U K air density

turbulent diffusion matrix

K

advection turbulence (1st-order closure)

advectionoperator

turbulentoperator

SOLVING THE EULERIAN SOLVING THE EULERIAN ADVECTION EQUATIONADVECTION EQUATION

• Equation is conservative: need to avoid diffusion or dispersion of features. Also need mass conservation, stability, positivity…

• All schemes involve finite difference approximation of derivatives : order of approximation → accuracy of solution

• Classic schemes: leapfrog, Lax-Wendroff, Crank-Nicholson, upwind, moments…

• Stability requires Courant number ut/x < 1 … limits size of time step

• Addressing other requirements (e.g., positivity) introduces non-linearity in advection scheme

i iC Cu

t x

VERTICAL TURBULENT TRANSPORT (BUOYANCY)VERTICAL TURBULENT TRANSPORT (BUOYANCY)

Convective cloud(0.1-100 km)

Model grid scale

Modelverticallevels updraft

entrainment

downdraft

detrainment

Wet convection is subgrid scale in global models and must be treated as a vertical mass exchange separate from transport by grid-scale winds.

Need info on convective mass fluxes from the model meteorological driver.

• generally dominates over mean vertical advection• K-diffusion OK for dry convection in boundary layer (small eddies)• Deeper (wet) convection requires non-local convective parameterization

LOCAL (CHEMISTRY) OPERATOR:LOCAL (CHEMISTRY) OPERATOR:solves ODE system for solves ODE system for n n interacting speciesinteracting species

1,i n

1( ) ( ) ( ,... )ii i n

dCP L C C

dt C C C

System is typically “stiff” (lifetimes range over many orders of magnitude)→ implicit solution method is necessary.

• Simplest method: backward Euler. Transform into system of n algebraic equations with n unknowns

( ) ( )( ( )) ( ( )) 1,i o i o

i o i o

C t t C tP t t L t t i n

t

C C

( )ot tC

Solve e.g., by Newton’s method. Backward Euler is stable, mass-conserving, flexible (can use other constraints such as steady-state, chemical family closure, etc… in lieu of Ct ) But it is expensive. Most 3-D models use higher-order implicit schemes such as the Gear method.

For each species

SPECIFIC ISSUES FOR AEROSOL CONCENTRATIONSSPECIFIC ISSUES FOR AEROSOL CONCENTRATIONS

• A given aerosol particle is characterized by its size, shape, phases, and chemical composition – large number of variables!

• Measures of aerosol concentrations must be given in some integral form, by summing over all particles present in a given air volume that have a certain property

• If evolution of the size distribution is not resolved, continuity equation for aerosol species can be applied in same way as for gases

• Simulating the evolution of the aerosol size distribution requires inclusion of nucleation/growth/coagulation terms in Pi and Li, and size characterization either through size bins or moments.

Typical aerosol size distributionsby volume

nucleation

condensationcoagulation

LAGRANGIAN APPROACH: TRACK TRANSPORT OF LAGRANGIAN APPROACH: TRACK TRANSPORT OF POINTS IN MODEL DOMAIN (NO GRID)POINTS IN MODEL DOMAIN (NO GRID)

Ut

U’t

• Transport large number of points with trajectories from input meteorological data base (U) + random turbulent component (U’) over time steps t

• Points have mass but no volume

• Determine local concentrations as the number of points within a given volume

• Nonlinear chemistry requires Eulerian mapping at every time step (semi-Lagrangian)

PROS over Eulerian models:• no Courant number restrictions• no numerical diffusion/dispersion• easily track air parcel histories• invertible with respect to time

CONS:• need very large # points for statistics• inhomogeneous representation of domain• convection is poorly represented• nonlinear chemistry is problematic

positionto

positionto+t

LAGRANGIAN RECEPTOR-ORIENTED MODELINGLAGRANGIAN RECEPTOR-ORIENTED MODELINGRun Lagrangian model backward from receptor location, with points released at receptor location only

• Efficient cost-effective quantification of source influence distribution on receptor (“footprint”)

• Enables inversion of source influences by the adjoint method (backward model is the adjoint of the Lagrangian forward model)

EMBEDDING LAGRANGIAN PLUMES IN EULERIAN MODELSEMBEDDING LAGRANGIAN PLUMES IN EULERIAN MODELS

Release puffs from point sources and transport them along trajectories, allowing them to gradually dilute by turbulent mixing (“Gaussian plume”) until they reach the Eulerian grid size at which point they mix into the gridbox

• Advantages: resolve subgrid ‘hot spots’ and associated nonlinear processes (chemistry, aerosol growth) within plume• Difference with Lagrangian approach is that (1) puff has volume as well as mass, (2) turbulence is deterministic (Gaussian spread) rather than stochastic

