Date post: | 17-Dec-2015 |
Category: |
Documents |
Upload: | alaina-lee |
View: | 218 times |
Download: | 2 times |
Interagency Monitoring of PROtected Visual Environmentsa State Perspective of the IMPROVE Steering Committee
1977 CAA Established National Visibility Goals• Applied to 156 Class I Federal Areas• Required EPA (to Require the States) to:
• Prevent Future & Remedy Existing Impairment &• Conduct Visibility Monitoring in Class I Areas
• AND Also Charged the Federal Land Managers with an “Affirmative Responsibility to Protect the Air Quality Related Values (Including Visibility)” in the Class I Areas They Manage
• BUT There Were No Standard Visibility Monitoring Methods in 1977, Nor Was There Much Available $
IMPROVE Steering Committee
• Established in 1985
• Composed of EPA + 4 FLMs (NPS, USFS, USFWS, BLM)
• Developed Methods & Pooled Resources
• IMPROVE Network (20 Sites, only 3 East) Started in 1988
• EPA $ from §105 Pot; STAPPA is Not Pleased
• 1990 WESTAR Proposal Led to Addition of: WESTAR, NESCAUM & STAPPA to Steering Committee
• 1991 Network Expansion added several New Eastern Sites (including Lye Brook & Brigantine) with FLMs subsequently adding Moosehorn (94) and Gt.Gulf (95)
• MARAMA added to Steering Committee in 1998
• Arizona Added as “Associate Member” in 1999
IMPROVE’s Been a Good Neighbor
• Allowed NESCAUM States to partially “Opt In” with 1988-93…5 NEPART NETWORK
• 1993-94 NEPART Enhancement in Champlain Basin
• IMPROVE Samplers in Casco Bay Toxics Study
• IMPROVE Samplers Loaned to CT for 1999-00 SURFUP
• Opt-In Option for State PM-2.5 Speciation Sampling
IMPROVE in Transition
• Huge ‘00 Network Expansion, New Samplers & Schedule
• Data will be Suddenly more “Relevant” with Regs
• FLMs Recent Addition of Several Top Data Analysts, will allow Improved Technical Assistance to States...
IMPROVE Meeting Agenda Inn at Estes Park - Columbine Room
8:30am September 6th – 12:00pm September 7th 2000
Wednesday, September 6th 8:30am Welcome and introductions Marc Pitchford8:45am Agenda review and revisions Marc Pitchford9:00am Network status (data recovery, technical issues, etc)
Optical John MolenarParticle Lowell Ashbaugh
10:00am Brea10:15am Network expansion - progress, problems Lowell Ashbaugh
& plans (sites yet to be installed, sampler & Bob Eldredperformance, flow & temperature data,
1day in 3, site documentation, etc)11:15am IMPROVE Protocol site status Lowell Ashbaugh
(9 FLM, 16 state/tribal, 8 CASTNET)11:30am Status of Quality Assurance Program Plan Bob Eldred12:00pm Lunch1:30pm Update on EPA’s guidance for tracking Marc Pitchford
regional haze trends2:00pm Update on EPA’s & WRAP’s approaches for Marc Pitchford
determining natural background3:00pm Visit Rocky Mountain IMPROVE monitoring site Carpool to site,
5:00pm Carpool to dinner for those interested.
Thursday, September 7th
8:00am Reconstructed extinction for use in Mark Scruggs the Regional Haze Rule
8:30am Review of data recovery goals Marc Pitchford8:50am Approaches to cope with missing data Marc Pitchford9:20am Break
Reports by subcommittees9:45am Data availability Scott Archer, Bill Malm, & Rich Poirot10:00am IMPROVE web site Mark Scruggs, Scott Archer & Bill Malm10:15am View monitoring Mark Scruggs, Sandra Silva & Bill Malm10:30am Next generation view monitoring John Molenar11:00am Image archival & distribution Bob Bachman11:30am Review action items & next meeting Marc Pitchford12:00pm Adjourn
Starting Ht. 1000 mb (Surface Flows)
Starting Ht. 925 mb (Synoptic Flows
Canadian Hemispheric Trajectory Model (HTM) 72-hour Back Trajectories at 0Z on 7/17/99 for
Montreal, PQ, Lye Brook, VT & Martha’s Vineyard, MA
From Mario Benjamin, Environnement Canada - Region du Quebec
Receptor Model Assessment Tools for Regional Haze
• Class I Areas in Few, Fixed “Receptor” Locations
• Haze Regulations Are Based on Aerosol Data
• and Apply to Long Averaging Times
Statistical Models Trajectory Methods
Chemical Mass Balance Cluster Analysis
Residence-Time Positive Matrix Factorization Analysis
Potential Source UNMIX Contribution Function
Receptor Model Limitations
• None are ”Stand Alone” Techniques
• None have “Standard EPA Guidelines”
• All Require User Skills & (Subjective) Judgement
• Trajectory methods are not very “Transferable”, & Limited by Quality, Detail, & Availability of Met Data
• Statistical Models Don’t Handle Secondarys Well, & are Limited by “Resolving Power” of Input Data
However, EPA is currently putting significant effort into Development, Evaluation & User-Interface of CMB8, UNMIX & PMF
• Latest, User-friendly Windows Interface; Will be Free!
