VISTAS Modeling Overview

May 25, 2004

Mt. Cammerer, Great Smoky Mtns. National Park

VISTAS is evaluating visibility and sources of fine particulate mass in the Southeastern US

View West from Shining Rock Wilderness Area, NC

VISTAS Air Quality Modeling

Objectives:• Accurately represent meteorology, emissions,

and air quality MM5, SMOKE, CMAQ

• Model base year to support both regional haze and PM2.5 regulatory requirements

• Model future year and control strategies for regional haze states responsible for PM2.5 attainment

demonstrations

VISTAS Air Quality Modeling

• Phase I: Evaluate different model configurations for 3 episodes: Jan 02, July 99, July 01 Recommend annual modeling protocol Mar 04

• Phase II: Annual regional modeling 2002 base year modeling begins Feb 04 Future year control strategy runs completed 2005

2009 interim year 2018

2003 episodes: evaluate model performance using continuous speciated

data from VISTAS Focus sites (Smokies, Cape Romain, Raleigh)

Jan 2004Revised 2002 VISTAS Em Inv

Feb-Apr 2004MM5 Met runs

Apr 2004Draft 2018 National Inv

Sep 2004Revised 2002National Inv

Apr-Sep 2004Annual 2002 CMAQ modelperformance

Dec 042018 Base Run

Apr 2004:DDM in CMAQ

May-Oct 20042018 Emissions Sensitivity Runs

Sep 2004“Typical” 2002Modeling Inv

Oct-Dec 2004:Control Strategies and Inventories

Jan - Jun 20052003 episodes: em inv, met, aq model

Jan-Jun 20052018 Control Strategy Runs

Sep 2004CART:selectepisodes

July-Dec 2005:ObservationsConclusionsRecommendations

Jan 2005Interim Future Year Model Runs

Dec 2004Interim Future Year Inventories

Emissions, Meteorological, Air Quality Modeling Deliverables

State Regulatory Activities

Jan-Sep 2004Define BART sources

Sep 2004Identify BART controls

Draft 5/19/04

EPA- approvedModeling Protocol

June 2005Economic Analyses

Mar 2004Em Modeling QA + Fill Gaps

Jan 2004Met modelingprotocol

Mar 2004AQ modelingprotocol

Mar 2004AQ Phase I wrapup

Jan-Mar 2004Define inv growth and control assumptions

Dec 2004Revised 2002Base Run (model performance)

Oct 2004:Revised 2018Em Inv

Sept 2004MM5 Met Final Report Dec 2004

“Typical” 2002 Run (compare to 2018 runs)

Aug 2004 / Dec 2005Draft / Final Natural Background and Reasonable Progress Goals

Emissions Inventories• Objective:

– Define emissions for PM2.5 and regional haze modeling• Responsibility: Stella, MACTEC, Pechan• Deliverables:

2002 1st draft inventory – Aug 20032002 2nd draft inventory – Jan 20042018 1st draft inventory – Apr 2004– 2002 revised (3rd draft) inventory – Sep 2004– 2002 revised plus “typical” EGU and fires – Sep 2004– 2018 revised plus BART and “typical” EGU and fires – Oct 2004– 2009 draft inventory – Oct 2004

Meteorological Modeling• Objective:

– Accurately represent meteorological conditions for atmospheric modeling using MM5 model

• Responsibilities: Baron Applied Meteorological Systems• Deliverables:

Evaluate meteorological model for 3 episodes – Dec 2003Recommend met modeling protocol for 2002 – Dec 2004Deliver met runs for 2002 – Mar 2004– Deliver met runs for selected 2003 episodes – fall 2004 – Reporting and data transfer - ongoing

Emissions Modeling• Objective:

– Prepare inventories for atmospheric modeling using SMOKE model, including temporal and spatial allocation

• Responsibilities: Alpine Geophysics• Deliverables:

Emissions for 3 Phase I modeling episodes – Sep 032002 emissions for VISTAS states, US, Canada, Mexico –

Mar 04– 2002 revised and “typical” emissions – Oct 042018 initial and revised emissions – Apr and Nov 04– 2009 emissions – Dec 04

Air Quality Modeling• Objectives:

– Accurately represent air quality using CMAQ model – Evaluate benefits from emissions controls

• Responsibilities: ENVIRON, University California Riverside, Alpine Geophysics

• Deliverables:Evaluate model configuration for 3 episodes – Feb 04Recommend model protocol for annual 2002 run – Mar 04– Model 2002 using initial, revised, and revised plus typical

EGU and fire emissions – Apr to Nov 04– Model 2009 and 2018 with typical emissions – Dec 04– Model control strategies – Jan to Jun 05

