Validation of CH4 surface emission using forward chemistry‐transport model
P. K. Patra, X. Xiong, C. Barnet, E. J. Dlugokencky,
U. Karin, K. Tsuboi, D. Worthy
18th International Emission Inventory Conference “Comprehensive Inventories – Leveraging Technology and Resources”
Baltimore, USA; 14‐16 April 2009
Acknowledgements: T. Nakazawa, T. Saino, JAMSTEC for funding
Research Institute for Global Change, JAMSTECYokohama, Japan
Introduction
Emission Inventory(EI) for MCF, CH4,
CO2, SF6…
Site-basedobservation
Process-based model
Remote-sensing products
ChemistryTransport
Model (CTM)
Atmospheric Observations
(site & satellite)
Transport ProcessesMeteorology
Radicals (OH, Cl)
Validation of Transport (SF6)
Validation of OH (CH3CCl3 - MCF)
Model-ObservationComparison of CH4
Multiple Tracers is key for Forward modeling and Validation research
Space‐Time scales
Hour Day Month Year
Loca
l R
egio
nal
Hem
isph
eric
Int
er-H
em
PBL
Synoptic/Frontal
Hadley Cell,ITCZ & SPCZMonsoon
Inter-hemisphericExchange
Connectin
g proce
sses
, e.g., M
eteoro
log
Atmospheric Observables, e.g., Concentration
Inpu
t par
amet
er, e
.g.,
Emis
sion
Inve
ntor
y
y
ACTM transport validation: inter‐hemispheric exchange time (τex)
Patra et al., ACP, 2009
En = NH emissionEs = SH emissionCn = NH concentrationCs = SH concentration
ACTM framework for CH4 simulation (Patra et al., JMSJ, in acceptance)
• CCSR/NIES/FRCGC AGCM‐based CTM (ACTM) run at resolution T42 L67 (top 90km)
• NCEP‐2 reanalysis meteorology (U,V,T nudged)
• Hadley Center Sea‐Surface Temperature & Sea‐Ice Cover
• CH4 chemistry (Sander et al., JPL Pub. 06‐2, 2006) as:
• All the radicals are taken from CHASER/STRAT (Sudo et al., Takigawa et al.) models at monthly (or hourly) intervals
CH4+O1D Products (KO1D =1.5×10–10)
CH4+OH CH3 + H2O (KOH= 2.45×10–12 exp(-1775/T)CH4+Cl CH3 + HCl (KCl=7.3×10–12 exp(-1280/T)CH4+hυ Products (wavelength dependent; not considered here)
Validation of CHASER+STRAT OH using MCF
The comparison for the recent times raises a couple of questions:1.Is MCF emission decreasing since 2000 or so. If yes, at what rate and what is present day emission magnitude2.Is the OH production in CHASER is higher than the real values? If so, at which latitudes, longitudes or heights
Surface flux types and annual budget of CH4
Mikaloff Fletcher et al., GBC, 2004
Range Estim.Reported by IPCC [2001]
92‐237
25‐10080‐11520‐2023‐5575‐109
35‐7310‐155‐10500‐600
450‐51040‐4610‐30
Patra et al., JMSJ, 2009, in acceptance
CH4 emission, sinks, and
concentration
CH4 lifetime and budgets
Instantaneous CH4 Lifetime = 1/[KOH∙OH + KO1D∙O1D + KCl∙Cl]
(useful for understanding the dominance of dynamics vs. chemistry on variability)
Atmospheric Lifetime = burden/loss(Prather et al., IPCC, 2001)
CH4 L. T. = 4999 Tg/580 Tg yr‐1 = 8.62 years
Estimates Atmospheric Lifetime
References
IPCC TAR 8.4 Prather et al.