• Not quite Bug Free/ Available Yet
• Need to Develop Regional Source Profiles
• Contact: Tom Coulter, EPA ([email protected])
• Available Free (from Ron Henry) but Requires MATLAB($)
• May be Re-compiled in C (Won’t require MATLAB)
• Currently being extensively evaluated by EPA
• Contact: Ron Henry ([email protected])
• Available 30 day Free, $400 License from Pentti Paatero
• Good User’s Manual, but Steep Learning Curve
• Currently Evaluated at EPA, May be replaced by MLE
• Contacts: P. Paatero ([email protected]), Phil Hopke ([email protected])
CMB8
UNMIX
PMF
Trajectory Cluster Analysis Groups a Large Number of Trajectories into a Small Number of Clusters, which Represent Predominant Meteorological Flow Patterns to the Receptor.
Then Various Summary Statistics Can Be Calculated for the Pollutant Concentrations, Pollutant ratios, deposition, etc. for Days in Each cluster.
Example from Barbara Stunder on NOAA READY Site Tour
Example Application of Cluster Analysis to 108 1993 Trajectories to State College AIRMoN Wet Deposition Site. From Barbara Stunder, NOAA-ARL.
Highest Sulfate Dep. Associated with Flows from NW-Short, SW-Short and SW-Long.
Lowest SO4 with NW-Long and South.
4-X Comparison of Receptor Methods
Applied to IMPROVE-Type Speciated PM-2.5 Data and CAPITA Monte Carlo Trajectories from Underhill, VT
Raw Data Included Fine Mass and 27 Trace Elements for 854 Sample Days with Many Variables Occasionally or Frequently Below MDL or Missing
Example UNMIX Results from 834 IMPROVE-Type Speciated PM-2.5 Samples from Underhill, VT Values represent percentages of each Input species (left column) accounted for by each identified source (top row)
UNMIX (top) & PMF (bottom) Seasonal
Source Contributions to 1989-95 PM-2.5
Mass at Underhill, VT
UNMIX: 7 Sources; PMF: 11 Sources,
7 of which were “Similar” to UNMIX
Notes: Both Results Projected to Residence-Time Grid Domain and Employ PSCF Inverse Distance Weighting - to minimize influence of Sparsely Covered Squares
Notes: Incremental probabilities are multiplied by 103
Green (Negative) Probabilities show where Source is Least Likely
• Midwestern Sulfur Sources are Predominant on Worst PM Days, Especially During Summer
• “Local” Eastern Sources, Especially Woodsmoke, Become Relatively More Important on Clean Days
Conclusions & Recommendations
• Many “Useful” Receptor Tools are “Available”
• None are Completely “Ready for Prime Time”
• Much Development/Evaluation is Underway
• Start Exploratory NE Receptor Model Workgroup(s)
1. Trajectory Modeling: Could Begin by selecting Model(s), Met Data, Defaults, etc. and Begin to Develop long-term Trajectory database for Each of Our (few) Class I Areas.
