Data AnalysisMonitoring Session OAQPS Updates

1
Data Analysis/Monitoring SessionOAQPS Updates

• Neil Frank
• RPO National Workgroup Meeting
• Dallas TX
• December 3-4, 2002

2
Topics

• Reconciliation of IMPROVE/STN data and 6 sites
  Study (5 minutes)
• Urban Excess and Elevation Adjustments (20
  minutes)
• Status of Draft Guidance Documents (5 minutes)
• Access to Canadian Data (5 minutes)
• Discussion (10 minutes)

3
IMPROVE/STN Intercomparisons
4
The 6 Site Intercomparison

• Collocated sampling at 6 sites
• 12 months sampling completed Oct 20 02
• To continue most sites through July 03
• One additional site operated by NYS
• Data will be compared for all common analytes
• 1st year data analyses ready in March 2003

5
Pinnacles St Park, NY RP Sampler
6
New York PM-2.5 Species Data (April Nov 01 STN
vs. IMPROVE (provided by Dirk Felton, NYSDEC)
IMPROVE 0.89STN - 0.00 R2 0.88
7
What else will help us understand IMPROVE vs. STN?

• Inter-lab comparison of the same quartz filters
• Carbon measurements associated with the different
  DRI and STN analytical procedures
• Detailed analysis of data and the thermograms
• With help from EPA/ORD, will look for
• Seasonal or site differences
• Potential effect of predominant emission sources
  ?
• Study of shipping procedures during spring/summer
  03
• Study of blank filter corrections for STN
• Current focus on carbon

8
Urban Excess and Elevation Adjustments
9
Ambient PM2.5 Composition in Rural Areas SO4 and
associated NH4 are prevalent in Rural Eastern
PM2.5
IMPROVE network, 1999
From Latest Findings on National Air Quality,
2001 Status and Trends
10
Ambient PM2.5 Composition in Urban Areas More
Carbon and Nitrates in Urban Areas vs. Rural Areas
EPA STN network, 2001
From Latest Findings on National Air Quality,
2001 Status and Trends
11
What Data are Selected for Comparison of Urban
STN and Rural IMPROVE Sites?

• Time Period Mar 2001-Feb 2002
• Air Quality 12-month average concentrations
• Measurements to account for PM2.5 mass
• Sulfates, Ammonium, Nitrates
• Organic Carbon (OC), Elemental Carbon(EC)
• Certain Elements Al, Si Ca, Fe, Ti
• To estimate crustal component
• Required data completeness
• For each calendar quarter
• 50 of observations (i.e. gt 15)

12
Data Handling Protocols How did we treat
Ammonium (NH4) ?

• Although STN measures NH4 at all sites, NH4 for
  STN and IMPROVE are estimated
• from sulfate (SO4-) nitrate (NO3-)
  concentrations, assuming
• Sulfates are ammonium sulfate, (NH4)2SO4
• Nitrates are ammonium nitrate, NH4NO3
• For consistency with IMPROVE

13
Data Handling Protocols How did we treat
carbon?

• OC and EC data are not separately presented
• IMPROVE and STN use different thermo-optical
  techniques to measure carbon
• Total Carbonaceous Mass is estimated as
• TCM k OC EC,
• Range of TCM is considered, with k 1.4 and 1.8
• OC is blank corrected
• for STN using network-wide estimates
• IMPROVE data is already corrected

14
Preliminary OC Blank Correction for STN

• Derived from network average quartz filter field
  blanks
• Appropriate to adjust annual averages
• Varies by 24-hr sampler volume
• MetOne (SASS) 9.6 m3
• Anderson (RASS) 10.4 m3
• RP 14.4 m3
• URG (MASS) 24 m3
• IMPROVE 32.8 m3

Avg OC blank, ug/m3
15
Evaluation of STN OC blank correction for
12-month averages using measured OC vs PM2.5 mass
OC concentration, ug/m3
PM2.5 Mass Concentration, ug/m3
16
Data Handling Protocols How did we estimate
crustal component?

