Technical Highlights of EPA’s 7 Conference on Air Quality ... · iii Fax Question Sheet APTI Workshop T-029 Technical Highlights of EPA’s 7th Conference on Air Quality Modeling

EnvironmentalPrograms

Technical Highlights ofEPA’s 7th Conference onAir Quality Modeling

Workshop GuideAPTI Workshop T-029DAY 2

Developed by Environmental Programs - North Carolina State UniversityEPA Cooperative Assistance Agreement CT-825724________________________________________________________________________________

Industrial Extension Service College of Engineering North Carolina State University

ii

This project has been funded wholly or in part by the United States Environmental Protection Agency(U.S. EPA) under Cooperative Assistance Agreement CT-825724 to North Carolina State University.The contents of this document do not necessarily reflect the views and policies of U.S. EPA, nor doesmention of trade names or commercial products constitute endorsement or recommendation for use.

2000 North Carolina State University

All rights reserved. State and local air pollution control agencies, U.S. EPA offices, and federal officesdesignated by EPA are authorized to make royalty-free copies of this document in connection withtelecourses. Otherwise, no part of this publication may be reproduced or transmitted in any form or byany means electronic or mechanical, including photocopying, recording, or information storage andretrieval systems, for any purpose without the express written permission of North Carolina StateUniversity.

Printed on recycled paper in the United States of America.

(Revision: 8/00)

iii

Fax Question Sheet

APTI Workshop T-029Technical Highlights of EPA’s 7th Conference onAir Quality Modeling DAY 2August 2, 2000

Voice: (800) 742-9813 Fax: (800) 553-7656

____________________________________________________________________________________Please write your question and direct it to the appropriate presenter if possible.

Question for:__________________________________________________________________

Question from: ________________________________________________________________

City/State:____________________________ Office Phone Number: ( ) ______________

KeyPoint Services International, Inc. Trouble Line: (800) 231-8945

iv

(This page intentionally left blank.)

v

APTI Workshop T-029

Technical Highlights of EPA’s 7th Conference on Air Quality Modeling

Presented by OAQPS

Broadcast Agenda

August 2, 2000 1:00pm ET DAY 2

SECTION TOPIC

1 Introduction Jim Dicke

2 Alternative ModelsADMS David Carruthers, Ph.D.CAMx Ralph MorrisSCIPUFF Ian Sykes

10 MIN. BREAK

3 Alternative ModelsHYROAD Introduction Edward CarrHYROAD Intersection Model Robert IresonUAM-V Edward Carr

4 SummaryJoe Tikvart

10 MIN. BREAK

Questions and Answers and Wrap up

Technical Highlights of EPA’s 7th Conference on Air Quality Modeling

Presenters

Edward CarrICF ConsultingSan Francisco, [email protected]

Dr. David J. CarruthersCambridge Environmental ResearchConsultants, Cambridge, [email protected]

Alan J. CimorelliRegional Modeler, EPA Region IIIPhiladelphia, [email protected]

Julie Ann DraperFederal Aviation AdministrationWashington, [email protected]

Chuck HakkarinenEPRIPalo Alto, [email protected]

Robert IresonICF ConsultingSan Francisco, [email protected]

John S. IrwinMeteorologist, NOAAAir Quality Modeling Group, [email protected]

Ralph E. MorrisENVIRON International Corp.Arlington, [email protected]

Robert J. Paine

ENSR CorporationActon, [email protected]

Warren D. PetersEnvironmental EngineerAir Quality Modeling Group, [email protected]

Joseph S. ScireVice President, Earth Tech, Inc.Concord, [email protected]

R. Ian SykesARAP GroupTitan [email protected]

Theodore G. ThrasherCSSI, Inc.Washington, [email protected]

Joe TikvartLeader, Air Quality Modeling Group, EPAResearch Triangle Park, [email protected]

John VimontNational Park ServiceDenver, [email protected]

Roger L. Wayson, Ph. D.University of Central Florida and Visiting Scientist, [email protected]

Jeffrey C. WeilCIRES, University of ColoradoBoulder, [email protected]

Rob WilsonRegional Modeler, EPA Region 10

Seattle, [email protected]

Presentation 1 1

Technical Highlights of EPA’s 7th Conference on Air Quality ModelingADMS Atmospheric Dispersion Modeling System

Dr. David J Carruthers

Dr. David J CarruthersCambridge Environmental

Research Consultants

ADMSAtmospheric Dispersion

Modelling System

ADMS! Development commissioned in

1988 following a CERC report toregulatory authorities in the UK

! The CERC report highlighted theadvantages of the use ofsurface/boundary layer scalingover Pasquill Gifford stabilitycategories

ADMS! Sponsors include UK’s

Environment Agency, UK Healthand Safety Executive, major powerand chemical companies

Presentation 1 2



! Development by:

" CERC(including Prof. Julian Hunt,Dr. David Carruthers, Dr. ChristineMcHugh, Dr. Rex Britter)

" University of Surrey(Prof. Alan Robins)

" UK Meteorological Office(Dr. David Thomson)

ADMS

ADMS! ADMS is the leading European

Short Range Air Dispersion Modeland is used extensively in the UKand across Europe

! ADMS has featured in all 6European Workshops onHarmonisation of DispersionModels (1991-present)

Key Features of ADMS! Continuous or discrete releases

! Point, line, area, volumeand jet sources

" treatment depends onreceptor location

Presentation 1 3



! Skewed-Gaussian model usinglocal boundary layer variables

! Meteorological preprocessor

! Integral plume rise model

Key Features of ADMS

Key Features of ADMS! Building effects

! Complex terrain

! Coastline

! Wet and dry deposition

! Chemical transformation

! Radioactive decay & gamma dose

! Jets and directional releases

! Concentration fluctuations module

! Condensed plume visibility module

Key Features of ADMS

Presentation 1 4



Regulatory Applications! Multiple buoyant or passive

industrial emissions

! Surface, near surfaceor elevated releases

! Urban or rural areas

! Short (seconds) to long (annual)term averaging times

Flat Terrain Validation

Table 1

Table 2

Summary Scores for ISC3, ADMS and AERMOD(Different model input parameters)

Table 1 from Hanna et al, 6th Workshop on Harmonisation, France Oct 1999Table 2 from Hanna et al, AWMA Meeting, US, June 2000

ISC3 ADMS AERMODBest 4 8 10Middle 10 15 11Worst 10 1 3

ISC3 ADMS AERMODBest 5 19 6Middle 2 5 11Worst 17 0 7

5000SO2 emission rate (g/s)

130Temperature (°°°°C)

22Exit velocity (m/s)

13Stack diameter (m)

200Stack height (m)

Typical input data

Meteorological data:

1 year of hourly sequential datafrom Manchester, UK, 1995

7.52.6Annual mean

3.20.290th

543.995th

1275198th

1809799th

52118899.9th

66931099.99th

681398100th

AERMODADMS 3

Maximum values of Percentilesand Annual Mean Concentration

Power StationComparison

Presentation 1 5



ModellingBuildingEffects

Two plumemodel

Based onflow field

large building

Building Validation -Wind Tunnel Experiment

small building

Building Validation - Results

0.00001

0.0001

0.001

0.01

0.1

1

0.0

000

1

0.0

001

0.0

01

0.0

1

0.1 1

BRE Dimensionless Concentration K

AD

MS

Dim

ensi

on

les

s C

on

cen

trat

ion

K S 1-opening

W 1-opening

X 1-opening

S 4-openings

W 4-openings

X 4-openings

S 9-openings

W 9-openings

X 9-openings

S 15-openings

W 15-openings

X 15-openings

ADMS=BRE

ADMS=FACTOR 2 > BRE

ADMS=FACTOR 2 < BRE

Comparison of ADMS with wind tunnel results for the large buildingwith 1, 4, 9 and 15 openings, increasing buoyancy flux (S, W and X)

