MOZART Development, Evaluation, and Applications
at GFDL
MOZART Users’ MeetingAugust 17, 2005
Boulder, CO
Arlene M. Fiore Larry W. Horowitz
Outline: MOZART Development, Evaluation, and Applications at GFDL
• Surface Ozone Bias over the United States- Comparison with observations (EPA AQS; CASTNet)- Sensitivity- Policy-relevant background
• Evaluation with 2004 ICARTT observations*
• Vertically distributed biomass burning
• Trends (historical, future) in ozone and aerosols
• Methane control for climate and air quality– 1990-2004 CMDL CH4
*
*Special thanks to George Milly, the ICARTT Science Team, CMDL
MOZART-2 Comparison with AIRS: July 2001 1-5 p.m. Surface O3 (ppbv)
Mean Bias = 24±10 ppbv; r2 = 0.50
Fires Landbiosphere
Humanactivity
Ocean
NORTH AMERICA
Lightning
“ Background” air
Outside naturalinfluences
Long-range transportof pollution
Processes Contributing to Surface Ozone over North America
stratosphere
lightning
“POLICY RELEVANT BACKGROUND” (PRB) OZONE:Ozone concentrations that would exist in the absence of anthropogenic emissions from North America
X
Daily afternoon (1-5 p.m. mean) surface ozone from all CASTNet sites for March-October 2001:
PRB ozone over the U.S. is typically 20-35 ppbv
CASTNet sites
GEOS-CHEM Model
PRB 26±7 ppbvGEOS-CHEM PRB 29±9 ppbv
MOZART-2
Both models predict consistent PRB range despite large surface O3 bias in MOZART-2
MOZART-2 Model
MOZART-2 bias associated with domestic ozone production
MODEL – OBSERVED
Daily mean 1-5 p.m. June 1 – Aug 31 at CASTNet stations
-50 0 50 100 150
Bac
kgro
und
O3
Nor
th A
mer
ican
O3
80
60
40
20
0
200
150
100
50
0
-50 0 50 100 150
Slope = -0.14; intercept=25;r=-0.38
Slope = 0.81; intercept=33;r=0.65
Substantial O3 sensitivity to the uncertain fate (and yield) of organic isoprene nitrates
OH RO2 NO(very fast) NO2
High-NOx
Isoprene nitratesSink for NOx?ISOPRENE
O3
Change in July mean 1-5 p.m. surface O3 when isoprene nitrates (at 12% yield) act as a NOx sink
4-12 ppbv impact!
ppbv
MOZART-2
Fiore et al., JGR, 2005
MOZART-4 Fully Interactive Base case +21 ppbv
0
-3
1.5
2
-1
-1
-1.5
-2
-3
-7
-8 -3 2
Change in eastern U.S. surface ozone bias due to sensitivity simulations
Base case = simulation with isop. nitrates as a NOx sinkMOZART-2 Base case +19 ppbv
O3 deposition velocities*1.5PAN, NO2 dep vels. = O3
Including alkyl nitrate formationSYNOZNo isop. peroxide recyclingDaytime PBL increased to 2 km
Xactive phot/emis (not drydep)All Xactive; no clouds in phot.
Daytime PBL increased to 1 km
Xactive drydep/emis (not phot)
Outline: MOZART Development, Evaluation, and Applications at GFDL
• Surface Ozone Bias over the United States- Comparison with observations (EPA AQS; CASTNet)- Sensitivity- Policy-relevant background
• Evaluation with 2004 ICARTT observations
• Vertically distributed biomass burning
• Trends (historical, future) in ozone and aerosols
• Methane control for climate and air quality– 1990-2004 CMDL CH4
*
COMPARISON WITH ICARTT : Mean % BiasMOZART-4 (preliminary version) NCEP T62, 1999 NEI
vs. All INTEX DC-8 observations June-Aug 2004
-200-150-100-50
050
100150200
CO
_UC
IGC
OZO
NE
PAN
HN
O3_
CIT
CH
4_U
CIG
C
CH
2O_N
CA
RC
H2O
_UR
I
HC
HO
_GIT
H2O
2_U
RI
H2O
2_C
IT
HN
O3_
UN
H
CH
3OO
H_U
RI
CO
_MIX
ING
_RA
TIO
ISO
PREN
E
AC
ETA
LDEH
YDE
AC
ETO
NE
MEK NO
X
OH
HO
2
ETH
AN
E
PRO
PAN
E
HO
2NO
2
%
OZONE(ppb)
CO(ppb) H2O2_CIT
(ppt)
HNO3_CIT(ppt)
PAN(ppt)
NOx(ppt)
Altit
ude
(km
)
Campaign Mean Vertical ProfilesModel vs. INTEX DC-8 Observations
Ozone Chemical RegimeModel vs. INTEX DC-8 Observations (day; <2km; east of 100 °W)
Model more HOx-rich (i.e., NOx-sensitive) and shows a stronger HOx-NOx correlation than observed.
