Use of National PM2.5 and Speciation Network Measurements

Use of National PM2.5 and Speciation Network Measurements

Use of National PM2.5 and Speciation Network Measurements for Model Evaluation For presentation at PM Model Performance Workshop February 10-11, 2004: Research Triangle Park, NC Neil Frank OAQPS/USEPA Scope Networks FRM EPA Urban Speciation (aka STN) IMPROVE CASTNET Measurements PM2.5 mass Major Chemical Components Sulfur and Nitrogen Species Data interpretation related to Model Evaluation National Speciation Networks IMPROVE EPA spec. CASTNET filters + HNO3 denuder

filters + HNO3 denuder Simple Filter pack Main Purpose Visibility PM NAAQS Deposition Time avg 24-hr 24-hr weekly Particle size PM2.5 PM2.5 >=PM2.5 Frequency 1 in 3 days 1 in 3 / 1 in 6 complete

No. Sites 110 + 54=164 54 + 186 = 240 87 ? Sampler types 1 6 1 Reporting Local conditions Local conditions STP (&local) Sulfates Ambient Ambient Ambient <=ambient? <= ambient

Ambient <=ambient Ammonium pNO3 Ambient HNO3 - - >=ambient TNO3 - - ~ambient OC/EC DRI method (TOR) EPA/NIOSH (TOT) - Crustal

Estimated from Al, Si, Ca, Fe, Ti same - CASTNET Network Dense in Ohio Valley and Eastern US 8 Supplementary PM2.5 Speciation Sites (1993-2002) PM2.5 speciation sites (became IMPROVE in 2001 ) FilterPacks IMPROVE Network, 2002 98 Sites with complete data, Mar 01-Feb 02 Sulfate Variation in Rural Areas Comparison of CASTNET and IMPROVE Measurements in 2001 CASTNET 2001 IMPROVE March 01 Feb 02 Sulfate 0 - 0.4 0.4 - 0.5 0.5 - 0.7 0.7 - 1 1 - 1.3

1.3 - 1.7 1.7 - 2.2 2.2 - 3.5 3.5 - 4.2 4.2 - 5 Concentrations, STP Nitrate Variation in Rural Areas March 01 Feb 02 CASTNET CASTNET 2001 2001 IMPROVE March 01 Feb 02 Nitrate 0 - 0.1 0.1 - 0.2 0.2 - 0.3 0.3 - 0.4 0.4 - 0.5 0.5 - 0.6 0.6 - 0.7 0.7 - 1 1 - 1.5 1.5 - 2.5 Concentrations, STP Better resolution will come from new ( 2002-2003) data Gradient is overstated

HNO3 exhibit different spatial pattern TNO3 also shows a MidWest to East gradient NO3 HNO3 Concentrations, STP 2.7 + 2.1 = 4.8 ug/m3 0.8 + 2.0 = 2.8 ug/m3 NO3 HNO3 NO3 TNO3 HNO3 TNO3 CASTNET Comparisons Sulfates and Nitrates Western and Eastern Sites Ames RB, Malm WC (2001) Comparison of sulfate and nitrate particle mass concentrations measured by IMPROVE and the CDN. ATMOSPHERIC ENVIRONMENT 35 (5): 905-916. Sulfate: Comparison of 4-week mean IMPROVE and CASTNET Great Agreement for Sulfates (after adjustment to LTP) Particle Nitrate: Comparison of 4-week mean IMPROVE and CASTNET

Relative Bias for particle Nitrates /DC 20 Nitrate difference as a % of CM D NO 3 - / CM (%) and Temperature ( o C) 18 BBE Temperature 16 PIN 14 CHA SEK CAN 12 YOS

10 MOR 8 ROM GLR MEV GRB GRC LAV PND YEL SEK 6 PIN 4 BBE MEV YOS LAV GRB GRC ROM PND CAN YEL MOR 2

0 GLR CHA -2 -50 0 50 100 150 D NO 3 - (%) From: Rodger B. Ames and William C. Malm Comparison of sulfate and nitrate particle mass concentrations measured by IMPROVE and the CDN CASTNET filterpak and PM2.5 nitrates in the Eastern US: better agreement in Midwest and during later years Hypothesis: Better NH4NO3 retention on teflon with free ambient NH3 4.0 9.0 8.0 IL 7.0 6.0 BVL filterpak amm_nitrate (PM2.5) 5.0

4.0 3.5 OH 3.0 2.5 QAK filter pak ammonium_nitrate 2.0 IMPROVE 3.0 1.5 2.0 1.0 1.0 0.5 0.0 4.5 5.0 4.0

4.5 3.5 KY 3.0 4.0 3.5 3.0 2.5 CDZ filter pak ammonium_nitrate 2.0 2.5 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0.0 01-2

