Source Attribution for Southeastern US: Weight of Evidence June 9, 2005 National RPO Meeting Denver, CO Pat Brewer VISTAS Mt. Cammerer, Great Smoky Mtns. National Park

Regional Haze Rule requires states to protect visibility in 156 “Class I” natural parks and wilderness areas. VISTAS is evaluating visibility and sources of fine particle mass in the Southeastern US I want to welcome y'all to Asheville. I am very grateful to live in WNC. This photo is of Cold Mountain on the morning after Hurricane Ivan went through last month. This is probably as close to natural background as I expect to see in So App. Our field trip on Wed afternoon will be to Shining Rock Wilderness Area and the beginning of the trail to Cold Mtn. We will be passing IMPROVE site for Shining Rock. Cold Mountain, Shining Rock Wilderness Area

Weight of Evidence Monitoring Trends SAMI Modeling Results for 2010 Emissions Sensitivities using 2018 initial inventory for July 2001 and Jan 2002 episodes Back Trajectory and Residence Time Analyses Meteorological Classification – CART analyses Receptor Modeling (MRPO-MARAMA 2003) Regional Air Quality Modeling (CMAQ, CAMx, REMSAD) Evidence for Role of Fire

There are 18 Class 1 areas within VISTAS region, IMPROVE operates monitors at 15 of the 18 Class I areas.

Light Extinction on 20% Haziest Days - IMPROVE 2000 - 2002 250 200 CM Soil 150 Organics EC Extinction (Mm-1) 100 NH4NO3 (NH4)2SO4 50 Rayleigh On the Haziest days NH4SO4 dominates. At the VISTAS coastal sites (illustrated on right) NH4SO4 ranges from 50 to100 Mm-1. At the So App sites (illustrated on left), NH4SO4 ranges from 100-150 Mm-1. In contrast, on the haziest days, OC contributes 10-30 Mm-1 NH4NO3 contributes 5-10 Mm-1, less from EC, soils, and coarse mass. Cadiz, KY Sipsey, AL Dolly Sods, WV Saint Marks, FL Shenandoah, VA Okefenokee, GA Everglades, FL Mammoth Cave, KY Shining Rock, NC Linville Gorge, NC Swan Quarter, NC Cape Romain, SC Chassahowitzka, FL James Rvier Face, VA Great Smoky Mtns, TN

Light Extinction on 20% Clearest Days - IMPROVE 2000 - 2002 10 20 30 40 50 60 CM Soil Organics Extinction (Mm-1) EC NH4NO3 (NH4)2SO4 Rayleigh On clearest days, NH4SO4 10-25 Mm-1, OC and NH4NO3 are 5-10 Mm-1 Cadiz, KY Sipsey, AL Dolly Sods, WV Saint Marks, FL Shenandoah, VA Okefenokee, GA Everglades, FL Mammoth Cave, KY Shining Rock, NC Linville Gorge, NC Swan Quarter, NC Cape Romain, SC Chassahowitzka, FL James Rvier Face, VA Great Smoky Mtns, TN

SAMI Final Report 2002 SO4 Fine Particle Response (%) Annual SO4 Fine Particles Response to 10% Reduction in SO2 Emissions from 2010 A2 strategy SO4 Fine Particle Response (%) -8.0 -6.0 -4.0 -2.0 0.0 Sipsey, AL Cohutta, GA Joyce Kilmer, NC Great Smoky Mtn, TN Shining Rock, NC Gorge, NC Linville James River Face, VA Shenanhoah, VA Otter Creek, WV Dolly Sods, Non-SAMI states SAMI states usedUrban-Regional Multiscale (URM) Model and 9 episodes 1991-1995

SO4 & its Change on July 15, 1995 for a 10% Reduction of 2010-A2 SO2 Emissions from Different Regions SAMI 2010-OTW Midwest Northeast Central FL + MS mg/m3 +0.3 +0.1 - 0.1 - 0.3 - 0.5 - 0.7 - 0.9 - 1.1 (Class 5 day)

