SEMAP Modeling Plans and Status James Boylan Georgia EPD – Air Protection Branch SEMAP Air Quality Modeling Lead 2012 Spring Grants/Planning Meeting May 23, 2012 – Atlanta, GA

SEMAP Project SouthEastern Modeling, Analysis, and Planning (SEMAP) Project Managed through SESARM Same group of states that were involved with SAMI, VISTAS, and ASIP AL, FL, GA, KY, MS, NC, SC, TN, VA, WV

Outline Emission Inventory Development WRF, SMOKE, CMAQ Future Year Projections for O3, PM2.5, RH Ozone Emission Sensitivities Regional Haze Periodic Reporting

Important Note SEMAP states do not have any moderate or serious ozone nonattainment areas & all PM2.5 monitors are meeting the NAAQS. Regional Haze periodic reports do not require photochemical modeling. Most likely, none of this modeling will be used in any SIPs.

Emission Inventory Emissions Inventory Contractors AMEC/Alpine  Point source (EGU and non-EGU), fire, and on-road mobile (MOVES) emissions. TranSystems  Area and non-road (including MAR) emissions. “Actual” Emission Inventory (2007) Used for model performance evaluations “Typical” Emission Inventory (2006-2008) EGUs and fires Used for model projections and RRF calculations Future Year Inventory – 2020 Growth & Control factors for EGU point, non-EGU point, area, and MAR Rerun NONROAD Inventory mode ratio method for MOVES (proxy FY) Inventory mode ratio method for MOVES (proxy FY): SMOKE ready emissions by each state (could be more than one county); can be different ratio for different areas in a state; for GA four PM2.5 maintenance areas done (ATL, Macon, Floyd, Chattanooga) and additional run for the rest of state

Point Source NOx

Point Source SO2

MOVES Emissions New tools had to be developed for QA Rate Per Distance Rate Per Profile Rate Per Vehicle

NOx from Wildfires

NOx from Ag, Open, Rx Burning

Non-Road and Area NOx

Non-Road and Area SO2

EGU Point SO2

Meteorological Modeling WRF Model Set-Up 36-km CONUS and 12-km Eastern US 35 Vertical Layers Four-Dimensional Data Assimilation Morrison (moisture physics), PX (land surface), and ACM2 (PBL) Modeling Centers NC DENR and IA DNR Model Performance Evaluation Atmospheric and Environmental Research (AER) NC DENR and GA EPD

WRF Modeling Domain

Surface Layer

Air Quality Modeling System Meteorology (WRF) Air Quality (CMAQ) Emissions (SMOKE) Emissions Inventory (NIF) MOVES Inputs

Emissions Modeling GT/UNC - SESARM contractor Created SMOKE inputs (NIF to ORL format) Ran SMOKE and SMOKE-MOVES Created biogenic emissions (BEIS) Added Canada/Mexico and other RPO Emissions Part 75 data replacement for CEMS Details on next slide All 2007 SMOKE files are finished and ready to be used.

1. Detect and correct atypical data 2. Gap filling for partial reporters

2007 Air Quality Modeling GT/UNC - SESARM contractor 2007 annual modeling with CMAQv5.0 36 and 12 km grids Model Performance Evaluation Diagnostic Sensitivity Testing (only if MPE poor) New CMAQv5.0 Model Options Additional new crustal species  YES In-line plume rise for point sources  YES In-line lightning NOx  YES In-line bi-directional NH3 flux  NO In-line biogenic emissions  NO In-line wind blown dust  NO

CMAQ Modeling Domain

SEMAP 12-km Modeling Domain

2020 Air Quality Modeling 2020 Annual Emisisons Modeling SMOKE for most source sectors MOVES: Ratio Approach using Inventory Mode 2020 Annual CMAQ Modeling Projections of ozone, PM2.5, and regional haze Follow EPA’s RRF approach Ozone Emission Sensitivity Modeling Statewide 30% NOx and VOC emission reductions Ozone season (5 months) at 12 km Performed for 10 individual SEMAP states, Maryland, and neighboring RPOs (28 model runs) For all sensitivities, CAMx with DDM could be better.

Georgia NOx vs. VOCs Surface NOx reductions were made in Atlanta, therefore small impacts on Augusta since wind was blowing from SC on the two RRF dats. Larger NOx impacts from SC on Augusta during those two RRF days. VVOC=Vistas VOC 4 11 23 2 3 6 6 Red numbers indicate number of days used in RRF calculations

VOC/NOx Sensitivity Updates Previous Sensitivity Modeling Planned Sensitivity Modeling CMAQv4.4 CMAQv5.0 1-Month Summer Episode 5-Month Ozone Season VISTAS 2009 OTW BaseD SEMAP 2020 SEMAP-wide VOC Reductions State-wide VOC Reductions County/State NOx Reductions State-wide NOx Reductions MOBILE6 MOVES Absolute Difference RRF Approach >70 ppb Cutoff >60 ppb Cutoff

SEMAP Modeling Schedule

Regional Haze Obligations 5-year Periodic Reports First reports due December 2012 (NC, SC) 2018 Regional Haze SIP Revisions Due July 31, 2018 Work must begin by mid-2014 For all sensitivities, CAMx with DDM could be better.

SEMAP Regional Haze TSD Current Visibility Conditions Examine 2000- 2010 IMPROVE Measurements Emission Inventory Changes VISTAS 2002 (actual), 2007 SEMAP (actual), VISTAS 2009 (projected), and VISTAS 2018 (projected) Update 2007 SEMAP with 2009, 2010, 2011 CEMS Status of Emission Controls in SIP Federal and State Controls Assessment of Current SIP For all sensitivities, CAMx with DDM could be better.

Regional Haze at COHU Surface NOx reductions were made in Atlanta, therefore small impacts on Augusta since wind was blowing from SC on the two RRF dats. Larger NOx impacts from SC on Augusta during those two RRF days. VVOC=Vistas VOC

Regional Haze SO2 Emissions Surface NOx reductions were made in Atlanta, therefore small impacts on Augusta since wind was blowing from SC on the two RRF dats. Larger NOx impacts from SC on Augusta during those two RRF days. VVOC=Vistas VOC

Contact Information Jim Boylan, Ph.D. Georgia Dept. of Natural Resources 4244 International Parkway, Suite 120 Atlanta, GA 30354 James.Boylan@dnr.state.ga.us 404-362-4851