Transcript Document

GLOBAL MODELS OF ATMOSPHERIC COMPOSITION
Daniel J. Jacob
Harvard University
HOW TO MODEL ATMOSPHERIC COMPOSITION?
Solve continuity equation for chemical mixing ratios Ci(x, t)
Lightning
U = wind vector
Ci
 U  Ci  Pi  Li
t
Pi = local source
Lagrangian form:
Transport
Chemistry
Aerosol microphysics
Volcanoes Fires
Eulerian form:
Li = local sink
dCi
 Pi  Li
dt
Human
Land
biosphere activity
of chemical i
Ocean
EULERIAN MODELS PARTITION ATMOSPHERIC DOMAIN
INTO GRIDBOXES
This discretizes the continuity equation in space
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
• Chemical Transport Models (CTMs) use external meteorological data as input
• General Circulation Models (GCMs) compute their own meteorological fields
OPERATOR SPLITTING IN EULERIAN MODELS
Reduces dimensionality of problem
• Split the continuity equation into contributions from transport and local terms:
Ci  Ci 
 dCi 



t  t TRANSPORT  dt  LOCAL
 dC 
Transport  advection, convection:  i 
  U  Ci
 dt TRANSPORT
Local  chemistry, emission, deposition, aerosol processes:
 dCi 
 Pi (C)  Li (C)
 dt 

 LOCAL
… and integrate each process separately over discrete time steps:
Ci (to  t )  (Local)•(Transport)  Ci (to )
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)
SPLITTING THE TRANSPORT OPERATOR
• Wind velocity
U
has turbulent fluctuations over time step t:
U(t )  U  U '(t )
Time-averaged
component
(resolved)
Fluctuating component
(stochastic)
• Split transport into advection (mean wind) and turbulent components:
Ci
1
 U  Ci    KCi
t

advection
  air density
K  turbulent diffusion matrix
turbulence (1st-order closure)
• Further split transport in x, y, and z to reduce dimensionality. In x direction:
Ci
Ci 1 
Ci
 u

( K xx
)
t
x  x
x
advection
operator
turbulent
operator
U  (u, v, w)
SOLVING THE EULERIAN
ADVECTION EQUATION
Ci
Ci
 u
t
x
• 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 ut/x < 1
… limits size of time step
• Addressing other requirements (e.g., positivity)
introduces non-linearity in advection scheme
VERTICAL TURBULENT TRANSPORT (BUOYANCY)
• 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
Convective cloud
(0.1-100 km)
detrainment
Model
vertical
levels
updraft
downdraft
entrainment
Model grid scale
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.
LOCAL (CHEMISTRY) OPERATOR:
solves ODE system for n interacting species
For each species
i 1, n
dC i
 Pi (C)  Li (C)
dt
C  (C1 ,...Cn )
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 C(t  t )
o
Ci (to  t )  Ci (to )
 Pi (C(to  t ))  Li (C(to  t ))
t
i  1, n 
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 C/t ). But it is expensive. Most 3-D models use
higher-order implicit schemes such as the Gear method.
SPECIFIC 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.
condensation
nucleation
coagulation
Typical aerosol
size distributions
by volume
LAGRANGIAN APPROACH: TRACK TRANSPORT OF
POINTS IN MODEL DOMAIN (NO GRID)
• 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
position
to+t
U’t
position
to
Ut
• 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
LAGRANGIAN RECEPTOR-ORIENTED MODELING
Run 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 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
S. California fire plumes,
Oct. 25 2004
• 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
GEOS-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, lightning
Tropospheric
ozone
precursors
Volatile organic compounds (VOCs)
Methane
Sources: wetlands, livestock, natural gas…
Non-methane VOCs (NMVOCs)
Sources: vegetation, combustion
Carbon monoxide (CO)
Sources: combustion, VOC oxidation
RADICAL CYCLE CONTROLLING TROPOSPHERIC OH
AND OZONE CONCENTRATIONS
GEOS-Chem simulation for tropospheric ozone includes 120 coupled species
to describe HOx-NOx-VOC-aerosol chemistry
O2
hn
O3
STRATOSPHERE
8-18 km
400
TROPOSPHERE
global sources/sinks
in Tg y-1
4300
O3
NO2
NO
OH
HO2
hn, H2O
4000
Deposition
hn
700
CO, VOCs
SURFACE
H2O2
GLOBAL DISTRIBUTION OF TROPOSPHERIC OZONE
Climatology of observed
ozone at 400 hPa in July
from
ozonesondes
and
MOZAIC aircraft (circles)
and corresponding GEOSChem model results for
1997 (contours).
GEOS-Chem tropospheric
ozone columns for July 1997.
Li et al., JGR [2001]
COMPARISON TO TES SATELLITE OBSERVATIONS
IN MIDDLE TROPOSPHERE
(July 2005)
averaging
kernels
Zhang et al. [2006]
TES ozone and CO observations in July 2005 at 618 hPa
North America
Asia
TES observations of ozone-CO correlations test GEOS-Chem simulation of
Zhang et al., 2006
ozone continental outflow
GEOS-Chem GLOBAL BUDGET OF TROPOSPHERIC OZONE
Present-day
Preindustrial
O2
Chem prod in
troposphere,
Tg y-1
Transport from
stratosphere,
Tg y-1
hn
O3
Burden, Tg
STRATOSPHERE
8-18 km
4300
1600
400
400
360
230
Chem loss in
troposphere,
Tg y-1
Deposition,
Tg y-1
Lifetime, days
4000
1600
700
400
28
42
TROPOSPHERE
hn
O3
Deposition
NO2
NO
OH
HO2
hn, H2O
CO, VOC
H2O2
IPCC RADIATIVE FORCING ESTIMATE FOR TROPOSPHERIC
OZONE (0.35 W m-2) RELIES ON GLOBAL MODELS
…but these underestimate the observed rise in ozone over the 20th century
Preindustrial
ozone models
}
Observations at mountain
sites in Europe
[Marenco et al., 1994]
RADIATIVE FORCING BY TROPOSPHERIC OZONE
COULD THUS BE MUCH LARGER THAN IPCC VALUE
Global simulation of late 19th century
ozone observations [Mickley et al., 2001]
Standard model:
F = 0.44 W m-2
“Adjusted” model
(lightning and soil NOx decreased,
biogenic hydrocarbons increased):
F = 0.80 W m-2
IMPLICATION OF RISING BACKGROUND
FOR MEETING AIR QUALITY STANDARDS
Europe AQS
(8-h avg.)
Europe AQS
(seasonal)
0
Preindustrial
ozone
background
20
40
U.S. AQS
(8-h avg.)
60
80
U.S. AQS
(1-h avg.)
100
120 ppbv
Present-day ozone
background at
northern midlatitudes
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)
as the ozone concentration that would be present in U.S. surface air
in the absence of N. American anthropogenic emissions
• This background cannot be directly observed, must be estimated from models
• Because chemistry is strongly nonlinear, sensitivity simulations are necessary
(1) Standard simulation; include all sources
(2) Set U.S. or N. American anthropogenic
emissions to zero a infer policy-relevant
background
(3) Set global anthropogenic emissions to zero a
estimate natural background
Difference between (1) and (2) a
regional pollution
Difference between (2) and (3) a background
enhancement from hemispheric pollution
Summer 1995 afternoon (1-5 p.m.) ozone in surface air over the U.S.
Observations
GEOS-CHEM
standard simulation
Fiore et al. [2002]
r = 0.66, bias=5 ppbv
CASTNet observations
Regional
Model
  pollution
Background
Hemispheric


