CRGAQS: CAMx Sensitivity Results

1
CRGAQSCAMx Sensitivity Results

• Presentation to the
• Gorge Study Technical Team
• By
• ENVIRON International Corporation
• November 15, 2006

2
Todays Presentation

• Describe CAMx sensitivity simulations
• Model configurations
• Performance for PM and light scattering/extinction
  
• Show all August results first
• Then November results
• Next Steps

3
Model Configuration

• Episodes
• August 10-22, 2004
• Meteorology from MM5 Run 6, 36/12-km grids
• Flexi-nesting to the 4-km grid
• November 4-18, 2004
• Meteorology from MM5 Run 3, all grids
• Identified issues in Run 1
• Small temporal profile problem for fires
• No on-road vehicle ammonia emissions in 4-km grid

4
Model Configuration

• Review Run 1 configuration
• Maximize model speed
• Mechanism 4 CF (static 2-mode PM chemistry)
• Bott advection solver
• No PiG
• OMP parallel processing on Linux quad-CPU
• OBrien Kv profile with 0.1 m2/s minimum
• 10-day model spin-up period
• 36-km grid only first 8 days
• 36/12-km grid last 2 days

5
Model Configuration

• Run 2
• Fix known emission issues
• Find new fire emission problems
• Run 3
• Use CMAQ Kv profiles with 1.0 m2/s minimum
• Run 4
• Double NH3 inventory in 4/12-km grids
• Run 5
• Halve POAPEC fire emissions (Aug only)

6
Performance Evaluation
7
August Performance Evaluation
8
August Performance Evaluation
9
August Performance Evaluation
10
August Performance Evaluation
11
Run 2 Statistics
August NO3
August SO4
12
Run 2 Statistics
August NH4
August OC
13
Run 2 Statistics
August EC
August Primary Fine
14
Run 2 Statistics
August Primary Coarse
August Total PM2.5
15
Run 2 Statistics
August Total PM10
August Bscat
16
Dry Light Scattering
17
Predicted Organic Aerosol Components
18
Run 2 Results

• High SO4 and NO3 from incorrect fire emissions
• Gorge not affected performance OK for SO4 and
  NO3
• But NH3 is low
• OC shows extremely high localized peaks from the
  fire emissions (POA)
• OC over predictions in Gorge from biogenic SOA
• Much less impacts from other SOA components
• Under prediction of EC (especially at high
  observed values)
• PM2.5 tended to be too high
• PM10 was too low at IMPROVE sites
• Why apparently too high at Gorge sites?

19
Run 2 Results

• 24-hr dry Bscat at Gorge sites tended to be too
  high
• Hourly dry Bscat at Gorge monitors show over
  predicted spikes not observed by the
  nephelometers.
• Individual PM concentrations at Mt. Zion show
• Spikes in the FINE COARSE mass accounted for a
  majority of the Bscat
• The spikes corresponded well with the spikes in
  Bscat
• During the peak Bscat of 206 Mm-1 , 150 was from
  this group
• OC, whose maximum hourly concentration was 15
  ug/m3, contributed a maximum of 60 Mm-1 to Bscat

20
Run 3 Results
21
Run 3 Results

• Secondary PM changed little compared to Run 2
• Primary PM and Bscat showed small reductions as
  the enhanced mixing diluted the concentrations
• Gorge sites closest to Portland (Steigerwald, Mt.
  Zion, and Strunk Rd) had the largest reductions
  in Bscat
• Hourly time series also show lower spikes in the
  Bscat in Run 3 compared to Run 2

22
Run 4 Results
23
Run 4 Results

• Doubling NH3 emissions had no impact on SO4
• NO3 increases generally occurred in Puget Sound
  area
• Smaller NO3 increases extend across eastern WA
  and OR
• Daily Bscat changed less than 3 Mm-1 on all dates
  at the Gorge sites.
• Hourly Bscat was not significantly different from
  Run 2

24
Run 5 Results
25
Run 5 Results

• EC and OC were reduced the most in the Washington
  interior, and only slightly at the Gorge
  monitoring sites.
• Gorge EC and OC were reduced most during
  northerly winds (August 13)
• Monitors in the east showed the largest reduction
• Hourly Bscat show similar patterns between Run 2
  and Run 5
• Some slightly lower spikes, particularly at the
  eastern sites (like Towal Rd)

26
November Performance Evaluation
27
November Performance Evaluation
28
November Performance Evaluation
29
November Performance Evaluation
30
Run 2 Statistics
November NO3
November SO4
31
Run 2 Statistics
November NH4
November OC
32
Run 2 Statistics
November EC
November Primary Fine
33
Run 2 Statistics
November Primary Coarse
November Total PM2.5
34
Run 2 Statistics
November Total PM10
November Bext
35
Dry Light Scattering(west Gorge)
36
Dry Light Scattering (east Gorge)
37
Run 2 Results

• Nitrates were more abundant in November than in
  the August simulation
• SO4 and OC were highest near urban centers
  (Portland and Seattle), and along I-5
• Near the Gorge, SO4 was highest on November 12
  and 13
• SO4 at IMPROVE sites was well-predicted
• SO4 at Gorge monitors tended to be underpredicted
• NO3 performance was scattered
• OC, fine, and coarse matter had a few over
  predicted extremes
• PM2.5 at the FRM network was over predicted

38
Run 2 Results

• Hourly Bscat at western Gorge monitors is over
  predicted (large spikes)
• Bscat and PM time series at Mt. Zion show
• Spikes in OC correspond to the Bscat spikes
• The largest contribution to Bscat from OC
  4OCmax 475 300, which accounts for only
  half of the total computed Bscat
• Spikes in FINE and COARSE matter also match the
  Bscat spikes -- this adds up to 200 to Bscat at
  Mt. Zion
• Hourly Bscat at eastern Gorge sites is under
  predicted (miss big haze event)

39
Run 3 Results
40
Run 3 Results

• Enhanced mixing diluted concentrations of all
  species (esp. primary PM)
• Changes were much greater than in the August Kv
  sensitivity test
• Gorge sites closest to Portland showed the
  greatest reductions in Bscat spikes
• Steigerwald Bscat dropped nearly 300 Mm-1 on
  November 6 compared to Run 2 (OBrien Kv)
• No siginficant change at eastern sites

41
Run 4 Results
42
Run 4 Results
43
Run 4 Results

• Nitrates were much higher with doubling of
  ammonia emissions
• Western Gorge area showed the largest increases
  early in the episode (Nov 6-9)
• Sulfate increases were more spatially confined
• Bscat, PM2.5 and PM10 were slightly higher
• Bscat changes at Gorge sites were higher at the
  western sites
• However, changes were relatively insignificant

44
Moving Forward

• More sensitivity/diagnostic runs
• Revise fire emissions
• Reduce primary fine/coarse emissions in both
  episodes
• Check role of biogenic emissions in OC over
  predictions
• Identify observed components of high scattering
  at eastern Gorge sites