VOC Lumping and Its Potential Impacts on Tropospheric Ozone Modeling Results

1
CMAQ Nucleation Algorithms and their impact on PM
modeling results in the Lower Fraser Valley
Helmut Roth, Weimin Jiang, Dazhong Yin, and Éric
Giroux
National Research Council of Canada Ottawa, Canada
2

• Introduction
• Nucleation Rates
• Impact on PM
• Conclusions future work

3
Introduction

• Episode Pacific 93 (1993 Jul 31 Aug 07)
• Emissions SMOKE, Giroux, É. et al
• Meteorology MM5, Yin, D. et al
• Domain
• Nucleation Algorithms

4
Horizontal domains
5
Nucleation parameterizations algorithms

• HK98 Harrington and Kreidenweis (1998)
• Fully coupled equations of 2-mode 2-moment model
• implemented in aero2
• KLPx Kulmala et al (1998)
• Equations of classical nucleation theory
• Implemented in aero2, aero3 KLP2, KLP3
• Differences in number distribution, density
• sulphuric acid water system

6
Nucleation rates modeling
7
1st layer domain-average number rates
8
1st layer domain average ASO4I rates
9
1st layer number rate statistics
all maxima occur in cell (17,8) Neptune Beach
10
Time-average, 1st layer number rate
11
Correlation at time of peak nucleation
r20.9 Aug 05 1300
12
Impact on PM concentrations model runs

• Common inputs
• Full gas phase calculations
• All aerosol processes
• Different nucleation algorithms
• NOOP HK98 KLP2 KLP3

13
Number concentration
PM2.5
i-mode
14
Time-average, 1st layer NUMATKN
15
Sulphate (ASO4) concentration
PM2.5
i-mode
16
Time-average, 1st layer ASO4I
17
Main Conclusions

• KLP2 and KLP3 results very similar
• KLPx had greatest individual cell rate.
• HK98 had greatest average rates.
• HK98 more responsive to SULAER fluctuations.
• HK98 increases number concentration over NOOP
  KLPx does not.
• No increase in PM2.5 SO4 mass over NOOP.

18
Future work

• Impact of nucleation algorithms on model
  performance using ambient number concentration
  data

19
Acknowledgements

• U.S. EPA CMAQ source code
• Environment Canada emissions and met data,
  ambient PM mass data
• Program of Energy Research and Development (PERD)
  in Canada funding support

20

• Thank you !

21
Time-average, 1st layer mass rate
22
(No Transcript)