Transcript NASA Air Quality Applied Sciences Team (AQAST)

Use of satellite and suborbital observations
to constrain North American methane emissions
in the Carbon Monitoring System
Daniel Jacob (PI), Steven Wofsy (Co-I), Kevin Wecht,
Alex Turner, Greg Santoni, Melissa Sulprizio
Harvard University
Vivienne Payne (Co-I), Kevin Bowman (Co-I), Meemong Lee (Co-I), John Worden
NASA JPL
Importance of methane for the Carbon Monitoring System
• Present-day emission-based forcing of methane is 0.95 W m-2 (IPCC AR5)
• Climate impact of methane is comparable to CO2 over 20-year horizon
• Methane is a low-hanging fruit for climate policy
• Natural gas and hydrofracking are changing US sources
• Methane is a central piece of the President’s Climate Action Plan
Building a methane monitoring system for N America
integrated into the CMS
EDGAR emission
Inventory for methane
Can we use satellites together with suborbital observations of methane
to monitor methane emissions on the continental scale and test/improve
emission inventories in a manner useful to stakeholders?
Methane bottom-up emission inventories for N. America:
EDGAR 4.2 (anthropogenic), LPJ (wetlands)
N American totals in Tg a-1
Surface/aircraft studies suggest that these emissions are too low by ~factor 2
Methane observing system in North America
Satellites
AIRS, TES, IASI
Thermal IR
GOSAT
3-day, sparse
SCIAMACHY
6-day
Shortwave IR
2002
Suborbital
2006
2009
TROPOMI
1-day
20015
GCIRI
geo
2018
1/2ox2/3o grid of GEOS-Chem chemical transport model (CTM)
INTEX-A
CalNex
SEAC4RS
High-resolution inverse analysis system
for quantifying methane emissions in North America
EDGAR 4.2 + LPJ
a priori bottom-up emissions
Observations
GEOS-Chem CTM and its adjoint
1/2ox2/3o over N. America
nested in 4ox5o global domain
Bayesian
inversion
Validation
Verification
Optimized emissions
at 1/2ox2/3o resolution
The same CMS inverse analysis system is used at JPL for CO2 (K. Bowman, PI)
Optimization of state vector
for adjoint inversion of SCIAMACHY data
Optimal clustering of 1/2ox2/3o gridsquares
Native resolution
1000 clusters
Correction factor to bottom-up emissions
Optimized US anthropogenic emissions (Tg a-1)
SCIAMACHY data cannot constrain
emissions at 1/2ox2/3o resolution;
use 1000 optimally selected clusters
34
posterior cost function
28
1
Kevin Wecht, Harvard
10
100
1000
10,000
Number of clusters in inversion
North American methane emission estimates
optimized by SCIAMACHY + INTEX-A data (Jul-Aug 2004)
SCIAMACHY column methane mixing ratio
Correction factors to a priori emissions
1000 clusters
1700
15
10
5
ppb
1800
EDGAR v4.2 26.6
US anthropogenic emissions (Tg a-1)
EPA
28.3
This work
32.7
0
Livestock
Oil & Gas
Landfills
Coal Mining
Other
Livestock emissions are underestimated by EPA, oil/gas emissions are not
Wecht et al., in prep.
GOSAT methane column mixing ratios, Oct 2009-2010
Retrieval from U. Leicester
Inversion of GOSAT Oct 2009-2010 methane
Correction factors to prior emissions (EDGAR 4.2 + LPJ)
Nested inversion
with 1/2ox2/3o resolution
Alex Turner, Harvard
Next step: clustering of emissions in the inversion, use of ACOS data
Testing the information content of satellite data
with CalNex inversion of methane emissions
CalNex observations
GEOS-Chem w/EDGAR v4.2
Correction factors to EDGAR
(analytical inversion)
May-Jun
2010
Emisssions, Tg a-1
S. Wofsy (Harvard)
3.5
3
2.5
2
1.5
1
0.5
0
1800
ppb
2000
0.1
1
3
2x underestimate
of livestock emissions
CA Air Resources Board
EDGAR v4.2
Santoni et al. Lagrangian (STILT) inversion
GEOS-Chem inversion
State of California
Los Angeles Basin
Wecht et al., in prep.
GOSAT observations are too sparse
to spatially resolve California emissions
GOSAT data (CalNex period))
Correction factors to methane emissions from inversion
GOSAT (CalNex period)
GOSAT (1 year)
Each point =
1-10 observations
0.5
Degrees of Freedom for Signal (DOFS)
in inversion of methane emissions
1.5
CalNex
aircraft data
GOSAT
(CalNex)
GOSAT
(1 year)
15
1.2
2.8
Wecht et al., in prep.
TROPOMI and GCIRI constrain state-level methane emissions
better than a dedicated aircraft mission
Correction factors to EDGAR v4.2 a priori emissions from a 1-year OSSE
TROPOMI (global daily coverage)
GCIRI (geostationary 1-h return coverage)
0.2
1
A priori
DOFS
California emissions (Tg a-1)
1.9
5
CalNex TROPOMI
GCIRI TROPOMI+GCIRI
15
17
23
26
3.1
3.0
3.0
3.0
Wecht et al., in prep.
Working with stakeholders at the US state level
State-by-state analysis of SCIAMACHY correction factors to EDGARv4.2 emissions
with Iowa Dept. of Natural Resources (Marnie Stein)
State emissions computed w/EPA tools too low by x3.5;
now investigating EPA livestock emission factors
Hog manure?
0
1
2
correction factor
with New York Attorney General Office (John Marschilok)
State-computed emissions too high by x0.6,
reflects overestimate of gas/waste/landfill emissions
Large EDGAR source from gas+landfills
is just not there
Melissa Sulprizio and Kevin Wecht, Harvard