Krish Vijayaraghavan, Prakash Karamchandani

1
Application of CMAQ-APT tothe Central California
Ozone Study

• Krish Vijayaraghavan, Prakash Karamchandani
• Christian Seigneur
• AER
• San Ramon, CA
• 3rd Annual CMAS Models-3 ConferenceOctober
  18-20, 2004
• Chapel Hill, NC

2
Overview

• Limitations of 3-D grid modeling
• CMAQ-APT Plume-in-grid (PiG) air quality model
• State-of-the science treatment of stack plumes at
  the sub-grid scale
• 3-D grid host model - CMAQ
• Reactive plume model SCICHEM
• Impact of PiG treatment on ozone and HNO3
  concentrations

3
Interface between CMAQ and SCICHEM
Domain, grid information geophysical
data meteorological data deposition velocities
Control File
Emissions, IC/BC
I/O API
I/O API
chemical concentrations
CMAQ
SCICHEM
Point source emissions
I/O API
I/O API
Standard SCICHEM output
Output
Output puff information
Dump puffs
chemical concentrations
Puff diagnostics (process analysis)
4
Improvements to CMAQ-APT

• Optional treatment for the effects of building
  downwash on plume rise and dispersion (PRIME)
• Incorporation of CMAQ code updates released in
  September 2003 (version 4.3)
• Incorporation of modifications in SCICHEM version
  1601 (January 2004 release)
• Young Boris chemistry solver

5
Application to Central California

• Central California Ozone Study (CCOS)
• Ozone episode July 30 to August 1, 2000
• Study domain
• 185 x 185 grid cells
• Horizontal grid resolution of 4 km
• 20 layers from surface to tropopause (surface
  layer 30 m)

6
Features of July/August 2000 CCOS episode

• Observed peak ozone during the modeling period
  was 134 ppb at San Andreas station on August 1,
  2000
• Prevailing winds from the coast to the Central
  Valley
• Wildfires in the southeast (Tulare County near
  the Sierra Nevada)

7
Model Inputs

• Meteorology driven by MM5
• CMAQ emissions, initial and boundary conditions
  from CAMx files (ARB)
• 3-D gridded emissions using SMOKE plume rise
  processor
• Ten largest NOx emitting plants (with 56 stacks)
  selected for plume-in-grid (PiG) treatment

8
PiG Sources
Top 10 NOx emissions

• Pittsburg power plant
• (16 Mg/day NOx)
• Riverside Cement
• California Cement
• Moss Landing power plant
• Martinez refinery
• Hanson Cement
• Unknown
• Portland Cement
• IMC Chemicals
• Contra Costa power plant

Total 101 Mg/day (4 of domain-wide NOx
emissions)
9
Simulations

• CMAQ base simulation
• All emissions in domain in 3-D gridded format
• CMAQ background simulation
• 3-D gridded emissions without PiG sources
• CMAQ-APT simulation
• 3-D gridded emissions other than PiG sources
• PiG sources treated separately with SCICHEM

10
Extended San Francisco Bay Area
11
CMAQ Base Surface Ozone 3 p.m. PDT on July 30,
2000
12
(Base Background) Surface Ozone 3 p.m. PDT on
July 30, 2000
13
(APT Base) Surface Ozone 3 p.m. PDT on July
30, 2000
14
Evolution of Plume Chemistry
3
2
Long-range Plume Dispersion
Early Plume Dispersion
Mid-range Plume Dispersion
1
Reduced VOC/NOx/O3 chemistry acid formation
from OH and NO3/N2O5 chemistry
NO/NO2/O3 chemistry
Full VOC/NOx/O3 chemistry acid and O3 formation
15
CMAQ Base Surface HNO3 3 p.m. PDT on July 31,
2000
16
(Base Background) Surface HNO3 3 p.m. PDT on
July 31, 2000
17
(APT Base) Surface HNO3 3 p.m. PDT on July 31,
2000
18
Comparison of CMAQ-APT results in CCOS and NARSTO

• CCOS
• APT produces up to 10 ppb lower ozone than Base
• and up to 1.5 ppb lower HNO3
• NARSTO
• APT produces up to 40 ppb lower ozone than Base
• and up to 24 ppb lower HNO3
• (Karamchandani et al., J. Geophys. Res.,107,
  4403, 2002)
• NOx emissions from PiG sources are about 50 times
  higher in NARSTO than in CCOS

19
Conclusions

• CMAQ-APT applied to July/August 2000 CCOS episode
• O3 concentrations using APT show both decrements
  (up to 10 ppb) and increments (up to 6 ppb) with
  respect to the base
• The VOC vs. NOx limited nature of the background
  environment explains the differences in O3
  production and destruction between the APT and
  base results

20
Conclusions

• Surface HNO3 concentrations are up to 1.5 ppb
  (about 10) lower in the CCOS APT simulation and
  24 ppb lower in the NARSTO APT simulation,
  compared to the base cases
• Effect of PiG treatment on HNO3 is important for
  PM nitrate and regional haze modeling.
• PiG treatment for PM (CMAQ-MADRID-APT) is
  currently being tested (Karamchandani et al.,
  AWMA, October 2004)

21
Acknowledgements

• California Energy Commission
• California Air Resources Board
• EPRI
• Titan/ARAP