Comparison of CAMx and CMAQ PM2.5 Source Apportionment Estimates

1
Comparison of CAMx and CMAQ PM2.5Source
Apportionment Estimates

• Kirk Baker and Brian Timin
• U.S. Environmental Protection Agency,
• Research Triangle Park, NC
• Presented at the 2008 CMAS Conference

2
Background

• Photochemical model source apportionment is a
  useful tool to efficiently characterize source
  contribution to PM2.5
• Implemented particulate source apportionment in
  CMAQv4.6
• Compared the source apportionment results with
  other model system CAMx
• Existing inputs developed for Milwaukee pilot
  project used for comparison of source
  apportionment results

3
PPTM PSAT

• The Particle and Precursor Tagging Methodology
  (PPTM) has been implemented in CMAQ v4.6
• Particulate Source Apportionment Technology
  (PSAT) has been implemented in CAMx v4.5
• Tracks contribution to mercury and PM sulfate,
  nitrate, ammonium, secondary organic aerosol, and
  inert species
• Estimates contributions from emissions source
  groups, emissions source regions, and initial and
  boundary conditions to PM2.5 by adding duplicate
  model species for each contributing source
• These duplicate model species (tags) have the
  same properties and experience the same
  atmospheric processes as the bulk chemical
  species
• The tagged species are calculated using the
  regular model solver for processes like dry
  deposition and advection as bulk species
• Non-linear processes like gas and aqueous phase
  chemistry are solved for bulk species and then
  apportioned to the tagged species

4
PM2.5 Source Apportionment Modeling for Milwaukee
Pilot Project

• CAMx v4.5 and CMAQ v4.6
• 12 km modeling domain
• 4 months in 2002
• Jan, Apr, Jul, Oct
• Evaluating 24-hr average contributions from 11
  source regions, the rest of the modeling domain,
  boundary conditions
• Emissions processed separately for each source
  region

5
Source Regions
Region 12 All non-tagged areas in domain Region
13 Boundary conditions
6
Model Performance

• Daily 24-hr PM predictions at Milwaukee
  (550790026) and Waukesha (551330027) county STN
  monitors over all modeled days
• Model-Model estimates shown at right
• CMAQ tends to predict more nitrate than CAMx

7
Model Performance
CMAQ
CAMx
8
Contribution Estimation

• Evaluated contribution at Milwaukee (5) and
  Waukesha (1) monitors
• PM2.5 SO4NO3NH4POCEC
• Examined 1) top 10 days, 2) average over all
  days, and 3) compared daily estimates
• Days included in top 10 analysis Q16, Q26,
  Q30, Q43
• Contribution from 11 source regions (counties),
  ICBC, all other non-tagged sources
• Did not track SOA due to low model estimations
  and resource constraints

9
Total PM2.5 Contribution Estimation
10
24-hr Avg Total PM2.5 Contribution Estimation
Top 10 Days
CMAQ
CAMx
11
4-month average total PM2.5 contributions from
source areas 1-6
CAMx
Region 1 2
3 4 5
6
CMAQ
12
4-month average total PM2.5 contributions from
source areas 7-11
CAMx
Region 7 8
9 10
11
CMAQ
13
Distribution of 24-hr avg Contribution Estimations
14
24-hr avg Contributions estimated by CMAQ and
CAMx
15
24-hr avg Contributions estimated by CMAQ and
CAMx
16
24-hr avg Contributions estimated by CMAQ and
CAMx
17
24-hr avg Contributions estimated by CMAQ and
CAMx
18
Domain Maximum 24-hr avg Initial Condition
Contribution
19
Remarks

• CMAQ estimates more nitrate and as a result
  estimates larger nitrate contributions
• CMAQ seems to estimate larger local contributions
  from primarily emitted species
• Spatial extent of average contributions similar
  between models
• Average contributions over high model days very
  similar at the Milwaukee/Waukesha monitors
• Initial contributions drop out of model after 5-7
  days
• Would like to compare with CMAQ-DDM for future
  work

20
Acknowledgements

• Tom Braverman, US EPA
• ICF International (Sharon Douglas and Tom Myers)

21
JAN APR JUL OCT
Kenosha County 24-hr max contribution Sulfate
Nitrate Primary OC