Presentation 1 6



Above: View of idealized hill

-3000 0 3000-3000

0

3000

Complex Terrain EffectsBased onmodelledflow field

Right: Flow field at 80mabove terrain and groundlevel concentration froman 80m stack

ADMS and AERMOD Comparison:Terrain amplification factors

-2000 -1500 -1000 -500 0 500 1000 1500 2000-2000

-1500

-1000

-500

0

500

1000

1500

2000

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2.0

2.1

2.2

-2000 -1500 -1000 -500 0 500 1000 1500 2000-2000

-1500

-1000

-500

0

500

1000

1500

2000

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

2.6

2.8

3.0

3.2

3.4

3.6

3.8

4.0

4.2

4.4

-2000 -1500 -1000 -500 0 500 1000 1500 2000-2000

-1500

-1000

-500

0

500

1000

1500

2000

400

500

600

700

800

900

1000

1100

1200

1300

1400

1500

1600

1700

1800

1900

-2000 -1500 -1000 -500 0 500 1000 1500 2000-2000

-1500

-1000

-500

0

500

1000

1500

2000

200

250

300

350

400

450

500

550

Left: ADMS CmaxRight: AERMOD Cmax

Ratio of complexterrain to flatterrain as functionof stack location

• Neutral conditions• 50m stack• Idealised hill

Left: ADMS XmaxRight: AERMOD Xmax

-1500 -1000 -500 0 500 1000 1500 20000

250

500

750

-1500 -1000 -500 0 500 1000 1500 20000

250

500

750

0.0

0.5

1.0

1.5

2.0

2.5

5.0

10.0

15.0

20.0

25.0

Ratio of complex terrainto flat terrain maximumconcentrations asfunction of stack heightand location

Wind tunnel, fromLawson et al

ADMS

AERMOD

ADMS and AERMOD Comparison:Complex terrain, Neutral flow

Presentation 1 7



-1500 -1000 -500 0 500 1000 1500 20000

100

200

300

400

-1500 -1000 -500 0 500 1000 1500 20000

100

200

300

400

0

1

2

3

4

5

6

7

8

9

10

20

30

40

45

50

ADMS

AERMOD

ADMS and AERMOD Comparison:Complex terrain, Stable flow

Ratio of complex terrainto flat terrain maximumconcentrations asfunction of stack heightand location

• Wind Speed (10m)= 0.76 m/s

• Monin-Obukhov length= 23 m

ADMS and AERMOD Comparison:Complex terrain results

369000 375000 381000437000

443000

449000

25

50

75

100

125

150

175

200

225

250

369000 375000 381000437000

443000

449000

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

MaximumConcentration (ug/m3)

Long Term AverageConcentration (ug/m3)

ADMS (Max=178) ADMS (Max=4.0)

AERMOD (Max=1162) AERMOD (Max=10.3)369000 375000 381000

437000

443000

449000

369000 375000 381000437000

443000

449000

Stack andsurrounding

terrain

ADMS-Urban• Advanced airquality modelfor major citiesand airports

• Includesalgorithms fordispersion fromroads andstreet canyons

• Used in majorcities in Europeand beyond

LondonAnnual average NO2 concentration, 2005

Presentation 1 8



Summary! ADMS includes in one model all

the features of AERMOD (exceptinput of observed boundary layerprofiles), ISC-PRIME and CTDMPLUS – potential difficulties arisingas to whether to use AERMOD orISC-PRIME avoided (e,.g, at sitewith buildings and tall stacks)

! Additionally ADMS includesconcentration fluctuation, plumechemistry and condensed plumevisibility algorithms

Summary

Summary! ADMS-Urban includes capabilities

of CALINE, most features of EDMSand other features

! The costs of ADMS are similarto commercially availableversions of AERMOD and ISC

Presentation 1 9



! ADMS was first released in1993 and has been used inmany critical applications

! There are over 500licenses worldwide

Summary

http://www.cerc.co.uk

Presentation 2 1

Technical Highlights of EPA’s 7th Conference on Air Quality ModelingComprehensive Air-quality Model with extensions (CAMx)

Ralph E. Morris

ComprehensiveAir-quality

Model with extensions(CAMx)

Ralph E. MorrisENVIRON International Corp.

! 3-D Eulerian troposphericphotochemical transport model

" treats emissions, chemistry,dispersion, removal of gaseous andaerosol air pollution

" scales range from individual pointsources (< 1 km) to regional (>1000 km)

CAMx Version 2.00

! Combines features required of“state-of-the-science” models

" new coding of several industry-accepted algorithms

" computationally and memoryefficient

" easy to use

CAMx Version 2.00

Presentation 2 2


Ralph E. Morris

" modular framework permits easysubstitution of revised and/oralternate algorithms

" publicly available (www.camx.com)

CAMx Version 2.00

! Technical Features:

! Grid nesting

" two-way horizontal and verticalnesting

" supports multiple levels

" variable meshing factors

CAMx Version 2.00

! Plume-in-Grid (PiG) sub-model

! Multiple, fast and accuratechemical mechanisms

! Mass conservative and massconsistent transport scheme

CAMx Version 2.00

Presentation 2 3


Ralph E. Morris

0 1 2 3 4 5 6 7 8 9 100

1

2

3

4

5

6

7

8

9

10

Coarse Grid Boundary Cells

Coarse Grid Boundary Cells

Co

ars

e G

rid

Bo

un

da

ry C

ells

Co

ars

e G

rid

Bo

un

da

ry C

ells

Fine Buffer Cells

Fine Buffer Cells

Fin

e B

uff

er

Ce

lls

Fin

e B

uffe

r C

ells

Boundary of Fine Grid 1

Boundary of Fine Grid 2

! Multiple map projections

" curvi-linear latitude/longitude

" Universal Transverse Mercator

" Lambert Conformal (MM5)

" Rotated Polar Stereographic (RAMS)

CAMx Version 2.00

! Ozone Source Apportionment(OSAT)

" tracks source region/categorycontributions to receptor ozoneconcentrations

" indicates if ozone formed in NOxor VOC-limited conditions

CAMx Version 2.00

Presentation 2 4


Ralph E. Morris

! Ability to use historical airquality model databasesdeveloped for other models

" OTAG, LMOS, COAST/Houston,Atlanta, Northeast Corridor

CAMx Version 2.00

Key Technical Components! Overview

" solves continuity equationfor each species

" time splitting operation

# each process solved individuallyfor each grid, each time step

" time step size maintains stablesolution of transport on each grid

# multiple transport steps permaster grid step required for

# nested grids

# multiple chemistry stepsper transport step required

" model developed to run onmeteorological modeling grid

# reduces error due tointerpolation and averaging

# multiple map projections available

Key Technical Components

Presentation 2 5


Ralph E. Morris

! Transport

" advection and diffusion solvers aremass conservative

" horizontal and vertical advectionlinked through the divergentcompressible atmospheric continuityequation

