Comparison of CMAQ and CAMx for an Annual Simulation over the South Coast Air Basin

1
Comparison of CMAQ and CAMx for an Annual
Simulation over the South Coast Air Basin

• Jin Lu1, Kathleen Fahey1, Bruce Jackson1, Paul
  Livingstone1, Daniel Chau1, and Ajith Kaduwela1,2

1Planning and Technical Support Division Air
Resources Board California Environmental
Protection Agency 2Department of Land, Air and
Water Resources University of California at Davis
2
DISCLAIMER

• The content of this presentation does not
  necessarily reflect the views and policies of the
  California Air Resources Board, nor does mention
  of trade names or commercial products constitute
  endorsement or recommendation for use.

3
South Coast Air Basin
One of the most polluted regions and the most
studied region in the world.
4
South Coast Air Basin

• 11680 5 km2 grid cells

5
Simulation Setup for 2005

• MM5 (South Coast AQMD)
• Emissions CARB/SCAQMD
• Measurements
• Rubidoux Riverside (RUB) and Los Angeles (LA)
• PM species concentrations STN database
• PM2.5, NO3, SO4, NH4, OC, EC
• Hourly gaseous species concentrations AIRS/AQS
  database
• O3, CO, NO, NO2, SO2
• Model predictions
• PM - 9 grid cells 24-hr average concentration
• O3, - 1 grid cell hourly Meteorology

6
Model Configuration
CMAQ Version 4.6 vs. CAMx Version 4.2
7
Species Mapping CMAQ vs. CAMx
8
Outline of Comparison and Analysis

• Scatter Plots
• CMAQ vs. Observation
• CAMx vs. Observation
• CMAQ vs. CAMx
• PM species, annual plots at Rubidoux
• O3, quarterly plots at Rubidoux
• Box Plots
• PM species, weekly
• Gas species, 24-hour
• Statistics
• FBIAS - Fractional Bias FERR Fractional Error
• NMB - Normalized Mean Bias NME Normalized Mean
  Error

9
PM2.5 Scatter Plots
10
PM2.5 Box Plots
11
PM Species Scatter Plots
12
O3 Scatter Plots
13
O3 Box Plots
14
Statistics Gaseous Species
15
Statistics PM species
16
Statistics at Rubidoux
17
Conclusions

• Annual simulations over South Coast Air Basin
  performed with both CMAQ and CAMx
• CMAQ and CAMx model performance evaluated at
  Roubidoux and Los Angeles
• Model predictions correlate well with each other
• Model predictions correlate reasonably well with
  observations

18
International Aerosol Modeling Algorithms (I AM
A) Conference

• Bi-annual in-depth look at the inner-workings of
  aerosol models (complements the annual CMAS
  Meeting)
• Wednesday, December 5th Friday, December 7th,
  2007 (just before the AGU meeting in San
  Francisco)
• University of California at Davis
• Impressive list of U.S. and international invited
  speakers
• Information at http//airquality.ucdavis.edu/