Options for Estimating Natural Background Visibility in the VISTAS Region

1
Options for Estimating Natural Background
Visibility in the VISTAS Region

• Ivar Tombach
• with benefit of material prepared by Jim Boylan
  and Daniel Jacob (Harvard University)
• Presentation to VISTAS Workgroups
• 15 January 2004

2
The Need

• VISTAS needs to define natural background
  visibility in 2064 and required rate of progress
  by 2018
• due Summer 2004
• Definitions should be specific for each VISTAS
  Class I area

3
Presentation Objective

• Identify issues related to estimating natural
  background visibility
• Provide insights into current scientific
  knowledge
• Discuss options and next steps
• (Update of 9/24/03 presentation by Jim Boylan)

4
The Concept
5
VISTAS Questions

• Should we use the EPA default values to determine
  natural extinction at Class I Areas in the VISTAS
  states?
• How can the EPA default values be refined and
  adjusted to better reflect local conditions?
• How do we deal with the impact of
  intercontinental transport of anthropogenic
  emissions on natural background conditions?
• What are the policy implications of adjusting the
  EPAs recommended approach?

6
Natural vs. Uncontrollable Pollution
7
EPAs Default Natural Conditions for the East
1Trijonis, et al, 1990. National Acid
Precipitation Assessment Program, State of
Science Report 24.
8
EPAs Approach for Determining Natural Extinction
on 20 Haziest Days at a Class I Area

• Start with default values
• Increase sulfate and nitrate extinction for
  particle growth due to humidity, using
  climatological-average f(RH) for site
• Add 10 Mm-1 for Rayleigh (clear air) scattering
• Use Ames Malm statistical procedure to
  determine 90th percentile value

Its all in EPAs guidance document.
9
Key Formulas

• Light Extinction (Mm-1)
• bext 3f(RH)SO4 3f(RH)NO3
  4ORG 10EC 1Soils
  0.6PMC bRay
• bRay 10 Mm-1
• Haze Index (dv)
• HI 10 ln (bext/bRay)

10
Example Great Smoky Mountains (Annual Average
Default Natural Background)
11
Major Issues with Default Approach

• Default Concentrations
• Same concentrations assumed at all Class I areas
  in the East
• Same concentrations assumed to occur every month
  of the year
• Fine sea salt and associated water not included
• Calculation of 20 Haziest Days
• Same frequency distribution assumed for every
  Class I area in the East

In other words, the approach assumes one size
fits all.
12
More Major Issues

• In Calculation of 20 Haziest Days
• 90th -ile assumed to represent haziest 20
• Organics not rolled back when estimating standard
  deviation

13
Possible Refinements to Default Approach for
VISTAS

• Fix Ames and Malm statistical procedure to
  reflect 20 haziest days -- adds 0.42 dv
• Lowenthal and Kumar, 2003
• Change carbon mass multiplier from 1.4 (urban) to
  2.1 (rural) to better represent natural
  conditions
• Turpin and Lim, 2001
• OC is the biggest contributor to default natural
  extinction, so this substantially affects the
  assumed natural background

14
Possible Refinements (contd)

• Consider oceanic aerosol impacts (sea salt and
  organics) near coast
• Average fine sea salt concentration estimates
  range from 0.3 to 1.3 µg/m3 at Cape Romain and
  Florida IMPROVE sites
• Salt is hygroscopic, so extinction impact is
  greater than for same amount of (non-hygroscopic)
  soil, although size distributions are similar
• Oceanic organic particle concentration 0.3 µg/m3
• Review default soil concentrations
• Current concentrations approximate default in
  northern portion of VISTAS region Default may
  be too large there

15
Possible Refinements (contd)

• Consider biogenic organic carbon (and, perhaps,
  sulfur) from forests
• Fine OC 9 µg/m3 during growing season in
  tropics
• Emission flux in some SE forests in summer could
  be comparable to that in tropics, although annual
  average emissions flux is less
• Consider carbon (EC and OC) impacts from natural
  fires
• Global modeling suggests mean naturally-emitted
  EC and OC in the East approximate the default
  concentrations

16
Possible Refinements (contd)

• Consider episodic impact of intercontinental dust
  transport
• Episodic African dust impacts are frequently
  greater than 3 µg/m3 in June-August in southern
  part of VISTAS region
• Episodic Asian dust impacts are greater than 1
  µg/m3 from spring through fall in northern part
  of VISTAS region
• These episodic concentrations exceed the average
  soil default value of 0.5 µg/m3
• Consider impact of intercontinental sulfate and
  nitrate transport (mostly anthropogenic)
• Global modeling suggests mean transported sulfate
  and nitrate concentrations are 2-3 times the
  default concentrations

