motivation model simulations for california have usually underestimated ozone concentrations aloft

32
Model Aloft Study Ola Persson, Jian-Wen Bao, Sara Michelson Jim Wilczak NOAA/ Environmental Technology Laboratory

Upload: kevlyn

Post on 12-Jan-2016

37 views

Category:

Documents


0 download

DESCRIPTION

Model Aloft Study Ola Persson, Jian-Wen Bao, Sara Michelson Jim Wilczak NOAA/ Environmental Technology Laboratory. Motivation Model simulations for California have usually underestimated ozone concentrations aloft CCOS IOP simulations have under-predicted surface ozone concentrations - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Model Aloft Study

Ola Persson, Jian-Wen Bao,Sara Michelson

Jim Wilczak

NOAA/ Environmental Technology Laboratory

Page 2: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

• Motivation– Model simulations for California have usually

underestimated ozone concentrations aloft– CCOS IOP simulations have under-predicted surface

ozone concentrations

• Hypotheses– Improper IC and BC’s– Insufficient model resolution– Too strong winds in stagnant conditions– Too weak of mixing of PBL O3 to layers above– Too weak recirculation of upslope flow back over CV– Errors in characterization of chemistry– Emissions

Page 3: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

NOAA & STI Components

• NOAA: O3 & meteorological fields– initial and boundary conditions– meteorological dependent O3 formation– mixing– transport

• STI: all chemical species– initial and boundary conditions– chemical mechanism– emissions

Page 4: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Planned Analysis (NOAA)

• MM5-CAMx and Hybrid comparisons for July-August and September 2000 IOP’s (utilizing ozone aloft data as well as meteorological obs)

• Produce new simulations with WRF-Chem model

• Run additional simulations with improved MM5-CAMx model.

Page 5: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

CAMx Experiments

• 'A60' MM5 without observational FDDA • 'A59' MM5 with observational FDDA• 'A53' meteorology from the CALMET/MM5

hybrid

• The two MM5-CAMx simulations used 40ppbv lateral and top boundary conditions

• Hybrid run used a truncated model to 5 km, with 70ppbv lateral and top boundary conditions

Page 6: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Ozone sonde evaluationJuly-August IOP

• Parlier (16 sondes)

• Granite Bay (12 sondes)

Page 7: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Parlier ozone sonde site

Page 8: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Granite Bay ozone sonde site

Page 9: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Year Day (2000)

Hei

ght (

m)

CAMx O3 concentration (ppb) at Parlier for exp. A59

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0 0 010 1020 2030 30 30

30

40 4050 5060 60

60

607070

70

7080

80

90110

80

80

70

50

80

80

Year Day (2000)

Hei

ght (

m)

sonde O3 concentration (ppb) at Parlier

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75

-75

-75

-65-65

-65

-65 -65

-55

-55

-55

-45

-45

-35-35

-35

-35

-25

-25-25

-25

-25

-25

-25

-15-5 5

5

-15

15

5

Year Day (2000)

Hei

ght (

m)

CAMx-sonde O3 concentration (ppb) at Parlier for exp. A59

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

80

Year Day (2000)

Hei

ght (

m)

CAMx O3 concentration (ppb) at Parlier for exp. A53

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0 010 1020 20 2030 30

30

40 4050 5060 60

60

60

70

7080 90

80

90

80

50

8080

80

80

80

Year Day (2000)

Hei

ght (

m)

sonde O3 concentration (ppb) at Parlier

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75

-75

-65-65

-65

-65

-55

-55

-45

-45

-45

-35-35

-35

-35

-25

-25

-25

-15

-15-15

-15

-15

-5

-5

5

1525

-35

-15

15

-35

-5

15

Year Day (2000)

Hei

ght (

m)

CAMx-sonde O3 concentration (ppb) at Parlier for exp. A53

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

50

90

Year Day (2000)

Hei

ght (

m)

CAMx O3 concentration (ppb) at Parlier for exp. A60

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0 010 10 1020 20

20

30 30

30

40 405060 60 60

6060

70

70 80

80

90

80

808080

90

80

90

Year Day (2000)