00-3 99-4 99-1 98-2 97-3 96-4 96-1 95-2 94-3 93-4 0.0 01-2 00-4 00-2 99-4 99-2 98-4 98-2

97-4 97-2 96-4 96-2 95-5 95-2 94-4 94-2 93-4 0.0 NW PA MKG filter pak ammonium_nitrate Less NO3 Agreement between CASTNET filter pack and PM2.5 nitrate as the sites move to the East Hypothesis: Poorer NH4NO3 retention on teflon in NH3 limited environments 3.5 3.0 2.5 2.0 CTH filter pak ammonium_nitrate 1.5

1.0 0.5 0.0 6.0 5.0 4.0 ARE filter pak ammonium_nitrate 3.0 Note: different than pNO3 loss with FRM measurements for PM2.5 2.0 01-2 00-3 99-4 99-1 98-2 97-3 96-4 96-1 95-2 94-3

0.0 93-4 1.0 Quarterly Average NO3: CASTNet vs PM2.5 Speciation Good agreement in rural IL. Relative Bias at Western NYS PM2.5 Speciation Bondville, IL CT H, NY 10 3 1:1 8 3:2 2 6 4 1 2 0 0

0 2 4 6 8 10 0 CASTNET NO3, ug/m3 1 2 3 Routine Estimates of Ambient Carbon More Uncertain than other measurements Carbon Inter-network differences in Measured C IMPROVE and STN use different thermo-optical techniques to measure carbon Many studies suggest that IMPROVE EC~=2x STN EC More recent results reveal more agreement Total Carbonaceous Mass is estimated as TCM = k* OC +EC Where k can be 1.2 to > 2.5 (+/- 30% regional uncertainty) IMPROVE uses 1.4

OC is blank corrected for artifacts using network-wide estimates but still sufficiently robust for model evaluation Blank corrections vary by Sampler Can Represent Substantial Portion of Measured Values * National Air Quality and Emission Trends Report 2003 Special Studies Edition http://www.epa.gov/oar/aqtrnd03/ IMPROVE MASS R&P MetOne (SASS): 9.6 m3 Anderson (RASS): 10.4 m3 R&P: 14.4 m3 URG (MASS): 24 m3 IMPROVE: 32.8 m3 OC blank, ug/m3 RAAS

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 Preliminary OC Blank Corrections Used by Rao, Frank, et al* SASS Derived from network average quartz filter field blanks to adjust annual averages Varies by 24-hr sampler volume Newest Blank Corrections Are Slightly Different Values used by Rao et al. 0.56 0.93 1.28 1.4 Newest Blank Corrections OC now corrected with Total Carbon value Reference: Analysis of Speciation Network Carbon Blank Data DRAFT REPORT, Flanagan et al

RTI International August 30, 2002 OC correction, ug/m3 1.8 1.6 1.4 1.2 1 For this analysis Newest estimates 0.8 0.6 0.4 0.2 0 SASS RAAS R&P MASS Evaluation of OC blank correction for 12-month averages using measured OC vs PM2.5 mass OC w Blank Adjustment smaller intercept OC wo Blank has adjustment OC concentration, ug/m3 10

y =y0.26x + 0.92 = 0.23x + 0.12 8 6 4 2 0 0 55 10 15 20 10 15 20 PM2.5 Mass Concentration, ug/m3 25 25 TCM Variation in Rural Areas March 01 Feb 02 Total Carbon Mass 0.7 - 1 1 - 1.3 1.3 - 1.5 1.5 - 1.8 1.8 - 2 2 - 2.6 2.6 - 3.3 3.3 - 4.2 4.2 - 5.4

5.4 - 7.0 TCM=1.8*OC+EC Based on TCM= OC*1.8+EC 13 Selected Urban Sites are Paired with Rural Sites for Urban PM2.5 Excess Calculations Missoula Cleveland Fresno SLC Indy S.L. Tulsa Birmingham 13 urban STN sites 16 rural IMPROVE sites Bronx Baltimore Richmond Charlotte Atlanta Urban PM2.5 is Higher than Nearby Rural Concentrations Gravimetric Mass 30 Bottom:

Regional Contribution Top: Urban Top: Urban Increment Bottom: Rural 20 15 10 5 Urban Increment Regional Contribution 12-month average PM2.5 mass from speciation samplers - PM2.5 STN mass is affected by high filter blanks prior to ~August 2001 Bronx/BRIG Baltimore/DOSO Richmond/JARI Charlotte/LIGO Cleveland/MKGO Atlanta/2 Sites Indy/LIVO Birmingham/SIP