HO2 vs. NO2
Outline: MOZART Development, Evaluation, and Applications at GFDL
• Surface Ozone Bias over the United States- Comparison with observations (EPA AQS; CASTNet)- Sensitivity- Policy-relevant background
• Evaluation with 2004 ICARTT observations
• Vertically distributed biomass burning
• Trends (historical, future) in ozone and aerosols
• Methane control for climate and air quality– 1990-2004 CMDL CH4
*
Vertically Distributed Biomass Burning (BMB) Emissions
1. IPCC AR-4 BASE CASE (met year 2000) -- monthly 1997-2002 mean van der Werf emissions -- levels with tops at 0.1, 0.5, 1, 2, 3, and 6 km
2. ICARTT Summer 2004 -- daily emissions from Rynda Hudman & Solene
Turquety, Harvard-- distributed up to 4 km, with 50% below 1 km
Change in SON composite max* CO concentrations (ppb) (Vertically distributed) – (All at surface)
300 hPa
750 hPa 995 hPa
500 hPa
*Composite max = daily max per grid point
increases just above the boundary layer decreases at surface and higher altitudes;
interplay btw emissions and convection?
Change in Tropospheric O3 Columns (DU)Composite Seasonal Maxima*
(Vertically distributed) - (All BMB emissions at surface)
JJA
SON
DJF
MAM
Maximum impact ~10% near source region*Composite max = daily max per grid point
-3 -2 -1 0 1 2 3
Outline: MOZART Development, Evaluation, and Applications at GFDL
• Surface Ozone Bias over the United States- Comparison with observations (EPA AQS; CASTNet)- Sensitivity- Policy-relevant background
• Evaluation with 2004 ICARTT observations*
• Vertically distributed biomass burning
• Trends (historical, future) in ozone and aerosols
• Methane control for climate and air quality– 1990-2004 CMDL CH4
*
NOx Emissions
0
20
40
60
80
100
120
SO2 Emissions
0
50
100
150
200
250
BC Emissions
0
5
10
15
20
25
30
0 0 0 0 0 0 0 0 0 0 0 0 0
1860
1880
1900
1920
1940
1960
1980
2000
2020
2040
2060
2080
2100
NOxEmissions
(Tg N)
SO2Emissions(Tg SO2)
BC Emissions
(Tg C)
120
100
80
60
40
20
250
200
150
100
50
0
25
20
15
10
5
0
O3 Burden
0
5
10
15
20
25
30
35
40
45
50
SO4 Burden
0
1
2
3
4
5
6
7
8
0 0 0 0 0 0 0 0 0 0 0 0 0
BC Burden
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1860
1880
1900
1920
1940
1960
1980
2000
2020
2040
2060
2080
2100
BC Burden(Tg C)
SO4 Burden(Tg S)
O3 Burden(DU)
5045403530252015105876543210
1.61.41.21.0.80.60.40.20
Historical A2 A1B B1
Emissiontrends inMOZART-2and resultingtroposphericburdens,used to drive GFDL climatemodelsimulationsfor IPCC
[Horow itz, in prep.]
1860: Mean=24.1 DU 2000: Mean= 34.0 DU
A2 2100: Mean 45.4 DU
Trends in Tropospheric O3 Columns
[Horow itz, in prep.]
First step; next we’ll examineclimate impacts on chemistrywith GFDL chemistry-climate model
Ozone Budgets in IPCC-AR4 from 19 Tropospheric Chemistry Models for Base Year 2000
PROD LOSS DEP STRAT
Ozo
ne (T
g yr
-1)
Budgets from Stevenson et al., 2005
X 19-Model MeanMOZART-2MOZART-4
X MOZECHGEOS-CHEM
Scenarios for 2030:• Current Legislation (CLE)• Maximum Feasible Reductions • SRES A2• Climate change (CLE emissions)
Emissions for 2000:• EDGAR 3.2• GFED 1997-2002 mean
for biomass burning
Outline: MOZART Development, Evaluation, and Applications at GFDL
• Surface Ozone Bias over the United States- Comparison with observations (EPA AQS; CASTNet)- Sensitivity- Policy-relevant background
• Evaluation with 2004 ICARTT observations
• Vertically distributed biomass burning
• Trends (historical, future) in ozone and aerosols
• Methane control for climate and air quality– 1990-2004 CMDL CH4
*
MOZART-2 Methane Study
Motivation: Methane controls benefit global air quality and climate by lowering background tropospheric O3
Question: Are prior results from steady-state simulations with uniform, fixed CH4 concentrations directly relevant to real-world emission controls?