Hei

ght (

m)

sonde O3 concentration (ppb) at Parlier

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75

-75

-75

-65

-65

-65

-55

-55

-45

-45

-35

-35

-35 -35

-25

-25

-25

-15

-15

-25-25

Year Day (2000)

Hei

ght (

m)

CAMx-sonde O3 concentration (ppb) at Parlier for exp. A60

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

MM5

MM5-FDDA

Sonde

Parlier

Hybrid80

Year Day (2000)

Hei

ght (

m)

CAMx O3 concentration (ppb) at Parlier for exp. A53

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0 010 1020 20 2030 30

30

40 4050 5060 60

60

60

70

7080 90

80

90

80

50

8080

80

80

80

Year Day (2000)

Hei

ght (

m)

sonde O3 concentration (ppb) at Parlier

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75

-75

-65-65

-65-65

-55

-55

-45

-45-45

-35-35

-35-35

-25

-25-25

-15

-15-15

-15-15

-5

-5

5

1525

-35

-15

15

-35

-5

15

Year Day (2000)

Hei

ght (

m)

CAMx-sonde O3 concentration (ppb) at Parlier for exp. A53

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

Page 10: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Year Day (2000)

Heig

ht (m

)

CAMx O3 concentration (ppb) at Parlier for exp. A59

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0 0 010 1020 2030 30 30

30

40 4050 5060 60

60

607070

70

7080

80

90110

80

80

70

50

80

80

Year Day (2000)

Heig

ht (m

)

sonde O3 concentration (ppb) at Parlier

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75

-75

-75

-65-65

-65

-65 -65

-55

-55

-55

-45

-45

-35-35

-35

-35

-25

-25-25

-25

-25

-25

-25

-15-5 5

5

-15

15

5

Year Day (2000)

Heig

ht (m

)

CAMx-sonde O3 concentration (ppb) at Parlier for exp. A59

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

80

Year Day (2000)

Heig

ht (m

)

CAMx O3 concentration (ppb) at Parlier for exp. A53

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0 010 1020 20 2030 30

30

40 4050 5060 60

60

60

70

7080 90

80

90

80

50

8080

80

80

80

Year Day (2000)

Heig

ht (m

)

sonde O3 concentration (ppb) at Parlier

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75

-75

-65-65

-65

-65

-55

-55

-45

-45

-45

-35-35

-35

-35

-25

-25

-25

-15

-15-15

-15

-15

-5

-5

5

1525

-35

-15

15

-35

-5

15

Year Day (2000)

Heig

ht (m

)

CAMx-sonde O3 concentration (ppb) at Parlier for exp. A53

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

MM5-FDDA

Hybrid

50

90

Year Day (2000)

Heig

ht (m

)

CAMx O3 concentration (ppb) at Parlier for exp. A60

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0 010 10 1020 20

20

30 30

30

40 405060 60 60

6060

70

70 80

80

90

80

808080

90

80

90

Year Day (2000)

Heig

ht (m

)

sonde O3 concentration (ppb) at Parlier

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75

-75

-75

-65

-65

-65

-55

-55

-45

-45

-35

-35

-35 -35

-25

-25

-25

-15

-15

-25-25

Year Day (2000)

Heig

ht (m

)

CAMx-sonde O3 concentration (ppb) at Parlier for exp. A60

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

MM5

Parlier Bias

Page 11: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

30

50

Year Day (2000)

Hei

ght (

m)

CAMx O3 concentration (ppb) at Granite Bay for exp. A60

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0 0 0

000

10

1010

20 20

2020

30 30

3030

40 40

40

50

5050

60 60

60

70

80

60

8080

70

70

Year Day (2000)

Hei

ght (

m)

sonde O3 concentration (ppb) at Granite Bay

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75-75

-75

-65

-65

-65

-55

-55 -55

-45-45

-45

-45

-35

-35

-25-25

-25

-25

-15

-15

-5

-5-5 515

-35

15

-25

25

25

5

Year Day (2000)

Hei

ght (

m)

CAMx-sonde O3 concentration (ppb) at Granite Bay for exp. A60

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

30

Year Day (2000)