SO4 & its Change on July 15, 1995 for a 10% Reduction of 2010-OTW SO2 Emissions from SAMI States KY WV VA TN 2010-OTW NC AL GA SC +0.15 +0.05 - 0.05 - 0.15 - 0.35 - 0.45 - 0.55 mg/m3 (Class 5 day)

SO4 & its Change on May 27, 1995 for a 10% Reduction of 2010-OTW SO2 Emissions from SAMI States KY WV VA TN 2010-OTW NC AL GA SC +0.15 +0.05 - 0.05 - 0.15 - 0.35 - 0.45 - 0.55 mg/m3 (Class 4 day)

VISTAS 12-km CMAQ Modeling Domain

SO2 Point Sources emitting > 5,000 tons per year 2002 Inventory Annual SO2 emissions 250,000 125,000 25,000

2018 Emissions Sensitivities In 2004, initial 2002 and initial 2018 inventories modeled by Georgia Tech for July 2001 and Jan 2002 episodes. For Southern Appalachian sites, 20% worst visibility days in summer, 20% best in winter For coastal sites, 20% best and worst days could be in either summer or winter Can interpret relative importance of pollutant or source sector reductions Not all Key meteorological conditions were represented

Great Smoky Mtns (TN) – 2 days Weighted B ext Response on 20% Haziest Days -80.0 -60.0 CM ) SOIL -1 m EC M -40.0 ( OC ext NO3 B D -20.0 NH4* SO4 0.0 All SO2 BCSO2 BCSO4 All Ground SO2 All Elevated SO2 Non-VISTAS SO2 VISTAS Coal Power Plant SO2 VISTAS Non Power Plant SO2 VISTAS other power plant SO2 30% Reduction from 2018 OTB

Elevated SO2 Emissions

VISTAS CPP SO2 Emissions

VISTAS NPP SO2 Emissions

Weighted Bext Response on Worst 20% days (July 01) 30% Reduction in SO2 emissions from 2018 OTB For Coal Fired Power Plants from VISTS states Great Smoky Mtns -7.0 -6.0 CM -5.0 ) SOIL -1 m -4.0 EC M ( -3.0 OC ext NO3 B -2.0 D NH4* -1.0 SO4 0.0 1.0 AL TN VA WV FL GA KY MS NC SC

in Elevated SO2 from Coal Fired Power Plants Bext Response to 30% Reduction from 2018 OTb in Elevated SO2 from Coal Fired Power Plants Great Smoky Mtn. Georgia North Carolina -20.0 -15.0 -10.0 -5.0 0.0 Jul 15 Jul18 Jul 21 Jul 24 Jul 27 D B ext (Mm -1 ) -20.0 -15.0 -10.0 -5.0 0.0 Jul 15 Jul18 Jul 21 Jul 24 Jul 27 D B ext (Mm -1 ) CM SOIL EC OC NO3 NH4* SO4

GA & NC CPP SO2 Emissions

Back Trajectories for Great Smoky Mtns. National Park (ending at 12 noon at 100 m)

Back Trajectories for Great Smoky Mtns. National Park

Average PM2.5 Components for Key Visibility Bins - Great Smoky Mtns., TN 12.00 8.00 EC NO3 Avg. Conc. (µg/m3) OM SO4 4.00 For Smokies, Class I clear days to Class 5 haziest days, PM2.5 dominated by SO4, OC second most important. 0.00 Class 1 Class 2 Class 3 Class 4 Class 5 (20% clearest days) (21-50%) (51-80%) (20% haziest days) (Days ranked in bins and classes from clearest to haziest days)

Categorical Comparisons Surface Characteristics: All Days (2000-2003)/5 Ex. Coeff. Categories * *Knoxville

Key Visibility Bins: 20% Worst Days Great Smoky Mtns. Bin 39 (7 days): July 18 (31.3%) Bin 38 (11 days): July 21 (68.7%) Surface Characteristics: All Days (2000-2003)/5 Ex. Coeff. Categories * *Knoxville

Residence Time based on Back Trajectories – All Days 2000-2003 Great Smoky Mtns. National Park

Residence Time – 20% Worst Days 2000-2003 Great Smoky Mtns. National Park

Residence Time Difference Worst Days – All Sample Days 2000-2003 Great Smoky Mtns. National Park

Probable Source Areas for Secondary Sulfate (work by Batelle for Midwest RPO, provided to VISTAS Jan 2003) Composite probability is illustrated here using results from Batelle study for Midwest RPO and MARAMA. This is the probable source area for days when secondary sulfate is elevated at Shenandoah National Park. Apologize if experts in the room would phrase the explanation differently.