St.Louis/3 Sites Tulsa/WIMO SLC/GRBA Missoula/MONT 0 Fresno/PINN ug/m3 25 Estimated Annual Urban Excess for Baltimore, MD Dolly Sods, WV Urban Excess Baltimore MD Rural Rural Concentrations Concentrations = Urban Rural concentrations Adjusted Superimposed for Elevation onsite differential Urban Rural IMPROVE site

STN urban (background) 9 9 6 3 Top bars are urban concentrations 6 Bottom bars are nearby rural concentration 3 3 0 0 0 Sulfate Ammonium Nitrate TCM1.8 Crustal

Sulfate Ammonium Sulfate Ammonium Nitrate Nitrate TCM1.8 TCM1.8 Crustal Crustal Crustal Ambient Urban Excess Concentrations for 13 example areas S u lf a t e : M is s o u la 0 .0 0 .4 0 .9 A m m o n iu m : 0 .0 0 .9 1 .9 N itr a te : B ro n x

C le v e la n d SLC B a ltim o r e In d y F re s n o 0 .4 3 .5 S t L o u is 6 .5 T C M (k = 1 .8 ): R ic h m o n d T u ls a C h a r lo tte (k=1.4) B ir m in g h a m 2 .9 Range of TCM based on k= 1.4 to 1.8 8 .1 1 3 .2

C r u s ta l: 0 .0 0 .4 A t la n t a 0 .8 Urban Excess = urban concentration regional urban excess, ug/m3 25 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 20 15 10 5 Gravimetric mass Chemical Components

Bronx/BRIG Baltimore/DOSO Richmond/JARI Charlotte/LIGO Cleveland/MKGO Atlanta/2 Sites Indy/LIVO Birmingham/SIPS St.Louis/3 Sites Tulsa/WIMO SLC/GRBA Missoula/MONT Fresno/PINN 0 excess carbon, % of PM2.5 Carbon is > 50-80% of the Urban Excess %TCM1.4 of total Note: GM excess is higher in part, because of bias in STN mass thru June 2001, and because GM contains water.

With straight inter-network comparison, We see a large OC increment, but. OC urban increment is potentially over stated **(OCM k=1.4) Regional Contribution Urban Increment Bronx/BRIG Baltimore/DOSO Richmond/JARI Charlotte/LIGO Cleveland/MKGO Atlanta/2 Sites Indy/LIVO Birmingham/SIPS St.Louis/3 Sites Tulsa/WIMO SLC/GRBA Missoula/MONT Fresno/PINN 12 10 8

6 4 2 0 Note: Comparisons based on different thermo-optical techniques * k for OCM in rural areas is likely > 1.4, further reducing urban increment as presented With straight inter-network comparison, we dont always see a large urban increment for EC Assuming IMPROVE EC > STN EC, urban increment is potentially understated Regional Contribution EC Urban Increment Bronx/BRIG Baltimore/DOSO Richmond/JARI Charlotte/LIGO Cleveland/MKGO Atlanta/2 Sites Indy/LIVO Birmingham/SIPS St.Louis/3 Sites Tulsa/WIMO SLC/GRBA

Missoula/MONT Fresno/PINN 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 Note: Comparisons based on different thermo-optical techniques By accounting for potential relative bias in reported EC The Urban EC Increment Can Be Bounded Assuming IMPROVE EC > STN EC > IMPROVE EC Regional (lower est) Regional or add'l urban incr. EC Urban Increment Bronx/BRIG Baltimore/DOSO Richmond/JARI Charlotte/LIGO Cleveland/MKGO Atlanta/2 Sites

Indy/LIVO Birmingham/SIPS St.Louis/3 Sites Tulsa/WIMO SLC/GRBA Fresno/PINN Missoula/MONT Upper Estimate EC urban excess 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 Note: Comparisons based on different thermo-optical techniques Other Issues Data reporting conventions STP vs LTP Using high elevation site data to represent regional concentrations Inter-annual variability

PM2.5 mass vs. Component species Estimated Annual Urban Excess for Baltimore, MD Elevation adjustment is a small technical correction to the Urban Excess calculation Concentration, ug/m3 9 Urban excess after elevation adjustment 6 3 0 Sulfate Ammonium Nitrate TCM1.8 Crustal Concentration at 1158m (Dolly Sods) is 12% lower than a sea level estimate Focus on Dolly Sods, WV Average Sulfate March 01 Feb 02 Sulfate 0 - 0.4 0.4 - 0.5 0.5 - 0.7