Hei

ght (

m)

CAMx O3 concentration (ppb) at Granite Bay for exp. A59

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0

000

10 10

1010

20 20

2020

30 30 30

303030

40

40

50

5050

60

60

70

70 80

60

Year Day (2000)

Hei

ght (

m)

sonde O3 concentration (ppb) at Granite Bay

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75

-75

-75

-65

-65

-65

-55

-55

-45-45

-45

-35-35

-35

-35

-25

-15

-15-15

-15

-5

-55

-25

15 255

-35

5 515 25

Year Day (2000)

Hei

ght (

m)

CAMx-sonde O3 concentration (ppb) at Granite Bay for exp. A59

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

60

Year Day (2000)

Hei

ght (

m)

CAMx O3 concentration (ppb) at Granite Bay for exp. A53

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0

000

10

1010

20

2020

30 30

3030

40

40

50 5060 60

60

70

70

80

70

80

70 70

Year Day (2000)

Hei

ght (

m)

sonde O3 concentration (ppb) at Granite Bay

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75

-75

-65-65

-65

-55

-55

-45-35-25

-25

-15-15

-15

-5-5-5 55

15

-5

15

-5

5

5

-5

5

51525

Year Day (2000)

Hei

ght (

m)

CAMx-sonde O3 concentration (ppb) at Granite Bay for exp. A53

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

30

50

Year Day (2000)

Hei

ght (

m)

CAMx O3 concentration (ppb) at Granite Bay for exp. A60

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0 0 0

000

10

1010

20 20

2020

30 30

3030

40 40

40

50

5050

60 60

60

70

80

60

8080

70

70

Year Day (2000)

Hei

ght (

m)

sonde O3 concentration (ppb) at Granite Bay

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75-75

-75

-65

-65

-65

-55

-55 -55

-45-45

-45

-45

-35

-35

-25-25

-25

-25

-15

-15

-5

-5-5 515

-35

15

-25

25

25

5

Year Day (2000)

Hei

ght (

m)

CAMx-sonde O3 concentration (ppb) at Granite Bay for exp. A60

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

Granite Bay

MM5

MM5-FDDA

Hybrid

Sonde

Page 12: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

30

Year Day (2000)

Heig

ht (m

)

CAMx O3 concentration (ppb) at Granite Bay for exp. A59

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0

000

10 10

1010

20 20

2020

30 30 30

303030

40

40

50

5050

60

60

70

70 80

60

Year Day (2000)

Heig

ht (m

)

sonde O3 concentration (ppb) at Granite Bay

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75

-75

-75

-65

-65

-65

-55

-55

-45-45

-45

-35-35

-35

-35

-25

-15

-15-15

-15

-5

-55

-25

15 255

-35

5 515 25

Year Day (2000)

Heig

ht (m

)

CAMx-sonde O3 concentration (ppb) at Granite Bay for exp. A59

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

60

Year Day (2000)

Heig

ht (m

)

CAMx O3 concentration (ppb) at Granite Bay for exp. A53

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0

000

10

1010

20

2020

30 30

3030

40

40

50 5060 60

60

70

70

80

70

80

70 70

Year Day (2000)

Heig

ht (m

)

sonde O3 concentration (ppb) at Granite Bay

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75

-75

-65-65

-65

-55

-55

-45-35-25

-25

-15-15

-15

-5-5-5 55

15

-5

15

-5

5

5

-5

5

51525

Year Day (2000)

Heig

ht (m

)

CAMx-sonde O3 concentration (ppb) at Granite Bay for exp. A53

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

30

50

Year Day (2000)

Heigh

t (m

)

CAMx O3 concentration (ppb) at Granite Bay for exp. A60

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

0 0 0

000

10

1010

20 20

2020

30 30

3030

40 40

40

50

5050

60 60

60

70

80

60

8080

70

70

Year Day (2000)

Heigh

t (m

)

sonde O3 concentration (ppb) at Granite Bay

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

-75-75

-75

-65

-65

-65

-55

-55 -55

-45-45

-45

-45

-35

-35

-25-25

-25

-25

-15

-15

-5

-5-5 515

-35

15

-25

25

25

5

Year Day (2000)