Residence Time – 20% Worst Days 2000-2003 Shenandoah National Park

2002 MANE-VU Weight of Evidence Relative RPO Contributions (to sulfate only; not including Canada yet) Acadia Brigantine LyeBrook Shenandoah

Uniform Rate of Reasonable Progress Glide Path Shenandoah, VA – 20% Worst Days MRPO VISTAS 2018 OTW-d

Percentage Reduction in 2018 (OTW-d) Compared to Reasonable Progress Goal Southern Appalachian Sites 0% 50% 100% 150% DOSO SHRO LIGO SHEN GRSM JARI COHU SIPS MACA Percent Reduction Achieved Assumes Method IB: average of 20% worst days in 2000-2003 with monthly RH

Average PM2.5 Components for Key Visibility Bins - a) Shining Rock, NC b) Cape Romain, SC a) 4.00 8.00 12.00 EC Avg. Conc. (µg/m3) NO3 OM SO4 0.00 b) Class 1 Class 2 Class 3 Class 4 Class 5 12.00 Slightly different story for Shining Rock here in WNC where one of the haziest days in the record was highly influenced by OC. Fire influence. Cape Romain, one of most difficult sites to catalogue. OC more comparable to SO4 on haziest days We can look at days in each bin to better understand which met conditions most important. 8.00 Avg. Conc. (µg/m3) 4.00 0.00 Class 1 Class 2 Class 3 Class 4 Class 5 (20% clearest days) (21-50%) (51-80%) (20% haziest days)

2002 20% Haziest Days at Swanquarter, NC Observations vs CMAQ 2002 Typical Typ_36_3 Fire not repesented in model? Model performance example for coastal site within +/-10% of reasonable progress goal.

2002 20% Haziest Days at Swanquarter, NC Reduction 2018 OTWc - 2002 Typical 2002 Typ_36_3 20% worst days occur between Jan and Dec; NO3 contributes on winter days

Example Back Trajectory Analysis for Day with Poor Model Performance (July 7, 2002) Swanquarter, NC HYSPLIT using EDAS (observations) or MM5 (model) meteorology

2002 20% Haziest Days at Brigantine, NJ Observations vs CMAQ 2002 Typical Typ_36_3 Model performance example for non-VISTAS site > 10% below reasonable progress goal.

2002 20% Haziest Days at Brigantine, NJ Reduction 2018 OTWc - 2002 Typical 20% worst days occur between Jan and Dec; mostly responses to SO4 and NO3

Example Back Trajectory Analysis for Day with Poor Model Performance (July 7, 2002) Brigantine, NJ HYSPLIT using EDAS (observations) or MM5 (model) meteorology

2002 20% Haziest Days at Everglades, FL Observations vs CMAQ 2002 Typical Typ_36_3

Julian Date 125

Julian Date 152

Example Back Trajectory Analysis for Day with Poor Model Performance (June 7, 2002) Everglades, FL Julian date 158 HYSPLIT using EDAS (observations) or MM5 (model) meteorology

Example Back Trajectory Analysis for Day with Poor Model Performance (Sep 14, 2002) Everglades, FL Julian Date 257 HYSPLIT using EDAS (observations) or MM5 (model) meteorology

Percentage Reduction in 2018 (OTW-d) Compared to Reasonable Progress Goal Southern Coastal Sites 150% 100% Percent Reduction Achieved 50% 0% CHAS SWAN SAMA EVER ROMA OKEF