Approach: Multi-decadal transient simulationsReduce global anthrop. CH4 emissions by 40%:
(1) All in Asia(2) Everywhere in the globe
(All simulations use 1990-2004 T62 NCEP winds, recycled as needed)
Methane Emissions in EDGAR inventory: early 1990s (Tg CH4 yr-1)
Anthropogenic: 248
Based on values in the literature, we increased biogenic CH4 emissions by 60 Tg
Energy,landfills,wastewaterRuminants
Rice
BiomassBurning
Ocean
Biogenic
9593
6086
204
10 Total: 548
Seasonal cycle, inter-annualvariability,increasing trend largelycaptured at remote sites
Underestimatespost-1998; indicating emissionsincrease?
MODEL CH4 CMDL CH4
1990 1992 1994 1996 1998 2000 2002
1780
1760
1740
1720
1700
1760
1740
1720
1700
1680
1780176017401720170016801660
High Bias at high northern latitudes
Low bias at high southern latitudes1990 1992 1994 1996 1998 2000 2002
MODEL CH4 CMDL CH4
1720
1700
1680
1660
1640
1800
1780
1760
1740
1720
1900
1850
1800
1750
1990 1992 1994 1996 1998 2000 2002
1990 1992 1994 1996 1998 2000 2002
Low Bias at high southern latitudes
Inter-hemispheric gradient too high
050
100150200250300350400
1990 2000 2010 2020
Change inMethaneConcentration(ppbv)
0
5
10
15
20
25
30
35
1990 1995 2000 2005 2010 2015 2020
Tg
CO
Decrease in Tropospheric CO Burden
140014501500155016001650170017501800
1990 2000 2010 2020
Met
hane
(ppb
v)
Standard
-40% Anthrop.Emissions
Transient simulations with EDGAR 1990 emissions, beginning 1990:(1) Standard (2) 40% decrease in global anthrop. emissions
(18% of total CH4 emissions)
0123456789
1990 2000 2010 2020
Tg O
zone
Difference (STD -GLOB40)
Global surface CH4 conc. Decrease in global surface CH4 conc.(standard – 40% anth. emis. decrease
Decrease in Tropospheric O3 Burden
Difference (STD -GLOB40)
CLIMATE IMPACTS: Change in July 2000 Trop. O3 Columns (to 200 hPa)
40% decrease in global anthrop.CH4 emissions
Zero CH4 emissions from Asia(= 40% decrease in global anthrop.)
No Asia – (40% global decrease)
Tropospheric O3 column response is independent of CH4 emission location except for small (~10%) local changes
Dobson Units
DU
U.S. Surface Afternoon Ozone Response in Summer also independent of methane emission location
MEAN DIFFERENCE MAX DIFFERENCE(Composite max daily afternoon mean JJA) NO ASIAN CH4
GLOBAL 40% DECREASE IN ANTHROP. CH4
Stronger sensitivity in NOx-saturated regions (Los Angeles),partially due to local ozone production from methane
Summary: MOZART Development, Evaluation, and Applications at GFDL
• Surface Ozone Bias over the United States– Typically 15-20 ppbv; sensitive to local chemistry
• Evaluation with 2004 ICARTT observations– Generally good; many species too high in boundary layer
• Vertically distributed biomass burning– Small mean effect, up to ~10% episodically
• Trends (historical, future) in ozone and aerosols– Past increases, future increases under some scenarios– First step towards studying chemistry-climate interactions– MOZART-2 near ensemble mean in IPCC 2030 comparisons
• Methane control for climate and air quality– Good agreement btw transient runs and remote surface obs.– Nearing steady-state after 30 years (~3 e-folding lifetimes) – 40% anthrop. CH4 decrease -9 Tg O3; -(1-3) ppbv U.S. JJA – Ozone response largely independent of CH4 source location