0.7 - 1 1 - 1.3 1.3 - 1.7 1.7 - 2.2 2.2 - 3.5 3.5 - 4.2 4.2 - 5 Elevation adjustment increases average DOSO sulfate to 4.8 ug/m3 Local Condition Concentrations < High Elevation STP Concentrations http://capita.wustl.edu/CAPITA/CapitaReports/LocalPM10/LocalP10.HTML#combpandt Local Condition Concentrations > Cold Area STP Concentrations http://capita.wustl.edu/CAPITA/CapitaReports/LocalPM10/LocalP10.HTML#combpandt Local Condition Concentrations vs STP Concentrations http://capita.wustl.edu/CAPITA/CapitaReports/LocalPM10/LocalP10.HTML#combpandt CASTNET sites Large Inter-annual Variability in NO3, 2000-02 Trend sites Northern MidWest/NE Southeast Q1 Average Nitrate Annual Average Nitrates 2.5 2.0

6 Northern 16 MWest/NE 1.5 1.0 0.5 0.0 15 Southeast 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 1995 Expressed as Ammonium Nitrates, 1.29*NO3 1997 1998 1999 2000 2001

2002 Use of PM2.5 Measurements FRM Mass not = Ambient PM2.5 Ambient PM2.5 = [Sulfates] + [Nitrates] + [Carbon Mass] +[Crustal] +[Other] Approximation used by IMPROVE program: PM2.5 = [SULFATE] + [NITRATES] + [OCM] + [LAC] + [fine soil] NH42SO4 and NH4NO3 estimated from 3*S and NO3 OCM=1.4*OC Fine soil estimated as 2.2[Al]+2.49[Si]+1.63[Ca]+2.42[Fe]+1.94[Ti] FRM mass does not retain all particle nitrates Includes particle bound water and other (e.g. passive PM2.5) OCM probably different than 1.4*OC = [Ammoniated Sulfate Mass] + [Retained Nitrate Mass] + [Retained Carbonaceous Mass] + [Metallic Metal Oxides] + [Other Components] Summary Many issues associated with Air Quality Measurement Uncertainties are relatively small for Model Evaluation Purposes

Recently Viewed Presentations

  • EXECUTIVE EDUCATION: Coaching for Perforance

    EXECUTIVE EDUCATION: Coaching for Perforance

    WHY WEST POINT - US MILITARY ACADEMY IS . THE BEST CEO SCHOOL IN AMERICA ? Because there are more fortune 500 CEO's that are West Point grads than any other school....Their Followership, Leadership, Ethics and Military Values are their...
  • Model-Driven Specification of Interoperable Service-Oriented ...

    Model-Driven Specification of Interoperable Service-Oriented ...

    C4ISR Architecture Framework v1.0 C4ISR Architecture Framework v2.0 DoDAF v1.0 MODAF v1.0 1996 1997 2003 2005 DoDAF v1.5 2007 MODAF v1.1 2007 NAF v1.0 2005 Scope of UPDM 1.0 Approved Sept 2008 MODAF Meta-Model (M3) expressed using UML Notation MODAF...
  • ILLNESS - WordPress.com

    ILLNESS - WordPress.com

    Illness behaviour-people who are ill generally adopt illness behaviour . These behaviour affect how people monitor their bodies, define and interpret their symptoms, take remedial actions and use the health care systems ( mechanic, 1982)-if people perceive themselves to be...
  • Kingdoms - St. Johns County School District

    Kingdoms - St. Johns County School District

    Which domain includes organisms that have a nucleus, a cell membrane, and mitochondria? A. Archaea B. Eukarya C. Prokarya D. Animalia Which group of organisms includes all of the following: species that are capable of photosynthesis, species that are consumer,...
  • Defect-free Ultra-Rapid Polishing/Thinning of Diamond Crystal Radiator Targets

    Defect-free Ultra-Rapid Polishing/Thinning of Diamond Crystal Radiator Targets

    Primary R&D challenge in Phase 2. Understand and overcome the thin diamond warping problem. SBIR / STTR Exchange Meeting, Gaithersburg, August 6-7, 20147
  • The Role of Earned Value in Fixed Price Projects

    The Role of Earned Value in Fixed Price Projects

    Compute the CV, CPI, SV, and SPI at every level you can . Determine where you have problems (e.g., CPI or SPI < 1.0) and determine the TCPI and/or TSPI required to recover. Do not assume that you and your...
  • Kidney Paired Donation: Update and Challenges

    Kidney Paired Donation: Update and Challenges

    Payer Recommendations …the designation of a national organization to administer and provide oversight to KPD would best meet the needs of expanding access to KT in a fair and equitable manner.
  • Building a SEM Organization: The Internal Consultant Approach

    Building a SEM Organization: The Internal Consultant Approach

    Building a SEM Organization as a Data Driven Consultant Jay W. Goff Vice Provost and Dean for Enrollment Management Larry Gragg, PhD History Department Chair and Professor Robert Montgomery, PhD Psychology Professor & former Department Chair www.enrollment.mst.edu Enrollment Planners Conference,...