# mass consistency


" order of east-west andnorth-south advectionalternates each master grid step

" three options available forhorizontal advection solvers:

# Smolarkiewicz (1983): diffusion-corrective forward-upstream scheme

# Bott (1989): area-preservingflux-form solver

# Piecewise Parabolic Method (PPM)


! Transport (concluded)

" vertical transport and diffusionsolved with an implicit scheme

" dry deposition rates are used asthe surface boundary condition

" horizontal diffusion solvedwith an explicit scheme intwo directions simultaneously


Presentation 2 6


Ralph E. Morris

! Pollutant Removal

" dry deposition velocities for eachspecies determined using resistanceapproach (Wesely, 1989)

# dependent upon: season, land cover,solar flux, near-surface stability,surface wetness, species solubilityand diffusivity

# for aerosols: size spectrumdictates sedimentation velocity


" wet scavenging based onMaul (1980) as implementedin CALPUFF (EPA, 1995)

# exponential decay

# decay rate dependent upon:rainfall rate, species solubility

# species removed from entiregrid column (all layers)


! Photochemistry

" CBM-IV (Gery et al., 1989)

# 3 variations available

" SAPRC97 (Carter, 1990)

# chemically up-to-date

# tested extensively againstenvironmental chamber data

# uses a different approachfor VOC lumping


Presentation 2 7


Ralph E. Morris

" all mechanisms are balancedfor nitrogen conservation

" photolysis rates derivedfrom TUV preprocessor

# generates lookup table over:zenith angle, altitude,ozone column, albedo, turbidity

# first two determined foreach grid cell internally

# last three provided by input files


" photolysis rates affected by clouds

# UAM-V approach: rates scaled byfractional cloud coverage only

# RADM approach: rates scaled byoptical depth and cloud coverage


! Chemistry Solver

" most “expensive” component ofphotochemical grid simulations

" CAMx solver increases efficiencyand flexibility

" adaptive hybrid approach:

# radicals (fastest reacting species)solved using implicit steady stateapproximation


Presentation 2 8


Ralph E. Morris

# fast state species solved usingsecond-order Runge-Kutta method

# slow state species solved explicitly

# “Adaptive” = number of fast statespecies changes according to thechemical regime


! Plume-in-Grid (PiG)

" fine resolution needed for near-source chemistry/dispersion of largeNOx plumes

" tracks stream of plume segments(puffs) in a Lagrangian frame

# each puff moved by winds in host cell

# puff growth (dispersion) determinedby diffusion coefficients in host cell


" GREASD PiG: faster, conceptuallysimpler

# reduced NOx chemistry set(NO-NO, NOx/ozone equilibrium, HNO3production)

# cross-sectional Gaussianpollutant distribution


Presentation 2 9


Ralph E. Morris

# puffs leak mass according togrowth rates and grid cell size

# puffs terminated due to ageor sufficiently dilute NOx


! Ozone Source Apportionment(OSAT)

" determines source area/categorycontributions to ozone anywherein the domain

" uses tracers to track precursoremissions and ozoneproduction/destruction


Presentation 2 10


Ralph E. Morris

" also tracks contribution of initialand boundary conditions

" estimates whether ozone is producedunder NOx- or VOC-limited conditions


" removes need to run modelrepeatedly to understand:

# chemical regime

# influences of various sources

" HOWEVER: cannot quantify ozoneresponse to NOx or VOC controls


! CAMx Version 2.00 PM Treatment

" Primary Particulate Matter (PM)

" Secondary Organic Aerosols(SOA) treated using aerosolyield approach in photochemistry

" Sulfate/Nitrate/Ammonia equilibriumaerosol thermodynamics usingempirical UAM/LC approach

" Empirical aqueos-phase (Sulfate)


Presentation 2 11


Ralph E. Morris

CAMx Version 3 AttributesComing Summer 2000

(www.camx.com)! Flexi-nesting

" Ability to add/delete nested-gridsduring a simulation

" Real time interpolation of fine-gridinputs from next coarser grid


(www.camx.com)! Decoupled Direct Method (DDM)

" Sensitivity coefficients providesinformation on the relationship ofCAMx-estimated ozone (or otherspecies) and sources of precursors(emissions, boundary conditions,and initial concentrations)

" Information is useful for

# Control Strategy Development

# Model performance evaluation

# Diagnostic analysis


(www.camx.com)

Presentation 2 12


Ralph E. Morris

Moving Sectional

Hybrid

CAMx Version 3 Attributes

AerosolSize Composition Distribution (fixed size grid)

Bulkaqueous-phase

chemistry

Size-resolvedaqueous-phase

chemistry

Equilibrium Dynamics

SCAPE2AERO-LC SEQUILIB2.1

Fixed Sectional

Aerosol Module

Cloud Microphysics/Chemistry Module

Advanced PM Treatment

ISORROPIA

Mass Transfer:

Advanced PM Treatment

Gas-PhaseChemistry:

Size Section:

1) SAPRC972) Enhanced CBM-IV

(monoterpines)

1) Fixed Section2) Moving Section

1) Equilibrium2) Hybrid3) Dynamic

CAMx Version 3 Attributes

SecondaryOrganicAerosol:

1) SOAM22) Aerosol Yields (existing)

Advanced PM TreatmentCAMx Version 3 Attributes

AerosolThermo-dynamic:

Aqueous-PhaseChemistry:

1) LCAERO (parameterized RFM)2) SCAPE2 (full science)3) ISORRPIA (significantly faster)

1) Bulk2) Size Resolved3) Empirical (existing)

Presentation 2 13


Ralph E. Morris

WetDeposition:

1) Rainout and washout as part of aqueous-phase module2) Existing gaseous wet deposition interface with aqueous-phase module

Advanced PM TreatmentCAMx Version 3 Attributes

Coagulation:

Nucleation:

DryDeposition:

1) CMU algorithm

1) CMU algorithm

1) Wesley gaseous (existing)2) AERO particle dry deposition

CAMx Postprocessingand Analysis Tools

! CAMxtrct

" Extracts and reformats CAMx outputfor multiple grids (Fortran)

! SURFER (by Golden Software)

" Visualization (PC based)

! MAPS (by Alpine Geophysics)

" Model evaluation and visualization(Fortran/NCAR graphics)

! PAVE (by MCNC)

" Visualization (Unix/LINUX)


Presentation 2 14


Ralph E. Morris


! VIS5D

" Visualization (various)

! CAMxDESK (by EnviroModeling)

" Visualization and analysis software(PC based)

Contribution to 1-Hour Ozone > 124ppb in the Lake Michigan Region for theJuly, 1995 Episode Composite using Base Case EPA1A (bar 1) and SIP Call

SIP2A (bar 2). EPA1A Conc = 141ppb; SIP2A Conc = 133ppb; n =1515

0

10

20

30

40

50

60

70

LakeMI IL BC W-MO OK E-MO AR KS TX KY TN IN

Source Region

Ozo

ne

Co

ntr

ibu

tio

n (

pp

b)

IC/BC

Area

Vehicles

Point

Biogenic

Contribution to 8-Hour Ozone > 84ppb in the Lake MichiganRegion for the July, 1995 Episode Composite using Base Case

