1

Neil Wheeler, Ken Craig, and Steve ReidSonoma Technology, Inc.

Petaluma, CA

Presented to:CCOS Technical Committee

Sacramento, CAOctober 1, 2008

905030

Central California Ozone Study (CCOS) Improvement of Air Quality

Model Aloft Performance

Final Presentation 05-2CCOS

2

Overview

• Introduction

• Analyses

• Summary and Conclusions

• Recommendations

3

Background

• Progress towards attaining the NAAQS• Model Performance Issues• Episodes

• July 8-15, 1999 – Limited data aloft• July 29 - August 2, 2000 – High ozone aloft • September 16-21, 2000 – Normal ozone aloft

• Roles• MM5 Meteorological Modeling – NOAA• Meteorological and Air Quality Modeling - ARB• Air Quality Model Improvement – STI

4

Performance Hypotheses (1 of 2)

1. Emission inventories are biased low

2. Wildfires are not adequately represented

3. Boundary concentrations are inaccurately specified• Lack of data• Transport from Asia• Transport from Southern California• Tropopause folding events (stratospheric ozone intrusion)

4. Grid resolution is insufficient

5. Meteorological models are unable to capture stagnant air conditions caused by terrain blocking of the flow

5

Performance Hypotheses (2 of 2)

6. Vertical mixing of pollutants in the planetary boundary layer is overestimated resulting in relatively clean air aloft being mixed downward

7. Mixing of high ozone surface air to the interior of the convective boundary layer is inadequate.

8. Recirculation of upslope flow from the Sierra Nevada Mountains and Coastal Range over the Central Valley are inadequately represented.

9. Chemical mechanisms underestimate ozone production efficiency at low precursor concentrations

10.Dispersion in areas of significant and steep terrain in central California is treated inadequately.

11.Photolysis rates at higher elevations are treated inadequately

6

Improvement Process

• Characterization

• Evaluation

• Identification of model performance issues

• Diagnosis of the possible causes

• Correction

• Reevaluation

7

Daily Weather Maps

8

Meteorological Soundings (1 of 2)

9

Meteorological Soundings (2 of 2)

10

Trajectories

11

Descriptive Statistics

Region 8: San Joaquin Valley South, 7/31/2000

-20

0

20

40

60

80

100

120

SHA OLD BGS BAC EDS MCKI ARV MCS

Site Code

O3

(pp

b)

12

Surface NMHC Plots

13

Structure of Air Quality Aloft

• Aircraft Spirals

• Aircraft Traverses

• Ozonesondes

• Hydrocarbons

14

Aircraft Spirals

37.0N 120.1W 35.9N 19.5W

15

Aircraft Traverses

16

Ozonesondes (1 of 2)

17

Ozonesondes (2 of 2)

Trinidad Head

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

0 20 40 60 80 100

Ozone (ppbV)

He

igh

t (m

) 07/21/00

08/01/00

08/30/00

09/21/00

18

Hydrocarbons AloftNMHC

0

10

20

30

40

50

60

70

80

90

100

7/30/2000 12:00:00AM

7/30/2000 12:00:00PM

7/31/2000 12:00:00AM

7/31/2000 12:00:00PM

8/1/2000 12:00:00AM

8/1/2000 12:00:00PM

8/2/2000 12:00:00AM

Date-Time

pp

bC

19

Aloft HMHC Plots

20

Transport Statistics

RWP

MM5

21

Time Series PlotsEpisode1, EDS, Surface Gases

0

24

48

72

96

120

144

168

192

216

240

7/2

8 1

2:0

0 A

M

7/2

8 1

2:0

0 P

M

7/2

9 1

2:0

0 A

M

7/2

9 1

2:0

0 P

M

7/3

0 1

2:0

0 A

M

7/3

0 1

2:0

0 P

M

7/3

1 1

2:0

0 A

M

7/3

1 1

2:0

0 P

M

8/1

12:0

0 A

M

8/1

12:0

0 P

M

8/2

12:0

0 A

M

8/2

12:0

0 P

M

8/3

12:0

0 A

M

8/3

12:0

0 P

M

8/4

12:0

0 A

M

Date

O3

, N

O, N

O2

(p

pb

)

0

1

2

3

4

5

6

7

8

9

10

O3

NO

NO2

NOY

CO

Episode1, EDS, Winds

0

3

6

9

12

15

18

21

24

7/2

8 1

2:0

0 A

M

7/2

8 1

2:0

0 P

M

7/2

9 1

2:0

0 A

M

7/2

9 1

2:0

0 P

M

7/3

0 1

2:0

0 A

M

7/3

0 1

2:0

0 P

M

7/3

1 1

2:0

0 A

M

7/3

1 1

2:0

0 P

M

8/1

12:0

0 A

M

8/1

12:0

0 P

M

8/2

12:0

0 A

M

8/2

12:0

0 P

M

8/3

12:0

0 A

M

8/3

12:0

0 P

M

8/4

12:0

0 A

M

Date

Ve

cto

r W

ind

Sp

ee

d (

m/s

)