Heigh

t (m

)

CAMx-sonde O3 concentration (ppb) at Granite Bay for exp. A60

211 211.5 212 212.5 213 213.5 214 214.5 2150

1000

2000

3000

4000

5000

Granite BayBias

MM5

MM5-FDDA

Hybrid

Page 13: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Parlier

MM5-FDDA Hybrid

SondeModelBias

Page 14: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Granite Bay

MM5-FDDA Hybrid

Page 15: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Bias

Parlier Granite Bay

MM5MM5-FDDAHybrid

Page 16: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

RMS difference

Parlier Granite Bay

MM5MM5-FDDAHybrid

Page 17: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Correlation Coefficient

Parlier Granite Bay

Cor Coef Cor Coef

Page 18: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Biggest differences aloft among 3 models are due to larger O3 value used as lateral

and top boundary condition

What is the most appropriate value for BC for ozone?

Page 19: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Trinidad Head (NOAA/CMDL)

Page 20: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Trinidad Head Summer 2000

11 soundings

Page 21: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Median ozone profiles at all IONS sites, July 1 – August 15, 2004

Owen Cooper, AL/NOAASummer 2004 Ozone Sondes

Page 22: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Owen Cooper, AL/NOAA

Page 23: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Owen Cooper, AL/NOAA

Page 24: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Owen Cooper, AL/NOAA

Page 25: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

AsianPlume

Qing Liang/UW

Page 26: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Observed Chemical Composition of Asian PlumesBackground Asian Plumes Observed Δ

CO, ppbv 94 113 +19

O3, ppbv 74 100 +26

HNO3, pptv 237 362 +125

NOx, pptv 345 372 +27

PAN, pptv 335 605 +270

HCN, pptv 271 417 +146

Ethane, pptv 744 924 +180

Acetylene, pptv 84 177 +93

Acetone, pptv 1471 2162 +691

Benzene, pptv 13 21 +9

Methanol, pptv 1364 2370 +1006

SO4=, pptv 76 61 -15

Qing Liang/UW

Page 27: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

• TH ozone sondes suggest that mean tropospheric lateral bc’s should be approximately 60 ppbv

• Significant departures from 60 ppbv can occur due to tropopause folding events and due to asian plumes

• Significant height variation exists in the mean TH profile, especially in lowest 2 km

Page 28: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

NWS Eta-CMAQ Boundary Conditions

• CMAQ-Eta simulations for the summer of 2004 in New England over-predicted surface ozone by ~10 ppb. Mathur (2005) traced this bias to too high of ozone values for lateral and top boundary conditions.

• Present CMAQ-Eta operational model uses O3 lateral and top boundaries from the GFS global model for heights > 6km, climatology below.

Page 29: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Satellite Measurements of Chemical Species

• O3• CO• CH4• CO2• NO2• HCHO• SO2• AOD, etc.

• AIRS• MODIS• SCIAMACHY• MOPITT• GOME, etc.

Page 30: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Recommended areas of investigation – High bias at coastal

sites and over land at night

• Sensitivity tests to height-dependent O3 profile in initial state and at lateral inflow boundaries

• Vertical resolution of CAMx near surface

• Chemical deposition over water

• Kz in stable conditions

• Chemical processing in off-shore stratus(?)

Page 31: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Recommended areas of investigation – Low bias at land

sites during day

• Sensitivity tests to larger O3 lateral and top boundary values (30-110 ppbv)

• Higher vertical resolution chemistry model simulation to reduce mixing between PBL and aloft

Page 32: Motivation Model simulations for California have usually underestimated ozone concentrations aloft

Next steps

• Analyze day and night profile statistics separately at Parlier and Granite Bay

• Compare 3 models with Trinidad Head O3 soundings

• Compare 3 models with aircraft observations• Repeat analysis using MM5-CMAQ from

Seasonal Modeling study• Add MM5-CAMx runs using hourly averaged

model winds• Add WRF-Chem run using CARB emission

inventory• Repeat all of analysis for September IOP