EPA1A (bar 1) and SIP Call SIP2A (bar 2). EPA1A Conc =105ppb; SIP2A Conc = 99ppb; n =5383

0

5

10

15

20

25

30

35

40

LakeMI IL BC W-MO E-MO OK AR KS IN KY TX IA

Source Region

Ozo

ne

Co

ntr

ibu

tio

n (

pp

b)

IC/BC

Area

Vehicles

Point

Biogenic

Presentation 2 15


Ralph E. Morris

Contributions to 8-Hour Ozone > 84 ppb by SourceType for the July, 1995 Episode for 2007 Episode

Composite Base Case

0

20

40

60

80

100

120

Atlanta

Birmin

gham

Boston

Buffalo

Charle

ston

Charlo

tte

Chatta

nooga

Chicag

o+Milw

auke

e

Cinci

nnati

Cleve

land

Colum

bus

Dayto

n

Detro

it

Evansv

ille

Ft Way

ne

Huntingto

n

India

napolis

Knoxvill

e

Louisvi

lle

Mem

phis

Nashvi

lle

New Y

ork

Parke

rsbur

g

Philadel

phia+B

alt

Pittsb

urgh

Portlan

d

Richm

ond

SW-M

ichig

an

St Louis

Terre

Hau

te

Toronto

Was

hingto

n DC

Receptors

Ozo

ne

Co

ntr

ibu

tio

n (

pp

b)

IC/BC

Biogenics

Area

Mobile

Utilities

Industrial

84.5ppb

-97.5 -90 -82.5 -75

Longitude

28.5

31

33.5

36

38.5

41

43.5

46

La

titu

de

1 m/s

Wind

0.02

0.04

0.06

0.08

0.1

ppm

O3

CAMxDESK Overlay of Ozone and Winds

290

294298

298

302

302

302

306

306

306

Longitude

28.5

31

33.5

36

38.5

41

43.5

46

La

titu

de

10

30

50

70

90

%

Urban

286

290

294

298

302

306

310

ºK

Tsfc

CAMxDESK Overlay of SurfaceTemperature and Urban Land use

10298

294

-75-82.5-90-97.5

Presentation 3 1

Technical Highlights of EPA’s 7th Conference on Air Quality ModelingSecond-order Closure Integrated Puff (SCIPUFF)

R. Ian Sykes

R. Ian SykesARAP Group

Titan Corporation

Second-order ClosureIntegrated Puff (SCIPUFF)

Overview! Modeling Approach

! Graphical User Interface

! Input/Output

! Model Evaluation Studies

Lagrangian Puff Model! Concentration field - collection of

overlapping puffs with Gaussiandistributions

! Concentration givenby sum over all puffs

! Solve ODE’s for puff moments

Presentation 3 2


R. Ian Sykes

Lagrangian Puff Attributes! Arbitrary range of scales without

numerical grid and associateddiffusion errors

! Arbitrary time-dependent, spatiallyinhomogeneous conditions

! Multiple sources with arbitrarytime-dependence

( ) ( ) ( )

( )( )

( ) ( ) ( )

( ) ( ) ( )

( )

( )

( )

1 1 ( )exp2( )

-1

i i j j

i

ij

c c Q G

G x x x xijV

Q

x

V

α α α

α α

α α α

α

α

α

ασ

σ

= =

= − − −

=

=

=

= =

∑ ∑x x x

where

total mass of puff (zeroth moment)

puff centroid (first moment)

spatial spread (second moment)

puff volume ( )13 2( )2

2α

σπ

( ) ( ) ( )

( )( )

( ) ( ) ( )

( ) ( ) ( )

( )

( )

( )

1 1 ( )exp2( )

-1

i i j j

i

ij

c c Q G

G x x x xijV

Q

x

V

α α α

α α

α α α

α

α

α

ασ

σ

= =

= − − −

=

=

=

= =

∑ ∑x x x

where

total mass of puff (zeroth moment)

puff centroid (first moment)

spatial spread (second moment)

puff volume ( )13 2( )2

2α

σπ

Concentration Field

Puff Moment Equations

( )

( ) ( )

( ) ( )

( )

( )

( ) ( )

0

( , )i i

i j j i

ij

j iik jk

k k

dQ

dtd

x u tdt

x u c x u cd

dt Q

u u

α

α α

α α

αα

α α

σ

σ σ

=

=

′ ′ ′ ′ ′ ′+=

∂ ∂+ +

∂ ∂

x

( )

( ) ( )

( ) ( )

( )

( )

( ) ( )

0

( , )i i

i j j i

ij

j iik jk

k k

dQ

dtd

x u tdt

x u c x u cd

dt Q

u u

α

α α

α α

αα

α α

σ

σ σ

=

=

′ ′ ′ ′ ′ ′+=

∂ ∂+ +

∂ ∂

x

Presentation 3 3


R. Ian Sykes

Turbulence Closure

( ) ( )( )

i j i j i j

d Aqx u c Q u u x u c

dt

α αα′ ′ ′ ′ ′ ′ ′ ′= −Λ

( ) ( )( )

i j i j i j

d Aqx u c Q u u x u c

dt

α αα′ ′ ′ ′ ′ ′ ′ ′= −Λ

where and Λ is theturbulence length scale.

2i iq u u′ ′=2i iq u u′ ′=

Turbulent Dispersion! Closure model gives direct

relationship between turbulencequantities and diffusion rates

! Provides single diffusion modelframework for a wide range ofatmospheric scales

Model Efficiency! Puff splitting allows accurate

treatment of wind shear

! Puff merging minimizesnumber of puffs

! Efficient adaptivetime-step algorithm

! “Static” puffs forsteady-state section of plume

Presentation 3 4


R. Ian Sykes

c

! Turbulent dispersion implies arandom concentration field

! 2nd-order closure model gives bothfluctuation variance, , and

! The probability distribution ofc is then modeled by the clippednormal distribution

Concentration Fluctuations

2c′2c′

Fackrell & Robins (1982)

0.1 1 10

10

100

1000

1

SCIPUFFMaximum mean dataGround-level data

x / H

CUH2

Q

0 100

2 4 6 8

1

2

3

4

5

6

x / H

ˆ c

cm

SCIPUFF

Concentration fluctuation intensityfor different source sizes

Mean concentration vs downwinddistance

0.1 1 10

10

100

1000

1

SCIPUFFMaximum mean dataGround-level data

x / H

CUH2

Q

0 100

2 4 6 8

1

2

3

4

5

6

x / H

ˆ c

cm

SCIPUFF

Concentration fluctuation intensityfor different source sizes

Mean concentration vs downwinddistance

Plume Rise! Associate dynamic vertical

momentum and temperatureperturbation integrals with each puff

! Add evolution equations for puffdynamics based on conservationof momentum and temperature

Presentation 3 5


R. Ian Sykes

Momentum Rise

2.54.06.08.0

13.215.4

R R=15.4

13.2

8.0

6.0

4.0

2.5

0 10 20 30 40 50 60 700

10

20

30

40

50

60

70

x r0

"z r0

Buoyant Rise

z

U

Wind Shear! Use complete (six-moment)