0

45

90

135

180

225

270

315

360

VEC

RES

Episode1, EDS (T), BSE (MH)

0

5

10

15

20

25

30

35

40

45

50

7/2

8 1

2:0

0 A

M

7/2

8 1

2:0

0 P

M

7/2

9 1

2:0

0 A

M

7/2

9 1

2:0

0 P

M

7/3

0 1

2:0

0 A

M

7/3

0 1

2:0

0 P

M

7/3

1 1

2:0

0 A

M

7/3

1 1

2:0

0 P

M

8/1

12

:00

AM

8/1

12

:00

PM

8/2

12

:00

AM

8/2

12

:00

PM

8/3

12

:00

AM

8/3

12

:00

PM

8/4

12

:00

AM

Date

De

gre

es

(C

)

0

200

400

600

800

1000

1200

1400

1600

1800

2000

Mix

ing

He

igh

t (

m)

AMB

MH

22

Mixing Depth Growth

bkf

0

100

200

300

400

500

600

700

7/ 30/ 2005 7/ 31/ 2005 8/ 1/ 2005 8/ 2/ 2005 8/ 3/ 2005

tmr

0

100

200

300

400

500

600

700

7/ 30/ 2005 7/ 31/ 2005 8/ 1/ 2005 8/ 2/ 2005 8/ 3/ 2005

23

Vertical Wind Profiles

RWP

TC

MM5

A53

24

Mass Flux Analysis – TC MM5

25

Mass Flux Analysis – Hybrid

26

Morning Ozone Predictions

A53 TC0700 PDT August 1

27

Model Performance Aloft

O3 Aloft ComparisonA53 vs. Observations

y = 0.3719x + 46.749

R2 = 0.2956

0

50

100

150

200

250

0 50 100 150 200 250

Obs

A53

Pre

d

O3 Aloft ComparisonTC vs. Observations

y = 0.3951x + 47.744

R2 = 0.4367

0

50

100

150

200

250

0 50 100 150 200 250

Obs

TC

Pre

d

28

Ozone Correlation by Level

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

3 4 5 6 7 8 9 10 11 12 17

CAMx layer

R-s

qu

are

d

0.9 km

1.8 km

7.5 km

0.25 km

29

Meteorological Modeling Issues

• MM5

• CALMET-MM5 Hybrid

• Mapping to CAMx vertical structure

• Evaluation of winds above 2-km

• Mechanical mixing at night

30

Improvements

• Meteorology (MM5)

• Vertical Structure

• Nighttime Mixing

• Photolysis Rates

• Wildfire Emissions

• Anthropogenic and Biogenic Emissions

• Emphasis on July-August 2000 Episode

31

Evaluation Process

• Aloft Model Performance for Ozone• Hourly Averages for Grid Cells• Scatter Plots• Correlation

• Inert Tracer Simulations for Meteorology Changes

• Vis5D Animations of Ozone & Inert Tracers• Difference Plots• Ground Level Model Performance

32

Results

• Meteorology (MM5)

• Vertical Structure

• Nighttime Mixing

• Photolysis Rates

• Wildfire Emissions (partial)

• Emission Inventory Improvements (incomplete)

33

Meteorology

• NOAA’s “Best” Simulation• Not our requested simulation

–No hourly average output–Interpolated observation FDDA

• Results• Not much change in performance or flow

structure• Vertically interpolated observation nudging

did not obscure or significantly weaken return flows

34

Winds

Original New

35

Vertical Structure

• CAMx and MM5 layers matched in first 2 km

• 38 Layers

• Results• Better vertical resolution of winds in CAMx• Little change in net circulation

• Transport through boundaries aloft

36

Ozone Transport Through Boundaries

1.3 km 5.0 km

Monday, July 31, 2000

37

Nighttime Mixing Issues

• Surface roughness and friction velocity

• Fire emissions

• Recirculation

38

Nighttime Mixing

• Mechanical Mixing Model• Neutral and stable: van Ulden and Holtslag

(1985)

• TKE profiles• Near neutral : Zhang et al. (1996) • Stable: Lenschow et al. (1988)

39

TKE Profiles

TKE (stable conditions)

0

50

100

150

200

250

300

350

0.000 0.500 1.000 1.500 2.000 2.500

TKE (m2/s2)

Hei

gh

t (m

)

TKE (neutral conditions)

0

50

100

150

200

250

300

350

0 0.5 1 1.5 2 2.5 3

TKE (m2/s2)

Hei

gh

t (m

)

TKE (unstablestable conditions)

0

50

100

150

200

250

300

350

TKE (m2/s2)

Hei

gh

t (m

)