+ Natural O3 level
pollution
X Stratospheric
*
Examine a clean site:
Voyageurs National Park, Minnesota
(May-June 2001)
}
}
High-O3 events:
dominated by regional
pollution;
minor stratospheric
influence (~2 ppbv)
regional
pollution
hemispheric
pollution
Fiore et al. [2003]
Background: 15-36 ppbv
Natural level: 9-23 ppbv
Stratosphere: < 7 ppbv
Probability ppbv-1
Compiling daily afternoon (1-5 p.m. mean) surface ozone from all
CASTNet rural sites for March-October 2001:
Policy-relevant background ozone is typically 20-35 ppbv
Natural 18±5 ppbv
GEOS-Chem
PRB 26±7 ppbv
GEOS-Chem
Fiore et al., JGR 2003
PRB 29±9 ppbv
MOZART-2
CASTNet sites
GEOS-Chem Model at CASTNet
EFFECT OF 2000-2050 CLIMATE CHANGE ON U.S. OZONE POLLUTION
Run GEOS-Chem driven by GISS GCM for present vs. 2050 climate
2000
2050 climate - 2000
• 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
Wu et al. [2007]
CONSTRAINING NOx AND REACTIVE VOC EMISSIONS
WITH NO2 AND FORMALDEHYDE (HCHO) MEASUREMENTS
FROM SPACE
GOME: 320x40 km2
SCIAMACHY: 60x30 km2
OMI: 24x13 km2
Tropospheric NO2 column ~ ENOx
Tropospheric HCHO column ~ EVOC
~ 2 km
BOUNDARY
LAYER
hn (420 nm)
hn (340 nm)
HCHO OH
CO
hours
hours
NO2
NO
O3, RO2
1 day
VOC
HNO3
Emission
NITROGEN OXIDES (NOx)
Deposition
Emission
VOLATILE ORGANIC COMPOUNDS (VOC)
TOP-DOWN CONSTRAINTS ON NOx EMISSION INVENTORIES
FROM OMI NO2 DATA INTERPRETED WITH GEOS-Chem
Tropospheric NO2 (March 2006)
Boersma et al. [2007]
OMI – GEOS-Chem difference
OMI
observations
GEOS-Chem
with EPA 1999
emissions
Fitting OMI NO2 with GEOS-Chem requires
• 25% decrease in power plant emissions
• 30% increase in vehicle emissions
relative to EPA 1999 official inventory
FORMALDEHYDE COLUMNS FROM OMI (Jun-Aug 2006):
high values are due to biogenic isoprene (main reactive VOC)
OMI
GEOS-Chem model w/best prior (MEGAN)
biogenic VOC emissions
MEGAN emission hot spots not substantiated by the OMI data
Millet et al. [2007]