Gaussian specification so thatshear distortions can beaccurately calculated

Presentation 3 6


R. Ian Sykes

Model Input! Source Data

# Pollutant physical andchemical properties

# Release type

! Meteorological Data

# Fixed winds

# Observational Input(surface and/or profile)

# Time-dependent3-dimensional gridded

Model Input

Model Input! Terrain for mass

consistent wind field

! Turbulence Data

# Planetary boundary layer

# Large scale variability

Presentation 3 7


R. Ian Sykes

! Boundary layer turbulence

# Profiles based on wind speed,roughness, and surface heat fluxcalculation

Model Input

GUI Input

Model Output! Sampler file

# Time history at receptor locations

! GUI plots

# Horizontal slices

# Vertical slices

# Integrated surfacedeposition and dose

# Probability

Presentation 3 8


R. Ian Sykes

GUI Output

GUI Output

Model Evaluation Studies! PGT curves

! Instantaneousdispersion data

! Lab dispersion andfluctuation data

! Continental-scaleANATEX fieldexperiment

! EPRI PMV&D tall-stack emissions

! CONFLUX (shortrange, fluctuations)

! Dugway field tests

! Model Data Archive

! ETEX

Presentation 3 9


R. Ian Sykes

MDA Passive

Pre

dic

ted C

on

centr

atio

n (

ppm

)

Observed Concentration (ppm)

103

10–1

101

10–1 101 103

a - Prairie Grassb - Hanford (Long time average)c - Hanford (Short time average)

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

aa

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

b

b

b

b

b

b

b

b

b

b

c

c

c

c

c

c

c

c

c

c

P

red

icte

d C

on

centr

atio

n (

ppm

)

Observed Concentration (ppm)

103

10–1

101

10–1 101 103

a - Prairie Grassb - Hanford (Long time average)c - Hanford (Short time average)

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

aa

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

a

b

b

b

b

b

b

b

b

b

b

c

c

c

c

c

c

c

c

c

c

Conflux

0 4 8 12 16 0 4 8 12 16C (ppm) C (ppm)

0

4

8

12

16

0

4

8

12

16

Z(m)

Z(m)

0 4 8 12 16 0 4 8 12 16C (ppm) C (ppm)

0

4

8

12

16

0

4

8

12

16

Z(m)

Z(m)

MeanConcentration

Conflux

ConcentrationFluctuation

Intensity

0 4 8 12 16 20 0 4 8 12 16 20$ c/C $ c/C

0

4

8

12

16

0

4

8

12

16

Z(m)

Z(m)

0 4 8 12 16 20 0 4 8 12 16 20$ c/C $ c/C

0

4

8

12

16

0

4

8

12

16

Z(m)

Z(m)

Presentation 3 10


R. Ian Sykes

ETEX24 Hours

SCIPUFF

Observed

44

49

54

59

64

Latit

ude

(deg

rees

)

48 Hours

44

49

54

59

64

Latit

ude

(de

gree

s)

-5 0 5 10 15 20 25

Longitude (degrees)

72 Hours

-5 0 5 10 15 20 25

Longitude (degrees)

-5 0 5 10 15 20 25

Longitude (degrees)

Model Availability

Available for downloading from theTitan website:www.titan.com/systems/prod.htm

Presentation 4 1

Technical Highlights of EPA’s 7th Conference on Air Quality ModelingThe Hybrid Roadway Intersection Model: HYROAD

Robert Ireson,Edward Carr

The Hybrid RoadwayIntersection Model:

HYROAD

Edward Carr andRobert Ireson

ICF Consulting

Contact InformationCrawford Jencks, P.E.Program Officer, NCHRP202-334-2379 / [email protected]

Edward L. Carr, CCMProject Manager, ICF Consulting415-507-7186 / [email protected]

Robert G. Ireson, Ph.D.Principal Investigator415-925-1440 / [email protected]

Overview! Background

! Scientific basis and modelformulation

! Model application -resource needs

Presentation 4 2



! Model sensitivity and performance

! Proposed applications

! Project status and next steps

Overview

Background! SIP, conformity, and EIS

“Hot-Spot” analysis forcarbon monoxide (CO)

! Sponsorship: NCHRP, FHWA

! Four phase study:

" Problem assessment -Site Monitoring Plan (1993-1995)

" Site monitoring and evaluation -Data Collection & Analysis(1993-1997)

" Model development -Evaluation & Testing (1997-2000)

" Development GUI & user guide(2000-2001)

Background

Presentation 4 3



PrimaryObjectives of Research

! Assemble a comprehensivenational database

# Model testing and evaluation

! Developed an improved fullyintegrated roadway intersectionmodel

# Dispersion

# Emissions

# Traffic

PrimaryObjectives of Research

Approach toModel Development

! Analysis of field program data tocharacterize dynamic processes

# Assess those elements mostimportant and incorporate intomodel framework

! Develop model based onunderstanding of key processes

Presentation 4 4



Scientific Basis andModel Formulation

! Intersection model components

! Limitations of existing models

! Field study design and findings

! HYROAD formulation

Intersection Dynamics! Traffic

# Queuing

# Acceleration, deceleration,and cruise

# Non-steady state

! Emissions and dispersion

# Modal effects(e.g., power enrichment)

# Buoyancy and vehiclewake turbulence

# Short transport distances

Intersection Dynamics

Presentation 4 5



Intersection ModelComponents

Outputs

The need for integration

DispersionModule

EmissionModule

TrafficModule

EmissionFactors

MeteorologicalInputs

Roadwayand Traffic

Inputs

Limitations ofExisting Models

! CAL3QHC simplifications

# Two emission states -- idle or cruise

# Steady state meteorology(PG sigma-y and sigma-z)

# Queuing based onquality of progression

! CALINE4 approaches

# Empirical emissionadjustment for acceleration

# Roadway turbulence and buoyancy

# Use of sigma-theta in place of sigma-y

Limitations ofExisting Models

Presentation 4 6



Field StudyElements - Continuous

! Traffic approach volumeand signal timing

! Meteorology

# 3 m and 10 m wind speedand direction

# Temperature, RH, sigma-theta, stability

(15 minute average)

! CO and CO2 at 16+ locations

! Sonic anemometers (2- and 3-axis)

Field StudyElements - Continuous

(15 minute average)

Field Study Layout

Presentation 4 7



Field Study Elements --Short-term Studies

! Floating car runs

# Time-distance data(1 second resolution)

! Coldstart survey(distance from trip origin)

! Tracer study (SF6)

# Light wind periods

# 15-minute averagesat multiple receptors

Field Study Elements --Short-term Studies

Field Study Findings --Flows and Turbulence

! Induced flows of 3+ m/s at roadside

! Induced flows and enhancedturbulence observable at >25 mfrom roadside

Presentation 4 8



! Observed vertical dispersionrate exceeds both PG andCALINE3/CAL3QHC

! SF6 tracer observed >100 m‘upwind’ of release point due toinduced flows

Field Study Findings --Flows and Turbulence

Field Study Findings --Traffic and Emissions

! Speed/acceleration distributionsdo not resemble any emissiontest cycles

! Power enrichmentoccurs on depart legs

! Constant g/gal emission ratesobserved at all locations forhigh concentration periods