S. California fire plumes,Oct. 25 2004

GEOS-Chem GLOBAL 3-D CHEMICAL TRANSPORT MODELGEOS-Chem GLOBAL 3-D CHEMICAL TRANSPORT MODEL

• Solves 3-D continuity equations on global Eulerian grid using NASA Goddard Earth Observing System (GEOS) assimilated meteorological data (1985-present) or GISS GCM output (paleo and future climate)• Horizontal resolution 1ox1o to 4ox5o, 48-72 vertical layers• Used by ~30 groups around the world for wide range of atmospheric composition problems: aerosols, oxidants, carbon, mercury, isotopes…

Illustrate here with Harvard work on tropospheric ozone

OZONE: “GOOD UP HIGH, BAD NEARBY”

Nitrogen oxide radicals; NOx = NO + NO2

Sources: combustion, soils, lightningVolatile organic compounds (VOCs)

MethaneSources: wetlands, livestock, natural gas…Non-methane VOCs (NMVOCs)Sources: vegetation, combustion

Carbon monoxide (CO)Sources: combustion, VOC oxidation

Troposphericozone

precursors

RADICAL CYCLE CONTROLLING TROPOSPHERIC OH RADICAL CYCLE CONTROLLING TROPOSPHERIC OH AND OZONE CONCENTRATIONSAND OZONE CONCENTRATIONS

O3

O2 h

O3

OH HO2

h, H2O

Deposition

NO

H2O2

CO, VOCs

NO2

h

STRATOSPHERE

TROPOSPHERE

8-18 km

SURFACE

GEOS-Chem simulation for tropospheric ozone includes 120 coupled speciesto describe HOx-NOx-VOC-aerosol chemistry

global sources/sinksin Tg y-1

4300

4000

700

400

Climatology of observed ozone at 400 hPa in July from ozonesondes and MOZAIC aircraft (circles) and corresponding GEOS-Chem model results for 1997 (contours).

GEOS-Chem tropospheric ozone columns for July 1997.

GLOBAL DISTRIBUTION OF TROPOSPHERIC OZONEGLOBAL DISTRIBUTION OF TROPOSPHERIC OZONE

Li et al., JGR [2001]

COMPARISON TO TES SATELLITE OBSERVATIONS COMPARISON TO TES SATELLITE OBSERVATIONS IN MIDDLE TROPOSPHEREIN MIDDLE TROPOSPHERE

Zhang et al. [2006]

averagingkernels

(July 2005)

TES ozone and CO observations in July 2005 at 618 hPaTES ozone and CO observations in July 2005 at 618 hPa

TES observations of ozone-CO correlations test GEOS-Chem simulation of ozone continental outflow

North America

Asia

Zhang et al., 2006

GEOS-Chem GLOBAL BUDGET OF TROPOSPHERIC OZONEGEOS-Chem GLOBAL 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,

Tg y-1

4300

1600

Chem loss in troposphere,

Tg y-1

4000

1600Transport from stratosphere,

Tg y-1

400

400

Deposition,

Tg y-1700

400Burden, Tg 360

230

Lifetime, days 28

42

Present-day Preindustrial

IPCC RADIATIVE FORCING ESTIMATE FOR TROPOSPHERIC IPCC RADIATIVE FORCING ESTIMATE FOR TROPOSPHERIC OZONE (0.35 W mOZONE (0.35 W m-2-2) RELIES ON GLOBAL MODELS) RELIES ON GLOBAL MODELS

Preindustrialozone models

}

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

…but these underestimate the observed rise in ozone over the 20th century

RADIATIVE FORCING BY TROPOSPHERIC OZONE RADIATIVE FORCING BY TROPOSPHERIC OZONE COULD THUS BE MUCH LARGER THAN IPCC VALUECOULD THUS BE MUCH LARGER THAN IPCC VALUE

Standard model:

F = 0.44 W m-2

“Adjusted” model

(lightning and soil NOx decreased,

biogenic hydrocarbons increased):

F = 0.80 W m-2

Global simulation of late 19th century ozone observations [Mickley et al., 2001]

IMPLICATION OF RISING BACKGROUNDIMPLICATION OF RISING BACKGROUND FOR MEETING AIR QUALITY STANDARDS FOR MEETING AIR QUALITY STANDARDS

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.)