Emissions Sensitivity Modeling• Objective:

– Evaluate air quality responses to modeled changes in emissions to inform emissions strategy design

• Responsibilities: Georgia Tech• Deliverables:

Demonstrate Decoupled Direct Method for aerosols (DDM-AERO) in CMAQ – Apr 04

• Time savings offset by added memory requirement, not continued– Evaluate emissions changes using Jul 01 and Jan 02 episodes and

2018 inventory • by source sector, geographic area – Oct 04

Characterizing Meteorology

• Objective:– Characterize relationship between meteorology, PM2.5,

visibility, for IMPROVE and STN sites, using Classification and Regression Tree (CART) analyses

– Recommend episodes in 2002 or 2003 for focused modeling analyses

• Responsibility: Systems Application International • Deliverables:

Initial CART analyses for IMPROVE sites – Apr 04– Initial CART analyses for STN sites – Jun 04– Recommend options for episodes – Aug 04

CMAQ Model Configuration

CMAQ Version 4.3 (August 2003)• Horizontal Advection and Vertical Advection

Piecewise Parabolic Method (PPM) • Gas-Phase Chemistry: CB-IV with MEBI/Hertel Solver• Aerosol Chemistry: AE3 / ISORROPIA / SORGAM• Aqueous-Phase Chemistry: RADM• Dry Deposition: Pleim-Xiu• MM5 Configuration:

Pleim-Xiu/ACM Soil/PBL models with MCIP2.2 Pass Through• 36 km national grid, 12 km VISTAS domain (eastern US)

VISTAS 36-km and 12-km CMAQ Modeling Domains

CMAQ Phase I Sensitivity Tests

• Vertical Layers: 19 vs 34 layers• Boundary Conditions: Ultra Clean, GEOS-CHEM• Ammonia Emissions: Jan 50% reduction, diurnal pattern• Vertical Diffusivity - Minimum Kz: 0.1 or 1.0 m2/s• Alternative meteorology - with Wesley Dry Deposition • Mass Conservation Patch• SAPRC-99 chemistry• CB-2002 chemistry• CB-AIMS chemistry with sectional approach• CAM-x trial with comparable configuration • Best and Final CMAQ configuration

CMAQ Configuration: Annual Modeling

• Lambert Comformal– 36 km origin (-2736, -2088) – 12 km origin (108, -1620) – 1-way nesting

• Vertical Diffusivity: Minimum Kz: 1.0 m2/s

• Boundary Conditions: – 2001 GEOS-CHEM seasonal averages initially– replace fall 2004 with 2002 GEOS-CHEM 3 hr ave

CMAQ Configuration: Annual Modeling

• Emissions used in Apr 04: – VISTAS states: 2002 state inventories

• Fires from state and federal records• Fires modeled as point sources where sufficient data

– Non-VISTAS states: NEI 2002 (Apr 04 release)– Canada and Mexico as available– Ammonia from CMU (Jan 04 version) with updates– Biogenics with BEIS3– Temporal profiles from SMOKE except:

• Point sources: CEM hourly profiles• MOBILE and NONROAD: 1 week/month

Performance Evaluation

• IMPROVE – daily • STN – daily 1 in 3 or 1 in6• SEARCH – daily and hourly• CASTNET – weekly• NADP – weekly• AQS – hourly

www.vistas-sesarm.org

www.cert.ucr.edu/vistas

2004 Modeling Workplan Tasks

1. Project Management2. IC/BC analysis: GEOS-CHEM3. Data Management

a. Air Quality Gatekeeper b. Meteorological Gatekeeper c. Emissions Gatekeeper d. Emissions Modeling: 2002, 2018 for annual modeling e. Emissions Modeling: 2018 for emissions sensitivity f. Emissions Modeling: 2009

2004 Modeling Workplan Tasks

4. CMAQ Modeling a. 2002 Initial Inventory b. 2002 Revised Inventory c. 2002 revised plus “typical” Fire and EGU d. 2018 “typical” e. 2009 “typical”

5. 2003 episodes6. Website, Data Transfer, Storage, Retrieval

2004 Modeling Workplan Tasks

7. Evaluate CB-IV vs SAPRC for responses to emissions changes

10. Journal publication Phase I13. 2002 GEOS-CHEM outputs prepared for

CMAQ IC/BC inputs15. Contingency

• Optional Tasks• Extended Performance Evaluation• Emissions Sensitivities• Source Apportionment/Tracers • Natural Background Analysis• Comparison of Alternate Models

• 2005 workplan focus: modeling control strategies

2004 Modeling Workplan Tasks