! Enrichment does NOT appreciablycontribute to high concentrations

Field Study Findings --Traffic and Emissions

Presentation 4 9



Emissions (g/gal) ShowEnrichment on Acceleration

0

1000

2000

3000

0 1000 2000 3000

Approach Leg

De

pa

rt L

eg Emissions > 3x Mean

High CO Events ShowLittle Enrichment

-2

0

2

4

6

8

10

12

14

16

18

20

-50 0 50 100 150 200 250 300

[COx] (ppm)

[CO

] (p

pm

)

Emissions > 3x Mean

Average Emissions

Concentrations > 6 ppm

HYROAD Formulation

Microsimulation

Speed-distribution-based, Modal-ready

Hybrid steady-state and puff

OutputsDispersion

ModuleEmissionModule

TrafficModule

EmissionFactors

MeteorologicalInputs

Roadwayand Traffic

Inputs

Presentation 4 10



Traffic Module -Micro-simulation

! Based on TRAF-NETSIM -simulation of vehicle movementsat 1 second resolution

! Explicit treatment of traffic patterns

# Turn lanes, signal phases, queuing

# Coordination of upstream signals

! Output: Speed/accelerationdistribution by 10 m roadwaysegment and signal phase

Traffic Module -Micro-simulation

Denver Speed-Acceleration Profile

-16

-12

-8 -4 0 4 8

12

16

0

20

40

60

0.0%

0.2%

0.4%

0.6%

0.8%

1.0%

Speed (fps)

Accelera t ion

(fps/s )

Id le = 38.6%

Idle = 38.6 %

Accel (fps/s)

Speed (fps)

Presentation 4 11



HYROAD Speed-AccelerationProfile (LOS C Test Case)

-12-9

-6-3

03

69

12

0

7

15

22

29

37

44

5159

66

0

100

200

300

400

500

600

700

800

900

1000

Vehic le-seconds

Acceleration (fps/s)

Speed (fps)

e

Idle = 32.3 %

Emission Module! Objective: make the best

possible use of cycle-basedemission factors

! Core rates from MOBILE5for the speeds of the speedcorrection cycles

! Multivariate regression weightsspeed correction cycles to matchnon-idle speed distribution fromNETSIM

Emission Module

Presentation 4 12



Emission Module! Total emissions calculated from

weighted average g/mi rate plusexcess idle (idle time not explainedby weighted cycles)

! Speed/acceleration distributionsused to calculate fuel consumptionby 10 m roadway segment andsignal phase

! Fuel consumption used as asurrogate to spatially andtemporally allocate emissions

Emission Module

HYROAD Speed Distribution

0

20000

40000

60000

80000

100000

120000

140000

160000

180000

200000

Idle 5 10 15 20 25 30 35 40 45 50 55 60Speed

Ve

hic

le s

eco

nd

s

(15

min

ute

sim

ula

tio

n)

Traffic Module

Emissions Module

Excess Idle Time Added

Presentation 4 13



CAL3QHC Speed Distribution

0

20000

40000

60000

80000

100000

120000

140000

160000

180000

200000

1 2 3 4 5 6 7 8 9 10 11 12 13Speed

Ve

hic

le S

eco

nd

s

Traffic ModuleCAL3QHC

Queued vehicle idle time

Induced Turbulenceand Flow Fields

! Based on ROADWAY(Eskridge, 1987)

! Turbulence and induced flowcalculated for each 10 mroadway segment using trafficvolume and speed

! Output for each signal phase:Gridded wind speed and eddydiffusivity for a 2.5 x 2.5 kmgrid of 10 m cells

Induced Turbulenceand Flow Fields

Presentation 4 14



Wind Field with Induced Flows

Dispersion Module! Gaussian model (moderate wind)

# Uses CALINE4 mixing zone, sigma-z, sigma-theta

! Puff model

# 1 puff per second per10 m roadway segment

# Puff transport with griddedwinds by phase

# Puff growth based on local stabilityfrom gridded eddy diffusivity

# Domain wind and stabilityused after 1 cycle

Dispersion Module

Presentation 4 15



Model Formulation -Conclusions

! HYROAD integrates acceptedmodeling approaches to treatimportant processes affectingintersection CO concentrations

# Induced flows and turbulence

# High spatial and temporalvariability of emissions

! Modular design allows updating(e.g., for modal emissions)

Model Formulation -Conclusions

Model Application -Resource Needs

! Standard inputs

# Intersection geometry(lanes, medians, etc.)

# Traffic (volume, turns,signal cycles, speeds, coordination)

Presentation 4 16



# Meteorology (wind speed, direction,stability, temperature, optional

sigma-theta

# Emission factors by temperature(constant cold-start), or hour-specific



! Input preparation time

# Geometry and NETSIM: 8 hours

# Emissions and Dispersion: 12 hours

! Run time(500 MHz Pentium, 128 Mb)

# Netsim: 30 sec / simulation hour

# Dispersion: 4 min / simulation hour

# Optimization possibilities


Presentation 4 17



HYROAD Sensitivityand Performance

! Sensitivity analyses

# Nominal intersection with gridof receptors in NW quadrant

# Modeling with bothGaussian and puff models

# Results for wind speed,wind direction, and stability

1 2 3 4 5 6 7 8

S1

S2

S3

S4

S5

S6

S7

S8

16-18

14-16

12-14

10-12

8-10

6-8

4-6

2-4

0-2

25 m

Concentration Patterns --Puff v. CALINE4

Puff Module CALINE4 Module

CO(ppm)

1 2 3 4 5 6 7 8

Wind


! Performance evaluation data sets

# Intensive data sets forthree intersections

" 528 hours with 15 minute data(6000 available)

" 10 or more receptor locations

Presentation 4 18



# 8 SLAMS/NAMS sites

" 1728 hours (75,000 available)

" Uncertain background concentrations



! Performance evaluation approach

# Concurrent evaluation of HYROAD(Gaussian and puff) with CAL3QHCwithout regulatory constraints(D, 1 m/s)

# Scatterplots, stratified by receptor,wind speed, and wind direction

# Standard statistics

" All data

" Max 25 paired

" Max 25 unpaired


Presentation 4 19



Predicted vs. Observed [CO]

-1

0

1

2

3

4

5

6

7

8

9

10

11

-2 -1 0 1 2 3 4 5 6 7 8 9 10 11

Observed (ppm)

Pre

dic

ted

(p

pm

)

HYROAD-Puff

HYROAD-CL4

CAL3QHC

Quantile-Quantile Plot

-1

0

1

2

3

4

5

6

7

8

9

10

11

-2 -1 0 1 2 3 4 5 6 7 8 9 10 11

Observed (ppm)

Pre

dic

ted

(p

pm

)

HYROAD-Puff

HYROAD-CL4

CAL3QHC

PreliminaryPerformance Results

! HYROAD produces more accuraterobust high concentration thanCAL3QHC

! Performance differences observedbetween receptor locations

! HYROAD appears to provideimproved treatment of problematicworst-case conditions