2. Statistical Models: Could Begin Exploratory Analysis with UNMIX, standard Factor Analysis, & possibly PMF on Historical IMPROVE data sets
Web References
7/16-17/99 Northeast Fine Particle and Ozone "Haze" Episodehttp://capita.wustl.edu/neardat/Activities/july99/july99.htm
Use of Airmass History Models & Techniques for Source Attribution http://capita.wustl.edu/capita/capitareports/airmasshist/EPASrcAtt_jul17/index.htm
Trajectory Cluster Analysis Primer on NOAA READY Site Tourhttp://www.arl.noaa.gov/slides/ready/climate/clim2.html
Harvard Forest Regional-scale Airmass Composition by PATHhttp://www.people.virginia.edu/~jlm8h/nigec.d/nigec.html
Air Mass History Pollution Climatology for Northeastern Forests and Parkshttp://capita.wustl.edu/NEARDAT/Reports/TechnicalReports/ForestSer_TrajProp/fstrjsum.htm
Estimation of U.S. Influence on Mean PM10, PM2.5 and Particle Sulphate Concentrations in Eastern Canadahttp://capita.wustl.edu/NEARDAT/Reports/ReportFragments/jbrook/uspminfl.html
Air Trajectory Pollution Climatology for the Lake Champlain Basinhttp://capita.wustl.edu/neardat/Reports/TechnicalReports/lakchamp/lchmpair.htm
Source Attribution Section in EPA Fine Particle Data Analysis Workbookhttp://capita.wustl.edu/PMFine/Workbook/PMTopics_PPT/WB_SourceAttr/Version3/index.htm
Receptor Modeling Section in EPA PAMS Data Analysis Workbookhttp://www.epa.gov/oar/oaqps/pams/analysis/receptor/factblrectxtsac.html#cmb
Fine Particulate Matter PM-2.5 Source Profiles Project Planhttp://capita.wustl.edu/PMFine/Workgroup/SourceAttribution/Reports/Proposed/ord_plan.pdf
Tracers of Opportunity: Potassiumhttp://capita.wustl.edu/PMFine/Workgroup/SourceAttribution/Reports/In-progress/potass/Kcover.htm
EPA Source Apportionment Study Literature Reviewhttp://capita.wustl.edu/PMFine/Workgroup/SourceAttribution/Reports/Completed/literatu.pdf
Local and Regional Contributions of PM2.5 to Urban Areas in the Mid-Atlantic and Southwestern UShttp://capita.wustl.edu/CAPITA/CapitaReports/PMFineAn/PM_vs_Tran/PMVsWnd_Pres/
August 1995 Forest Fire Impacts in New England and Atlantic Canadahttp://capita.wustl.edu/NEARDAT/Reports/TechnicalReports/smoke895/895smoke.htm
Recent PM Receptor Modeling Workshops & Conferences
Workshop on UNMIX and PMF As Applied to PM2.5, U.S. EPA, RTP, NC February 14-16, 2000
Workshop Report and Appendices will be available next week from EPA, Contact: Chuck Lewis ([email protected]) or Shelly Eberly ([email protected])
International Symposium on Measurement of Toxic & Related Air Pollutants, RTP, NC, Sept. 12-14, 2000
SESSION 5: PM RECEPTOR MODELING (Wed. Sept. 13)(Chair: Dr. Charles Lewis, U.S. EPA, NERL, Research Triangle Park, NC8:00 A.M. EPA Workshop on Source Apportionment Tools UNMIX and PMF as Applied to PM2.5 Synthetic Data Set Generation Shelly I. Eberly and Basil W. Coutant, U.S. EPA, Office of Air Quality Planning and Standards, Research Triangle Park, NC; Charles W. Lewis, U.S. EPA, NERL, Research Triangle Park, NC
8:25 A.M. Comparison of the Analyses of a Synthetic Data Set by Positive Matrix Factorization and UNMIX Xin-Hua Song, Ziad Ramadan, and Philip K. Hopke, Dept. of Chemistry, Clarkson University, Potsdam, NY; Ronald Henry, Dept. of Civil and Environmental Engineering, University of Southern California, Los Angeles, CA
8:50 A.M. Multilinear Receptor Analysis Utilizing Wind Direction and Wind Speed as Independent Variables Pentti Paatero, Dept. of Physics, University of Helsinki, Helsinki, Finland; Philip K. Hopke, Dept. of Chemistry, Clarkson University, Potsdam, NY
9:15 A.M. Comparison of Source Apportionment of Phoenix Fine Particle Mass by PMF and UNMIX Philip K. Hopke, Ziad Ramadan, and Xin-Hua Song, Dept. of Chemistry, Clarkson University, Potsdam, NY; Ronald Henry, Dept. of Civil and Environmental Engineering, University of Southern California, Los Angeles, CA
9:40 A.M. Break and Exhibition Viewing
10:10 A.M. Sources of Fine Particle Concentration and Composition in Northern Vermont Rich Poirot, Paul Wishinski, VT Air Pollution Control Division, Waterbury, VT; and Alexandr V. Polissar and Philip K. Hopke, Clarkson University, Potsdam, NY
10:35 A.M. The Potential Influence of Data Artifacts on Receptor Model Results Rich Poirot, VT Dept. of Environmental Conservation, Waterbury, VT; Philip K. Hopke, Dept. of Chemistry, Clarkson University, Potsdam, NY
11:00 A.M. Chemical Mass Balance Software: EPA-CMB 8.2 Thomas Coulter, U.S. EPA, NERL, Research Triangle Park, NC; Robert A. Wagoner, Pacific Environmental Services, Inc., Research Triangle Park, NC; Charles W. Lewis, U.S. EPA, NERL, Research Triangle Park, NC
11:25 A.M. Recent Improvements in SPECIATE: Source Profiles and Data Management William Hodan, Mark Saeger, Robert Wagoner, and Vicky Kriegsman, Pacific Environmental Services, Inc., Research Triangle Park, NC