• Crustal (fine soil)
• 2.2Al2.49Si1.63Ca2.42Fe1.94Ti
• For consistency with IMPROVE

17
What sites are considered?

• Selected Urban Sites
• From EPAs PM2.5 Speciation Trends Network (STN)
• 13 of 35 sites with complete data
• Selected Rural Sites
• From IMPROVE Network (including protocol sites)
• 98 sites to describe spatial patterns
• 16 sites paired with the urban sites for
  urban-rural comparisons

18
EPA Speciation Trends Network 35 complete sites
initially selected for analysis
35 STN sites
19
IMPROVE Network, 2002
98 Sites with complete data, Mar 01-Feb 02
20
13 Selected Urban Sites are Paired with Rural
Sites for Urban PM2.5 Excess Calculations
Missoula
Cleveland
Bronx
SLC
Indy
Fresno
S.L.
Baltimore
Tulsa
Richmond
Charlotte
Atlanta
Birmingham
13 urban STN sites
16 rural IMPROVE sites
21
Are the Rural IMPROVE Sites Indicative of
Regional Backgroundfor Comparison to Urban
Sites?Lets look at Spatial Patterns for SO4,
NO3 and Carbon
22
Rural Sulfates March 01 Feb 02
23
Rural Nitrates March 01 Feb 02
24
Rural Total Carbonaceous Mass March 01 Feb 02
TCM1.8OCEC
25
Now, lets estimate urban excess

• First in terms of gravimetric mass
• Then, specific species

26
Urban PM2.5 is Higher than Nearby Rural
Concentrations
Top Urban Bottom Rural
12-month average PM2.5 mass with speciation
samplers
27
Estimated Annual Urban Excess for Baltimore, MD
Dolly Sods, WV Rural IMPROVE site (background)
Baltimore MD STN urban site
Urban Excess
Urban Rural concentrations
Rural Concentrations Superimposed on Urban
Rural Concentrations Adjusted for
Elevation differential
Top bars are urban concentrations Bottom bars are
nearby rural concentration
28
Estimated Annual Urban Excess for Baltimore, MD
Elevation adjustment is a small technical
correction to the Urban Excess calculation
Concentration, ug/m3
Urban excess after elevation adjustment
Concentration at 1158m (Dolly Sods) is 12 lower
than sea level
29
Focus on Dolly Sods, WV Average Sulfate March 01
Feb 02
Elevation adjustment increases average DOSO
sulfate to 4.8 ug/m3
30
Examples of Other Urban-Rural Pairings
31
Ambient Urban Excess Concentrations for 13
example areas
Range of TCM based on k 1.4 to 1.8
Urban Excess
urban
concentration regional background
32
PM2.5 Urban Excess is Mostly Carbon
Concentration, ug/m3
TCM (k1.8)
TCM (k1.4)
Crustal
12-month average PM2.5 mass and components with
speciation samplers
33
Carbon Excess Percentage is even higher when FRM
mass is used
w. STN mass
w. FRM mass
34
Status of Draft RH Guidance Documents
35
Regional Haze Guidance

• Two guidance documents prepared to support States
  in preparing SIPs
• Draft Guidance for Tracking Progress Under the
  Regional Haze Rule
• Draft Guidance for Estimating Natural Conditions
  Under the Regional Haze Rule

36
Chronology of the Guidance Review

• Distributed for public review in October, 2001
• Notice of availability in Federal Register in
  December 2001 to spur additional comments
• Distributed for technical peer review in April,
  2002
• Responses in preparation by EPA/State/RPO/NPS
  working group (contract related delays)
• Revised Guidance expected January 2003

37
Technical Peer Review Committee

• Bruce Hill
• Clean Air Task Force
• Rich Poirot
• Vermont Department of Environmental Conservation
• Ivar Tombach
• Environmental Consultant
• James Watson
• Desert Research Institute
• Warren White
• Washington University

38
Response to Commentssome key points

• Technical peer review and public comment
  addressing scientific issues
• Substantial rewriting of several sections and new
  technical appendices for clarification (not new
  science)
• Refinements to substitution procedures

-- Revised monthly avg f(RH)

• -- Sensitivity analyses with new f(RH)
• Many editorial changes and clarifications

39
Access to Canadian Data
40
North American Particulate Matter Database

• North American Atmospheric Chemistry (NAAtChem)
  Database is under development
• Collaboration of National Atmospheric Chemistry
  (NAtChem) Database and Analysis Facility
  (operated by Environment Canada) and Clean Air
  Markets Division (EPA)
• with full web-based data access
• likely to be completed in about 6 to 12 months.
• In the meantime, data available by request
• Contacts
• Bob Vet (EC)
• Gary Lear (CAMD/EPA)

41
What data is in NATChem?

• NAtChem archives 3 types of data from regional
  scale networks in the USA and Canada
• Precipitation Chemistry, Particulate Matter and
  Toxics data
• The NAtChem Particulate Matter Database consists
  of
• a number of US and Canadian data sets

42
(No Transcript)
43
(No Transcript)
44
How network differences are treated?

• The differences between the networks/data are
  documented.
• NatChem has converted all data to 0 deg C, 760 mm
  Hg except those data at ambient conditions.
• Methods to further harmonize STP/LTP units
  underway

45
(No Transcript)