Presentation 4 20



PreliminaryPerformance Results

! Performance evaluation forother intersections is under way

! Screening methodology usingHYROAD will be developedand evaluated

Proposed Applications! Refined CO Applications

# SIPs

# Conformity

# EIS/EIR

! Particulate matter“hot-spots” assessment

! Air toxic risk assessment

Project Status! Current Status

# Select transition frompuff to line-segment

# Complete modelperformance evaluation

" Completed Tucson

" Virginia & Denver plus 8SLAMS/NAMS sites

Presentation 4 21



# Present Evaluation andRecommendations to NCHRP panel

Project Status

Next Steps! Develop Graphical User Interface

# Facilitate communicationbetween modules

# Ease burden for userin developing inputs

! Update Draft User Guide

# Reflect changes

! Schedule for Completion

# Complete Evaluation (August 2000)

# Develop GUI and Beta Testing(Fall 2000)

# Update User Guide(Winter 2000-2001)

Next Steps

Presentation 5 1

Technical Highlights of EPA’s 7th Conference on Air Quality ModelingThe Variable Grid Urban Airshed Model

Edward Carr

Edward Carr

The Variable Grid UrbanAirshed Model

(UAM-V, UAM-VPM)

Overview of the Variable-Grid Urban Airshed Model

(UAM-V)! Simulates the physical and

chemical processes governing theformation and transport of ozone inthe troposphere

" three-dimensional, Eulerian(grid-based) model

" requires specification ofmeteorological, emissions, land-use,and other geographic inputs

" output includes hourly concentrationsof ozone and precursor pollutants foreach grid cell within a(three-dimensional) modeling domain

Overview of the Variable-Grid Urban Airshed Model

(UAM-V)

Presentation 5 2


Edward Carr

Overview of UAM-V! Core model, supporting software,

user’s manuals, and examplemodeling database available fromSAI at no charge

! www.uamv.saintl.com

! Version 1.24 – OTAG version

" Updated isoprene chemistry (1996)

! Version 1.30 – Latest version

" Toxics chemistry, process analysis

Overview of UAM-V

UAM-V ModelingSystem Features

! Carbon-Bond-IV chemicalmechanism with enhancedisoprene and toxics chemistry

! Two-way interactivenested-grid capabilities

! Plume-in-grid (P-i-G) treatment

Presentation 5 3


Edward Carr

! Accepts output from a variety ofdynamic meteorological models(e.g. MM5)

! Contains “process analysis”capabilities

UAM-V ModelingSystem Features

Treatment ofProcesses in UAM-V

! Advective pollutant transport

" Smolarkiewicz scheme

! Turbulent diffusion

" Dispersion proportional toconcentration gradient – K Theory

! Surface removal

" Uptake of pollutants by varioussurface features - land use

! Chemistry

" Carbon Bond IV chemical mechanismwith updated isoprene chemistry andtoxics mechanism

Treatment ofProcesses in UAM-V

Presentation 5 4


Edward Carr

UAM-V Modeling System! Core model and input processing

software

! Emissions Preprocessing System(EPS2.5) to prepare ozone andparticulate emission inventories

! UAM-V Postprocessing System(UPS)

! Process analysis modelingsystem software

! Model Output Visualization andInput Evaluation Software(MOVIES) – Color animations

UAM-V Modeling System

UAM-V InputFile Requirements

! Meteorological input files

" wind

" temperature

" water-vapor concentration

" pressure

Presentation 5 5


Edward Carr

" vertical diffusivity(effective mixing height)

" cloud cover

" rainfall rate



! Emissions input files

" low-level anthropogenic emissions

# point sources

# area sources

# motor-vehicles

" elevated point source emissions

" biogenic emission estimates


! Air quality related input files

" initial conditions

" boundary conditions

! Chemistry input files

" chemical reaction rates

" photolysis rates

Presentation 5 6


Edward Carr

! Geographic/other input files

" land-use

" albedo, turbidity, and ozone column


UAM-V ProcessAnalysis Capabilities

! UAM-V process analysis providesdetailed information on thephysical and chemical simulationprocesses

! Process-level information includes

" photochemicalproduction/consumption

" horizontal advection/diffusion

" vertical advection

" vertical diffusion

" deposition

" emissions (for precursor pollutants)

UAM-V ProcessAnalysis Capabilities

Presentation 5 7


Edward Carr

Simulated Process Contributions to Ozone

-25-20-15-10

-505

101520

0 2 4 6 8 10 12 14 16 18 20 22

Chem Horiz. V. Adv. V Diff. Dep.

-25-20-15-10

-505

101520

0 2 4 6 8 10 12 14 16 18 20 22

Chem Horiz. V. Adv. V Diff. Dep.

Ozone

ppb

Hour

0

100

Simulated —and observedozone concentration

Simulated Process Contributions to Ozone

UAM-V Users/Applications! Over 60 registered users worldwide

(Version 1.30), unknown number ofunregistered users

" U.S. EPA, state/local agencies,and industry

" Environment Canada

Presentation 5 8


Edward Carr

" European research andregulatory agencies

" European and Japanese auto-makers; European oil companies

UAM-V Users/Applications

UAM-V Users/Applications! Registered users

" Research and regulatory groups inCentral and South America, Australia,New Zealand, and several Asiancountries including

# China

# Taiwan

# South Korea

# Philippines

# India

# Thailand (AIT)

UAM-V Users/Applications! Some completed applications

" Numerous U.S. regions/cities

# OTAG, Atlanta, Houston, BatonRouge, Chicago

" Vancouver,B.C.

" U.K.

" Paris

" Milan

" Mexico City

" Athens

Presentation 5 9


Edward Carr

Worldwide Users of theUAM-V Modeling System

(2)

(2)

(4)

(2)

(2)(3)

(3)(2)(2) (2)

(2) (2)

(3)

(2)

(5)

Current Applications! Gulf Coast Ozone Study (GCOS) –

Assessment of ozone formationand transport processes affecting1- and 8-hour ozone along the U.S.Gulf Coast

! Arkansas-Tennessee-MississippiOzone Study (ATMOS) –Assessment of potential 8-hourozone issues for Memphis,Nashville, Knoxville, Tupelo,Chattanooga, and Little Rock

! Mexico City - Demonstration for anew area/preliminary emissionssensitivity analysis

Current Applications

Presentation 5 10


Edward Carr

Example of a Multiple Nested Domain: Gulf Coast Ozone Study

4-kmgrids(GridsA,B & C)

2-kmgrid(Grid D)

36-kmgrid(Grid 1)

12-kmgrid(Grid 2)

Overview of theUAM-VPM

! UAM-V photochemical model(CB-2000)

! Particulate matter (PM) stand-alonebox model employing hybridmodal-sectional approach to PMrepresentation

! Gas-phase chemical mechanismgenerator for the UAM-V

Structure of the UAM-VPM! Features of PM dynamics that can

be well characterized by knownalgorithms are hard coded

! Features which are not well knownare user inputs or dynamicallyselected

! Allows for the best research gradealgorithms to be used in aregulatory and planning platform

Presentation 5 11


Edward Carr

PM Processes in UAM-VPM1) Modal discretization (new)

2) Nucleation(Fitzgerald, Hoppel, Gelbard, 1998)

3) Coagulation(Jacobson, 1994, 1999)

4) Condensation(Jacobson, 1997, 1999)

5) Dissolution (Jacobson, 1997, 1999)

6) Reversible chemistry (various)

7) Sectional remodalization (new)

PM Processes in UAM-VPM

Gas-to-particleTransformations

EmissionsInjected

(Gas ∧∧∧∧ PM)

HorizontalAdvection/Diffusion

(Gas ∧∧∧∧ PM)

VerticalAdvection/Diffusion

(Gas ∧∧∧∧ PM)