IPCC FAR (prescribed)
8.67±1.32(m#26)8.45±0.38 (m#12)
Stevenson et al., JGR, 2006
This work (full model)
8.62 Patra et al., JMSJ, 2009
CH4 Measurement Sites – can we track emissions?(~50 used here; >100s are in operation in 2007)
Contributing Institutes: 1. NOAA/ESRL, 2. FEA, Germany, 3. JMA, Japan, 4. EC, Canada, 5. NIWA
CH4 Latitudinal gradients: seasonal and longitudinal variations
1700
1750
1800
1850
1900
1950ACTM: JANACTM: APRACTM: JULACTM: OCT
1700
1750
1800
1850
1900
1950
CH
4 mix
ing
ratio
[pp
b]
ESRL: JANESRL: APRESRL: JULESRL: OCT
90S 60S 30S 0 30N 60N 90N1700
1750
1800
1850
1900
1950
90S 60S 30S 0 30N 60N 90N
a. East Pacific b. America
c. Atlantic d. Ind./Europe
e. Australia-Asia f. West Pacific
TD
F
EIC
SMO C
HR
ML
O MID TH
D
CB
A BR
W
RPB
KE
YB
MW
NW
RW
SA
AL
T
LE
F
TD
F
ASC
IZO
MH
D
STM
ZE
P
SYO
CR
Z
SEY M
KN H
UN
SCH Z
GT
SYO
SPO
CG
O
CFA
BK
T
WL
G
UU
M
AR
H BH
D
GM
I
MN
MY
ON
TA
P
RY
O
SHM
SPO
SPO
SPO
SPO
SPO
FRD
Seasonal cycle in the extra tropics/high latitudes is
controlled by chemical loss and surface em
ission
CH4 Seasonal Cycles in the tropics – produced by chemical loss and dynamics (monsoon)
Preliminary comparison with AIRS/Aqua satellite retrievals
AIRSACTM‐E1ACTM‐E2
80οE‐ 110οE, 20οN ‐35οN; 250‐350 mb
AIRS Precision : ~20 ppb AIRS – ACTM = 20 ppb
Source: NOAA/N
ESDIS/STA
R
E1 : 80 Tg‐CH4 Rice emisE2 : 40 Tg‐CH4 Rice emis
CH4 Synoptic variations – arising from synoptic meteorology and concentration gradients
0
20
40
60OBSACTM
-20
0
20
40
-40
0
40
80
CH
4 Syn
optic
var
iatio
n [p
pb]
-200
-100
0
100
200
-80
0
80
160
J F M A M J J A S O N D J F M A M J J A S O N D J F M A M J J A S O N D-40
0
40
80
(a) CGO
(c) YON
(e) NGL
(f) BRW
(d) AMY
(b) MLO
2002 | 2003 | 2004
GSN
CD
L
AM
Y
BR
W
TH
D
AL
T
FRD
ZSF
RY
O
NG
L
SMO
DE
U
MN
M
YO
N`
ML
O
CO
I
RPB
ZG
T
SCH
MH
D
IZO
CG
O
HA
T
0
0.2
0.4
0.6
0.8
Cor
rela
tion
(r)
2002-2005DJFMAMJJASON
0
0.5
1
1.5
2
NSD
(un
itles
s)
(a)
(b)
Measure of
Synopticpeakshape /phase
Measure of
Synoptic peakheight
(σobs/σmod)Ideal~1.0
CH4 diurnal cycles:caused by local emission
and PBL cycle
Obs. Mod.
0
50
100
150Daily Obs
Daily ACTM
Monthly Obs.
Monthly ACTM
0
50
100
150
CH
4 Diu
rnal
am
plitu
de [
ppb]
2002 2003 20040
200
400
600
(a) FRD
(b) SCH
(c) AMY
0 24 0 24 0 24 0 24 0 24 UTC
Mon
thly
-mea
n di
urna
l CH
4 cy
cle
(ppb
)
Conclusions
• ACTM CH4 simulations have been optimized for a combinations of fluxes, radicals and transport– Examining seasonal cycles of IHGs at a ‘set’ of stations helped to prepare flux‐
types combinations– Very low (~5ppb) CH4 longitudinal gradients between CRZ/CGO/EIC can be
tracked– Seasonal cycle, synoptic variability and diurnal cycle; all needed to be checked
for testing bottom‐up fluxes
• The ACTM‐AIRS comparison in the upper troposphere region reveal– A reversal of CH4 seasonal cycle at about 400‐300 mb height– Role of CH4 transport by Asian monsoon during high surface emission period
• Both transport and chemistry components in ACTM have been validated independently using two different chemical tracers, SF6 and MCF, respectively
Outlook• Reduction in radical (e.g., OH) uncertainty by independent
methodology– Can the MCF emission be available at more confidence?
e.g., global/country total emission inventory for the users?
– Time dependent gridded distribution of emissions
• Updating chemical (e.g., CH4) emission distribution and strength– A concerted effort?, e.g., EDGAR gives annual mean emission due to
rice cultivation
– IAVs in wetland or other natural emissions? Validation strategy
• Increased use of remote sensing data for long‐lived gases– Some issues persist with the product quality; more validation and
comparison would help probing the causes
– More dedicated satellite sensors, e.g., GOSAT, new‐OCO?