40

Alternate Photolysis Rates

• Modified CAMx 4.31 to read in terrain elevations and calculate layer heights above mean sea level (MSL)

• April 2006 release of TUV 4.01 allows levels to be specified

• Pseudo-spherical two-stream delta-Eddington scheme

• Increased number of photolysis levels in CAMx from 11 to 27

Case photorig: Photolysis rates used in TC simulation

Case newphot: Terrain-corrected photolysis rates

Spatial Differences – New Photolysis

42

Wildfire Emissions

• Stack parameters and plume rise• Plume rise:

2 km (intended) vs. > 10 km calculated in CAMx• Default stack parameters: T=295 ˚K and V=4 m/s

• Smoldering emissions• Fire effects models • Dominate VOC emissions from fires• Diurnal variations• Post burn smoldering• Issues with establishing defensible profiles

43

Wild Fire Tracers

16 km

44

Wild Fire Emissions Research

• JFSP Matrix

• JFSP FT-CSA

• USFS-FERA FCCS

• NASA Vertical Analysis

• JFSP SEMIP

45

EI Improvements

• Acquired draft improved EI for Emissions Reconciliation Project

• Initial review indicated the changes were not large enough, based on past sensitivity studies, to warrant a separate simulation

• Emissions Reconciliation Project indicated additional investigation warranted

46

Modeling Domain

• CAMx Simulations• Ozone mass flux analyses • Transported ozone and tracers through

southern and western boundaries

• SJVAQS 1990

• Animations

47

September 2000 Episode

• MM5 temperature bias

• No new MM5 simulations

• Biogenic emissions in southern SJV

• Underestimation of TNMOC:NOX emission ratios in southern SJV

48

High Ozone Day TemperaturesEpisode2, PLR, Surface Gases

0

24

48

72

96

120

144

168

192

216

240

9/15

12:

00 A

M

9/15

12:

00 P

M

9/16

12:

00 A

M

9/16

12:

00 P

M

9/17

12:

00 A

M

9/17

12:

00 P

M

9/18

12:

00 A

M

9/18

12:

00 P

M

9/19

12:

00 A

M

9/19

12:

00 P

M

9/20

12:

00 A

M

9/20

12:

00 P

M

9/21

12:

00 A

M

9/21

12:

00 P

M

9/22

12:

00 A

M

9/22

12:

00 P

M

9/23

12:

00 A

M

Date

O3

, NO

, NO

2 (

pp

b)

0

1

2

3

4

5

6

7

8

9

10

CO

(p

pm

)

O3

NO

NO2

NOY

CO

Episode2, PLR (T), FAT (MH)

0

5

10

15

20

25

30

35

40

45

50

9/15

12:

00 A

M

9/15

12:

00 P

M

9/16

12:

00 A

M

9/16

12:

00 P

M

9/17

12:

00 A

M

9/17

12:

00 P

M

9/18

12:

00 A

M

9/18

12:

00 P

M

9/19

12:

00 A

M

9/19

12:

00 P

M

9/20

12:

00 A

M

9/20

12:

00 P

M

9/21

12:

00 A

M

9/21

12:

00 P

M

9/22

12:

00 A

M

9/22

12:

00 P

M

9/23

12:

00 A

M

Date

De

gre

es (C

)

0

200

400

600

800

1000

1200

1400

1600

1800

2000

Mix

ing

Hei

gh

t (m

)

AMB

MH

49

Maximum Predicted Temperatures

September 19, 2000

50

Summary

• Air Quality Aloft

• Transport Processes

• Production Processes

• Fires

• Emissions

• Meteorological Model Biases

• Modeling Domain

51

Hypothesis Evaluation (1 of 2)

1. Emission inventories are biased low

2. Wildfires are not adequately represented

3. Boundary concentrations are inaccurately specified • Lack of data • Transport from Asia X• Transport from southern California • Tropopause folding events (stratospheric ozone intrusion) X

4. Grid resolution is insufficient ±

5. Meteorological models are unable to capture stagnant air conditions caused by terrain blocking of the flow X

52

Hypothesis Evaluation (2 of 2)

6. Vertical mixing of pollutants in the planetary boundary layer is overestimated resulting in relatively clean air aloft being mixed downward X

7. Mixing of high ozone surface air to the interior of the convective boundary layer is inadequate X

8. Recirculation of upslope flow from the Sierra Nevada Mountains and Coastal Range over the Central Valley are inadequately represented ±

9. Chemical mechanisms underestimate ozone production efficiency at low precursor concentrations ±

10.Dispersion in areas of significant and steep terrain in central California is treated inadequately ±

11.Photolysis rates at higher elevations are treated inadequately ±

53

Recommendations

• Implement new emission inventory

• Nest CCOS domain within a larger regional 12-km domain

• Continue to monitor research on fire emission in the Fire Sciences community