Baton Rouge 8-hr Ozone Modeling Technical Review Meeting

V:\corporate\marketing\overview.ppt

Baton Rouge 8-hr Ozone ModelingTechnical Review Meeting

Presentation to theLDEQ/AQSD

&8-hr Ozone SIP Coalition

July 27, 2006


Today’s Presentation

• Conceptual Model of 8-hr Ozone• Episode Analysis• Modeling System• Modeling Domain• Emission Inventory Development

– Area & Point Sources– Motor Vehicles (on-road & off-road)– Biogenics & Fires


Today’s Presentation

• Initial/boundary conditions• Use of Probing Tools


Conceptual Model

• Baton Rouge is a Marginal 8-hr Ozone nonattainment area– Single 2003 exceedance DV:

• LSU = 86 ppb– Four 2005 exceedance DVs:

• LSU = 96 ppb– 2006 exceedances:

• Will not attain standard by June 2007 as required


Conceptual Model– Likely “bump-up” to Moderate area

• Attainment date: June 2010

• Spatial distribution– Four key exceedance monitors

• LSU, Baker, Carville, Port Allen• Aligned south-to-north along river• Other monitors mostly outside & south of

Baton Rouge– Likely interaction between urban,

industrial, and biogenic emissions


Conceptual Model


Conceptual Model

• Temporal distribution– Diurnal: classic 12 – 4 PM peaks

• Evidence of ozone cloud transport site to site• No obvious late AM ROFEs or THOEs• Possible early PM ROFEs or THOEs

– Weekly: no clear-cut weekday/weekend dependencies – more analysis needed

– Seasonal: trends toward late spring, early fall multi-day, multi-site episodes• In last few years, mid-summer episodes are

typically 1-day, few sites


Conceptual ModelHourly Ozone: 28-30 September, 2004

0

20

40

60

80

100

120

140

160

9/28 9/29 9/30 10/1

Date

ppbv

BRBKCVPA


Conceptual Model

• Weather requirements– Stagnation, light/variable winds under

high pressure• Exceedances independent of wind direction

– Clear skies– Temperatures do not need to be hot

• Many exceedances in low 80’s F• Emissions-driven, not heat-driven

– Few prolonged hot summer episodes• Excessive PBL venting or Gulf breeze?

Recent interannual climate?


Conceptual Model

• Regional transport– Needs 2-3 day transport times

• Recent summer episodes are too short• Some multi-day episodes are clearly caused

by local stagnation, re-circulation• Some multi-day episodes establish

consistent transport corridors from midwest, Ohio Valley, south-east U.S.

– Choose episodes that represent mix of conditions• Season, meteorology, transport, WE/WD


Episode Analyses

• Screened 14 episodes from 2000-2004– 2005 de-emphasized

• Concern about availability of emissions data

• Reduced to 6 candidates– Max exceedance monitor-days at 4 key

monitors– Min number of modeling days– Different times of year– Recent episodes


Episode Analyses

• 6 candidates– May 19-30, 2003 (M, Sa, W, Th)– September 28-30, 2004 (W, Th)– April 12-30, 2003 (Su, M, F, Su, M, Tu)– October 4-6, 2003 (Sa, Su)– May 4-9, 2004 (Tu, W, Th, Sa)– August 11 – September 5, 2000 (F, Su,

Th, F, Sa, Su, M, F, Sa, W, Th, F, Sa, Su)


Episode Analyses• Comparison to CART analysis

• Bin 10 (22%)• Bin 20 (24%)• Bin 25 (33%)• Bin 27 (10%)• Bin 35 (11%)

Table 3-6. CART Bins for 4 ranked episodes for which data are available (Source: ICF, 2005).

Rank Episode CART Bins for Exceedance Days 3 April 2003 10, 18, 18, 25, 25 1 May 2003 6, 27, 28 5 May 2004 6, 10, 19, 20, 25 2 September 2004 29, 31