Shutting down N. American anthropogenic emissions in GEOS-Chem reduces frequency of European exceedances of 55 ppbv standard by 20%

The U.S. EPA defines a “policy-relevant background” (PRB) The U.S. EPA defines a “policy-relevant background” (PRB) as the ozone concentration that would be present in U.S. surface air as the ozone concentration that would be present in U.S. surface air

in the absence of N. American anthropogenic emissionsin the absence of N. American anthropogenic emissions

(1) Standard simulation; include all sources

(2) Set U.S. or N. American anthropogenic emissions to zero infer policy-relevant background

(3) Set global anthropogenic emissions to zero estimate natural background

Difference between (1) and (2) regional pollution

Difference between (2) and (3) background enhancement from hemispheric pollution

• This background cannot be directly observed, must be estimated from models• Because chemistry is strongly nonlinear, sensitivity simulations are necessary

Summer 1995 afternoon (1-5 p.m.) ozone in surface air over the U.S.Summer 1995 afternoon (1-5 p.m.) ozone in surface air over the U.S.

Observations

r = 0.66, bias=5 ppbv

GEOS-CHEMstandard simulation

Fiore et al. [2002]

Examine a clean site: Examine a clean site: Voyageurs National Park, MinnesotaVoyageurs National Park, Minnesota(May-June 2001)(May-June 2001)

CASTNet observationsModelBackgroundNatural O3 levelStratospheric

+

*

Hemisphericpollution

Regionalpollution}

}

Background: 15-36 ppbvNatural level: 9-23 ppbvStratosphere: < 7 ppbv

High-O3 events: dominated by regional pollution; minor stratospheric influence (~2 ppbv)

regional pollution

hemispheric pollution

X

Fiore et al. [2003]

Compiling daily afternoon (1-5 p.m. mean) surface ozone from all Compiling daily afternoon (1-5 p.m. mean) surface ozone from all CASTNet rural sites for March-October 2001: CASTNet rural sites for March-October 2001:

Policy-relevant background ozone is typically 20-35 ppbv Policy-relevant background ozone is typically 20-35 ppbv P

rob

abili

ty p

pb

v-1

CASTNet sites

GEOS-Chem Model at CASTNet

Natural 18±5 ppbvGEOS-Chem

PRB 26±7 ppbvGEOS-Chem

PRB 29±9 ppbv MOZART-2

Fiore et al., JGR 2003

EFFECT OF 2000-2050 CLIMATE CHANGE ON U.S. OZONE POLLUTIONEFFECT OF 2000-2050 CLIMATE CHANGE ON U.S. OZONE POLLUTION

2000 2050 climate - 2000

Wu et al. [2007]

Run GEOS-Chem driven by GISS GCM for present vs. 2050 climate

• Climate change decreases the background ozone because higher water vapor increases ozone loss;

• but it aggravates ozone pollution episodes due to less ventilation (fewer mid-latitudes cyclones), faster chemistry, higher biogenic VOC emissions

CONSTRAINING NOCONSTRAINING NOxx AND REACTIVE VOC EMISSIONS AND REACTIVE VOC EMISSIONS

WITH NOWITH NO22 AND FORMALDEHYDE (HCHO) MEASUREMENTS AND FORMALDEHYDE (HCHO) MEASUREMENTS

FROM SPACEFROM SPACE

Emission

NOh (420 nm)

O3, RO2

NO2

HNO3

1 day

NITROGEN OXIDES (NOx) VOLATILE ORGANIC COMPOUNDS (VOC)

Emission

VOC

OHHCHOh (340 nm)

hoursCO

hours

BOUNDARYLAYER

~ 2 km

Tropospheric NO2 column ~ ENOx

Tropospheric HCHO column ~ EVOC

Deposition

GOME: 320x40 km2

SCIAMACHY: 60x30 km2 OMI: 24x13 km2

TOP-DOWN CONSTRAINTS ON NOTOP-DOWN CONSTRAINTS ON NOxx EMISSION INVENTORIES EMISSION INVENTORIES

FROM OMI NOFROM OMI NO22 DATA INTERPRETED WITH GEOS-Chem DATA INTERPRETED WITH GEOS-Chem

Tropospheric NO2 (March 2006)

OMIobservations

GEOS-Chemwith EPA 1999 emissions

OMI – GEOS-Chem difference

Fitting OMI NO2 with GEOS-Chem requires• 25% decrease in power plant emissions• 30% increase in vehicle emissionsrelative to EPA 1999 official inventory

Boersma et al. [2007]

FORMALDEHYDE COLUMNS FROM OMI (Jun-Aug 2006):FORMALDEHYDE COLUMNS FROM OMI (Jun-Aug 2006): high values are due to biogenic isoprene (main reactive VOC) high values are due to biogenic isoprene (main reactive VOC)

OMIGEOS-Chem model w/best prior (MEGAN)

biogenic VOC emissions

MEGAN emission hot spots not substantiated by the OMI data

Millet et al. [2007]