draft, 2 june 20051 natural haze levels sensitivity assessment 1. project overview ivar tombach...
TRANSCRIPT
Draft, 2 June 2005 1IvarTombach, Ph.D.EnvironmentalConsulting
NATURAL HAZE LEVELS SENSITIVITY ASSESSMENT
1. Project Overview
Ivar TombachRegional Haze Data Analysis Workshop
8 June 2005
Draft, 2 June 2005 2IvarTombach, Ph.D.EnvironmentalConsulting
Project Sponsorship
• WRAP is lead, acting on behalf of the 5 RPOs
• Contract is with Western Governors’ Assoc.
• Marc Pitchford is WRAP technical lead for project
Draft, 2 June 2005 3IvarTombach, Ph.D.EnvironmentalConsulting
Purpose of Study
• States have the option of developing refined estimates of natural conditions that incorporate advances in scientific knowledge and consider local conditions
• The study assists in this process by evaluating the sensitivity of natural haze estimates at each Class I area to potential refinements
• Study results will help States evaluate where refinements could materially affect estimates of natural conditions and reasonable progress glide paths
Draft, 2 June 2005 4IvarTombach, Ph.D.EnvironmentalConsulting
Topics of Today’s Presentations
• 1. Overview of study (this presentation)• 2. Discussion of currently-prescribed
approach to estimating average and 20% best/worst natural conditions
• 3. Descriptive analysis of default natural concentrations and their relationships to current concentrations
Draft, 2 June 2005 5IvarTombach, Ph.D.EnvironmentalConsulting
Today’s Topics (cont’d)
• 4. Discussion of effects of uncertainties and procedural refinements on visibility improvement glide paths
• 5. Discussions of initial results on refinements to natural conditions estimates
Caveat -- These are initial results of a work in progress and could change
Draft, 2 June 2005 6IvarTombach, Ph.D.EnvironmentalConsulting
Three Types of Potential Refinements to Natural Conditions Estimates
• 1. Refinements to the IMPROVE extinction formula– May consider refinements to formula after this
workshop– Up to now this study has only considered effects of
including sea salt in the extinction formula and of adjusting the OC to OCM mass multiplier
Draft, 2 June 2005 7IvarTombach, Ph.D.EnvironmentalConsulting
Potential Refinements (2)
• 2. Refinements to EPA’s (Trijonis’) default estimates of annual average natural background concentrations– Consider new information on background
concentrations since Trijonis’ work– Consider geographic variability throughout US– Consider seasonal or episodic variability
Draft, 2 June 2005 8IvarTombach, Ph.D.EnvironmentalConsulting
Potential Refinements (3)
• 3. Refinements to EPA’s (Ames’ & Malm’s) approach for estimating haze indices for the 20% clearest/haziest natural haze days – Improved estimates of mean natural values– Improved estimates of standard deviations under
natural conditions– Improved choices of percentile values to
represent averages of worst and best 20%
Draft, 2 June 2005 9IvarTombach, Ph.D.EnvironmentalConsulting
Natural Background vs. Controllable Background -- How Does That Fit In?
S O UR C E O F BAC K G R O U ND AER O S OL
S O UR C E
L O C AT IO N
NA T URA L AN T HR O P OG EN I C
U . S .
( L O CA L)L NP L AP
N o n - U . S .
( TRA N SP O R T ED )T NP T AP
K e y: LN P = L oca l N at u ral Pol l utio n , e t c .
N atu r al B ac k gro u nd und er Re gional Ha z e R ul e ( L N P + T N P )
Po l lutio n t h at i s co n t r ollabl e by U. S. ( L A P )
Po l lutio n t h at i s not co n t r ollabl e by U. S . ( L N P + T N P + T A P )
Draft, 2 June 2005 10IvarTombach, Ph.D.EnvironmentalConsulting
Potential Refinements to Default Natural Concentrations
• Add sea salt and its hygroscopic growth to natural and current conditions (Fine and possibly coarse)
• Consider organics from vegetation (seasonal) and from the ocean
• Consider seasonal and episodic impacts from intercontinental transport of desert dust
• Consider episodic OC and EC impacts of natural fires• Consider episodic impacts from volcanic emissions
Draft, 2 June 2005 11IvarTombach, Ph.D.EnvironmentalConsulting
Issues with Refinements to Concentrations
• Robustness of scientific information for setting a refined concentration value at a specific location?
• To what degree do default concentration values already reflect refinements?
Draft, 2 June 2005 12IvarTombach, Ph.D.EnvironmentalConsulting
But, Note!
• For the objectives of this study– Exact numerical values are not needed to
illustrate whether refinements could potentially result in material changes to natural conditions or glide paths nor for qualitatively evaluating the importance of refining one component relative to another
– States/RPOs can determine more exact values if they actually decide to undertake refinements
Draft, 2 June 2005 13IvarTombach, Ph.D.EnvironmentalConsulting
Ground Rules for Study
• Studied all 156 mandatory Class I Areas, except for Bering Sea Wilderness
• Used 5 yrs of IMPROVE data (1999-2003) or as much of that period as was available to represent each Class I area. (No complete years of data for Breton Wilderness during that period.)
Draft, 2 June 2005 14IvarTombach, Ph.D.EnvironmentalConsulting
Ground Rules (cont’d)
• Followed EPA protocols for data handling and calculations (using data from IMPROVE and VIEWS web sites)– Substitution for missing data– Data completeness– Calculated annual and multi-year haze averages
by averaging deciviews (not bext)
Draft, 2 June 2005 15IvarTombach, Ph.D.EnvironmentalConsulting
Summary of Study Process
• Describe current conditions at each Class I area– Determine slopes of default glide paths for 20%
haziest days– Evaluate sensitivity of results to changing 90th
%-ile to 92nd %-ile and to choice of OC-OMC multiplier
Draft, 2 June 2005 16IvarTombach, Ph.D.EnvironmentalConsulting
Process (2)
• Estimate uncertainties in default estimates of natural extinction at each Class I area– Propagate Trijonis’ uncertainties– Look at geographic distribution of uncertainties.
• Varies with East vs. West and local f(RH) distribution over year
Draft, 2 June 2005 17IvarTombach, Ph.D.EnvironmentalConsulting
Process (3)
• Review scientific basis for refining default concentration values– Propose alternative values (or range of values),
if appropriate, and estimate uncertainties– Re-evaluate Trijonis’ error limits, by subregion
if possible
Draft, 2 June 2005 18IvarTombach, Ph.D.EnvironmentalConsulting
Process (4)
• Determine which PM components are most important contributors to glide path slopes at every Class I area– Include sea salt– Look for geographic and seasonal patterns
Draft, 2 June 2005 19IvarTombach, Ph.D.EnvironmentalConsulting
Process (5)
• Consider potential annual average and seasonal/episodic concentration refinements and evaluate their impacts on the glide path slopes at every Class I area– Relative impacts -- Helps States/RPOs decide
where to concentrate their efforts at developing refinements
– Absolute impacts -- Helps for optimizing regional control strategy over several Class I areas
Draft, 2 June 2005 20IvarTombach, Ph.D.EnvironmentalConsulting
Process (6)
• Synthesize results– Assess relative importances of various potential
refinements at each Class I area– Group Class I areas with similar responses to
refinements, to provide subregional assessments
Draft, 2 June 2005 21IvarTombach, Ph.D.EnvironmentalConsulting
Process (7)
• A study deliverable -- Methodology for assessing relative importance of natural conditions refinements at any Class I area
Draft, 2 June 2005 22IvarTombach, Ph.D.EnvironmentalConsulting
Schedule
• Draft report in mid-July
• Final report by September 1