Episode Analyses

• Final 2-3 episodes need to be selected from 6 candidates– Input from advisory group– Consider schedule, resources, and pre-

existing datasets– July 31 draft Protocol documents

conceptual model for each of 6 episodes


Episode AnalysesHourly Ozone: 19-30 May, 2003

0

20

40

60

80

100

120

140

160

5/19 5/20 5/21 5/22 5/23 5/24 5/25 5/26 5/27 5/28 5/29 5/30 5/31

Date

ppbv

BRBKCVPA


Episode AnalysesHourly Ozone: 28-30 September, 2004

0

20

40

60

80

100

120

140

160

9/28 9/29 9/30 10/1

Date

ppbv

BRBKCVPA


Episode AnalysesHourly Ozone: 12-30 April, 2003

0

20

40

60

80

100

120

140

160

4/12

4/13

4/14

4/15

4/16

4/17

4/18

4/19

4/20

4/21

4/22

4/23

4/24

4/25

4/26

4/27

4/28

4/29

4/30 5/1

Date

ppbv

BRBKCVPA


Episode AnalysesHourly Ozone: 4-6 October, 2003

0

20

40

60

80

100

120

140

160

10/4 10/5 10/6 10/7

Date

ppbv

BRBKCVPA


Episode AnalysesHourly Ozone: 4-9 May, 2004

0

20

40

60

80

100

120

140

160

5/4 5/5 5/6 5/7 5/8 5/9 5/10

Date

ppbv

BRBKCVPA


Episode AnalysesHourly Ozone: 11 Aug - 5 Sep, 2000

0

20

40

60

80

100

120

140

160

8/11

8/12

8/13

8/14

8/15

8/16

8/17

8/18

8/19

8/20

8/21

8/22

8/23

8/24

8/25

8/26

8/27

8/28

8/29

8/30

8/31 9/1

9/2

9/3

9/4

9/5

9/6

Date

ppbv

BRBKCVPA


Modeling System

• MM5 – Meteorological Model– Widely used to support regulatory

modeling• EPS3 – Emissions Processor

– LDEQ is familiar with EPS• CAMx – Air Quality Model

– Widely used to support regulatory modeling in south-central U.S.

– Full Chemistry PiG allows evaluation of HRVOC plumes


Domain Configuration

• Regional grid (36 km) based on TCEQ• Intermediate grid (12 km) captures

transport from Midwest and Southeast• Local grid (4 km) along Gulf Coast,

including Houston• Vertical grid based on St. Louis

modeling



-2592 -2232 -1872 -1512 -1152 -792 -432 -72 288 648 1008 1368 1728 2088 2448-1800

-1440

-1080

-720

-360

0

360

720

1080

1440

1800

Birm ingham

Washington

Miami

Tampa

Atlanta

Chicago

Indianapolis

W ichita

Louisville

New Orleans

Shreveport

Boston

Detroit

Minneapolis

Kansas City St. Louis

Charlotte

Omaha

Buffalo

New York

Cincinnati

Cleveland

Columbus

Oklahom a City Tulsa

Philadelphia Pittsburgh

Mem phis

Nashville

Austin

Dallas

Houston San Antonio

Virginia Beach

Milw aukee

36 km MM5

36 km

4 km

12 km



-72 48 168 288 408 528 648 768 888 1008 1128 1248 1368 1488 1608 1728 1848-1332

-1212

-1092

-972

-852

-732

-612

-492

-372

-252

-132

-12

108

228

B irm ingham

Jacksonville

Tam pa

A tlanta

Ind ianapolis

W ichita

Louisville

New O rleans

Kansas C ity S t. Louis

Charlotte

O m aha O m aha

C incinnati

Colum bus

O klahom a C ity

Tu lsa M em phis

Nashville

Dallas

Houston San Anton io

12 km MM5

12 km

4 km


Domain ConfigurationMM5 CMAQ/CAMx Layer Sigma Pres (mb) Height (m) Depth (m) Layer Pres (mb) Height (m) Depth (m)

34 (top) 0.000 100 18123 2856 16 100 18123 7987 33 0.050 145 15267 2097 32 0.100 190 13170 1659 31 0.150 235 11510 1374 30 0.200 280 10136 1173 15 280 10136 3106 39 0.250 325 8963 1024 28 0.300 370 7938 909 27 0.350 415 7030 817 14 415 7030 2866 26 0.400 460 6213 742 25 0.450 505 5471 680 24 0.500 550 4791 627 23 0.550 595 4163 582 13 595 4163 1635 22 0.600 640 3581 543 21 0.650 685 3038 509 20 0.700 730 2528 386 12 730 2528 664 19 0.740 766 2142 278 18 0.770 793 1864 269 11 793 1864 443 17 0.800 820 1596 174 16 0.820 838 1421 171 10 838 1421 338 15 0.840 856 1251 167 14 0.860 874 1083 164 9 874 1083 324 13 0.880 892 920 161 12 0.900 910 759 79 8 910 759 158 11 0.910 919 680 78 10 0.920 928 601 78 7 928 601 155 9 0.930 937 524 77 8 0.940 946 447 76 6 946 447 152 7 0.950 955 371 75 6 0.960 964 295 75 5 964 295 149 5 0.970 973 220 74 4 0.980 982 146 37 4 982 146 37 3 0.985 987 109 37 3 987 109 37 2 0.990 991 73 36 2 991 73 36 1 0.995 996 36 36 1 996 36 36

0 (ground) 1.000 1000 0 0 0 0 0 0


Emission Inventories

• Start-point inventories:– 2002 CENRAP Base B– 2002 VISTAS Base G– 2002 MRPO Base K– TCEQ inventories and other datasets

• Useful for August/September 2000 episode• Replace CENRAP data for Texas when

available for other years?