Deposition(Gas ∧∧∧∧ PM)

Chemistry(Gas Phase)

Gas Concentrations and PM Modes (low size resolution)UAM-VPM Flow Chart

DiscretizeModes

Nucleationadd mass to

smallest size bin

Solve homo-coagulationequations

Calculate diffusioncoefficients knudsennumbers, coagulation

kernels etc., for allcombinations of particles

Calculate diffusioncoefficients, knudsennumbers, coagulationkernels etc., for self-

similar particles

Solve hetero-coagulationequations

Determinedeliquescence points forsolids, and if below r.h.,eliminate reactions fromequilibrium mechanism

Equilibrate ions, solids, andgases to get ionic activitycoefficients (equilibrium

chemistry solver)

Determine watercontent w/Zdanovskii-

Stokes-Robinson (ZSR)relationship

DetermineeffectiveHenry’s

constants

Dissolution/Condensation(add or subtract mass)

Inert OrganicHenry’s constants, or

“effective” Henry’s constantssupplied by user

Solid ∧∧∧∧ AqueousInorganic

Calculate effectiveHenry’s constants

Solve Analytic Prediction ofDissolution (APD) for Each

Condensable Species

Solve Analytic Prediction ofCondensation (APC) for

Each Condensable Species

Renormalize SizeDistributionCoagulation

Homo-Coagulation

Hetero-Coagulation

Presentation 5 12


Edward Carr

ExampleUAM-VPM Species

H2O(aq) water Na+ sodium ion

H2SO4(aq) sulfuric acid Cl- chloride ion

HNO3(aq) nitric acid Na2SO4(s) sodium sulfate

NH3(aq) ammonia NaHSO4(s) sodium bisulfate

HCl(aq) hydrochloric acid NaCl(s) sodium chloride

H+ hydrogen ion NaNO3(s) sodium nitrate

OH- hydroxy ion (NH4)2SO4(s) ammonium sulfate

NH4+ ammonium ion NH4HSO4(s) ammonium bisulfate

NO3- nitrate ion NH4Cl(s) ammonium chloride

HSO4- bisulfate NH4NO3(s) ammonium nitrate

SO42- sulfate

ChemicalFormula

Chemical Name

ChemicalFormula

Chemical Name

Status of UAM-VPM: Plansfor 2000 and beyond

! Rigorous testing of box model

! Testing of full modeling system

! Complete initial application toVancouver for a 10-day 1993episode

! Initiate application to Alberta

Presentation 6 1

Technical Highlights of EPA’s 7th Conference on Air Quality Modeling7th Conference on Air Quality Modeling: Prepared Comments on the Second Day - June 29

Joseph A. Tikvart

7th Conference on AirQuality Modeling:

Prepared Comments on theSecond Day - June 29

Joseph A. Tikvart

AMS Committee PerspectiveJ. Weil

AWMA Committee PerspectiveR. Paine

Prognostic Meteorological Model PanelR. Schulze

STAPPA/ALAPCO AgenciesP. Hanrahan

Department of EnergyP. Lunn

Gas Research InstituteD.Blewitt

American Petroleum InstituteH. FeldmanK. Steinberg

Utility Air Regulatory GroupA. FieldR. Paine

Southern CompanyS. Vasa

Trinity ConsultantsR. Schulze

Personal StatementM. Sharan

Presentation 6 2


Joseph A. Tikvart

7th Conference on AirQuality Modeling

Questions EPA Asked on New Modeling Systems

Q1. Scientific Merit

Q2. Model Accuracy

Q3. Appropriate Regulatory Applications

7th Conference on AirQuality Modeling

Questions EPA Asked on New Modeling Systems

Q4. Implementation Issues

Q5. Resource Constraints

Q6. Additional Analyses

Notes on Summaryof Comments

! Public comment periodis open until 8/21/00

! Summary is for oral comments only

Presentation 6 3


Joseph A. Tikvart

! No responses have beenformulated at this time

! Comments on the scientificmerit of AERMOD, CALPUFF,and ISC-PRIME were generallyfavorable



! Summary of comments is basedon the previous 6 questions

" AERMOD

" CALPUFF

" Numerical Grid Models

" Data from Meteorological Models

" General Comments

! There were no specific oralcomments on other models


Presentation 6 4


Joseph A. Tikvart

AERMODQ1, Q2. Scientific Merit

& Model Accuracy

" Improvements over ISC, especiallyregarding PBL dispersion andcomplex terrain, are desirable


" Add PRIME and depositionalgorithms to AERMOD

" Develop AERSCREEN

AERMOD

AERMODQ4, Q6. Implementation Issues

& Additional Analyses

" Define separate uses ofAERMOD and ISC-PRIME

" Expand period for transitionto AERMOD (>12 months)

Presentation 6 5


Joseph A. Tikvart

Q5. Resource Constraints

" Application of PRIME -- separate fromAERMOD or included with AERMOD

" Use of electronic terrain data --need more training

" Additional EPA support

AERMOD

CALPUFFQ1. Scientific Merit

" Significant advancement as State-of-Practice for long range transport

" Model flexibility providesroom for growth

Q2. Model Accuracy

" Testing is adequate forinclusion in the Guideline

" More testing for short-rangeapplications is desirable

CALPUFF

Presentation 6 6


Joseph A. Tikvart

CALPUFFQ3. Appropriate

Regulatory Applications

" Appropriate for longrange transport (50 - 200km)

" Need more specific guidelinelanguage for other applications


" Clarify user's guideand default options

CALPUFFQ5. Resource Constraints

" Computer requirements

" User skills

" Additional EPA support

CALPUFFQ6. Additional Analyses

" Requirement for 5 years ofmeteorological data

" Dispersion coefficient treatment

" Example protocol

" Better chemistry for SOx/NOx

" Overall experience of dispersionmodel community

Presentation 6 7


Joseph A. Tikvart

Numerical Grid Models forUrban/Regional Scales

Q1. Scientific Merit

" Support removal of old grid modelfrom guideline

Q6. Additional Analyses

" Need more testing of Models-3/CMAQ

" Insure that models are in the publicdomain

Data fromMeteorological Models

Q1, Q2. Scientific Merit & Model Accuracy

" Use of data from meteorologicalmodels (e.g., RUC, MM5) is desirableand meets needs


" Applications need aresolution finer than 80 km


Presentation 6 8


Joseph A. Tikvart



" Need detailed resolution ofterrain and meteorological data

" CALMET can be used inconjunction with MM5 data

" What are the consequencesof using these data bases inregulatory programs


Q5. Resource Issues

" Need more modelers using RUC &MM5 meteorological model productsand greater exposure to data bases

" Need a repository of routine andeasily accessible prognosticmeteorological data (NCEP,RUC)

" Need training for MM5 and similarmeteorological models


Presentation 6 9


Joseph A. Tikvart

General Comments

Summary! Public comment period is open

until 8/21/00

! No responses have beenformulated at this time

! Comments on the scientific merit ofAERMOD, CALPUFF, and ISC-PRIME were generally favorable --implementation / resource issues

! There were no specificcomments on other models

! Next steps

Summary

Documents

Technical Highlights of EPA’s 7 Conference on Air Quality ... · iii Fax Question Sheet APTI Workshop T-029 Technical Highlights of EPA’s 7th Conference on Air Quality Modeling