– 2000-2005 LDEQ data, as available• Stationary point, area• On-road: VMT, MOBILE6• Non-road: NONROAD



• Inventory projections– Specific modeling years (2000-2004,

2009)• Regional: based on RPOs, EPA, and TCEQ• Local: future year projections are challenging

due to recent events– EGU and other large sources:

• Use CEM as available for base case model performance evaluation

• Use “typical” inventory rates for projected base and future years



• Biogenics from GloBEIS– All grids, episode-day specific– Driven by:

• MM5 and/or temperature observations• Satellite PAR data• Land cover/biomass data (GIS, etc.)

– Held constant into 2009 future year

• Fires (wild, agricultural, prescribed)– As needed, as available


Initial/Boundary Conditions

• Only needed on 36-km grid• Use 2002 VISTAS model output fields

– Derive monthly-average diurnally-varying IC/BCs for base case episodes

• Use 2009 VISTAS model output fields– If available, as described above– Otherwise, use 2002 fields


CAMx Probing Tools

• Ozone Source Apportionment Technology (OSAT)– Determines source area/category

contribution to ozone anywhere in the domain

– Tracks NOx and VOC precursor emissions, ozone production/destruction, and initial/boundary conditions

– Estimates ozone production uner NOx- or VOC limited conditions


CAMx Probing Tools

• OSAT– HOWEVER: it cannot quantify ozone

response to NOx or VOC controls– Chemical allocation methodologies:

• OSAT: standard approach• APCA: attributes ozone production to

anthropogenic (controllable) sources only• GOAT: tracks ozone based on where it

formed, not where precursors were emitted


CAMx Probing Tools

• PM Source Apportionment Technology (PSAT)– Parallel to OSAT operation– Tracks user-defined groups of species

for sulfate, nitrate, ammonium, SOA, Mercury, and primary PM


CAMx Probing Tools

• Decoupled Direct Method (DDM) for sensitivity analysis– Calculates first-order concentration

sensitivity to emissions, initial/boundary conditions

– Allows estimates of effects of emission changes

– Allows ranking of source region/ categories by their importance to ozone formation


CAMx Probing Tools

• DDM– Slower than OSAT, but:

• Provides information for all species (not just ozone)

• More flexible in selecting which parameters to track

• Better estimate of small emission perturbation impacts (e.g., control measures)

• Includes sensitivity from non-linear secondary effects


CAMx Probing Tools

• Process Analysis (PA)– Designed to provide in-depth analyses of

all physical and chemical processes operating in model

– Operates on user-defined species and any portion of the modeling grid

– Appropriate for evaluating base case performance

– Recent UNC enhancements


CAMx Probing Tools• PA

– Three components:• Integrated Process Rate (IPR): provides

detailed process rate information for each physical process (emissions, advection, diffusion, chemistry, deposition)

• Integrated Reaction Rate (IRR): provides detailed reaction rate information for all chemical reactions

• Chemical Process Analysis (CPA): like IRR, but designed to be more user-friendly and accessible


CAMx Probing Tools

• Reactive Tracers (RTRAC)– Tracks multiple independent reactive gas

and particle tracers• Tracers operate in parallel to the CAMx host

model• Allows for several generations of products• Decay/production uses standard gas-phase

mechanism photolysis and oxidants• Can output tracer decay rates to a separate

Lagrangian model for “fenceline” dispersion calculations


CAMx Probing Tools

• RTRAC– Designed for primarily for air toxics

• Assumes RTRAC species have minimal impact on photochemistry

• Each tracer can be “tagged” for source apportionment

– RTRAC works with IRON PIG• An optional “sampling grid” capability

passively samples near-source sub-grid RTRAC surface concentrations within reactive PiG plumes

Baton Rouge 8-hr Ozone Modeling Technical Review Meeting

Documents

Transcript of Baton Rouge 8-hr Ozone Modeling Technical Review Meeting