17
The 64,000 Question

• How much of each of these possible refinements is
  already reflected in the default values?
• Clearly --
• Default organics multiplier and Ames Malm
  statistical factor should be changed
• Some VISTAS locations clearly differ
  substantially from the default averages for
  entire East, at least during part of the year
• Coastal salt and organics
• African dust in summer
• Forest organics during growing season
• Episodic carbon from natural fires

18
Potential Impacts of Some Refinements on Dry
Extinction in VISTAS Region
Blue International Anthropogenic Transport
Adjustment Red Natural Background
Refinement This value represents a
hypothetical worst case scenario including
episodic effects (e.g., Everglades in the
summertime)
19
Potentially Biggest Effects

• All year
• Ames Malm statistics
• OC multiplier factor
• Sea salt in coastal areas
• Oceanic carbon in coastal areas
• Part time
• Forest organics during growing season
• African dust in summer in the south, especially
  during episodes
• Asian dust during episodes in the north, spring
  to fall
• OC and EC from wildfires, episodic

20
Implications for VISTAS (Example Great Smoky
Mountains)

• (1) Default situation
• Current extinction over 20 haziest days 203
  Mm-1 (30.1 dv)
• Default 2064 natural conditions goal 31.4 Mm-1
  (11.44 dv)
• Needed improvement over 60 years 18.7 dv, or
  3.11 dv per decade
• Requires reduction of non-Rayleigh extinction by
  28 per decade

21
GRSM Implications (contd)

• (2) Example with refined natural conditions
• Increase natural extinction on 20 haziest days
  in 2064 by 4 Mm-1 (especially, change organics
  multiplier and add some forest organics),
• This decreases the required slope to 2.9 dv per
  decade, corresponding to a reduction in
  non-Rayleigh extinction by 26 per decade

22
GRSM Implications (contd)

• (3) Most hazy days are in summer at GRSM
• Add 7 Mm-1 for African dust and peak impacts of
  vegetation emissions
• This further decreases the required slope to 2.5
  dv per decade, corresponding to a reduction in
  non-Rayleigh extinction by 24 per decade

23
GRSM Implications (contd)

• If only sulfates are reduced, then neededSO2
  reductions for the three scenarios are
• (1) 54 per decade
• (2) 38 per decade
• (3) 30 per decade

24
Implications (contd) (Another Example Cape
Romain)

• (1) Default situation
• Current extinction over 20 haziest days 142
  Mm-1 (26.5 dv)
• Default 2064 natural conditions goal 31.1 Mm-1
  (11.36 dv)
• Needed improvement over 60 years 15.2 dv, or
  2.53 dv per decade
• Requires reduction of non-Rayleigh extinction by
  24 per decade

25
ROMA Implications (contd)

• (2) Example with refined natural conditions
• Increase natural extinction on 20 haziest days
  in 2064 by 7 Mm-1 (same as GRSM plus sea salt and
  oceanic organics)
• This decreases the required slope to a bit more
  than 2.1 dv per decade, corresponding to a
  reduction in non-Rayleigh extinction by 21 per
  decade

26
ROMA Implications (contd)

• (3) If most hazy days are in summer at ROMA
• Add 3 Mm-1 for African dust (no further
  vegetation adjustment here)
• This further decreases the required slope to a
  bit leas than 2.1 dv per decade, corresponding to
  a reduction in non-Rayleigh extinction by 20 per
  decade

27
ROMA Implications (contd)

• If only sulfates are reduced, then needed SO2
  reductions for the three scenarios are
• (1) 32 per decade
• (2) 28 per decade
• (3) 27 per decade

28
Intercontinental Transport Issues

• What do you do with the anthropogenic pollutants
  transported to the US from Mexico, Canada, and
  Asia?
• Add their contributions to the natural background
  then determine the reasonable progress slope?
• Dont add them to the natural background, but
  account for them when you can not control anymore
  U.S. sources?
• Can result in changing the slope of reasonable
  progress line
• Similar implication as adding previously
  discussed refinements to EPAs default extinction
  values

29
Intercontinental Transport (contd)
20 Haziest Days
29.9
dV
Natural Background (with Intercontinental
Transport)
Natural Background
2000 YEAR 2064
30
Carbonaceous Aerosol from GEOS-CHEM Modeling
This includes anthropogenic vegetation burning,
however.

• Model indicates that EPA default natural
  concentrations are too low by factors of 2-3
  except for OC and EC in Eastern U.S.
  quantifying fire influences is critical
• Transboundary pollution influences are
  relatively small compared to EPA default natural
  concentrations, except for EC from Canada/Mexico

31
Sulfate-Nitrate-Ammonium Aerosol from GEOS-CHEM
Modeling

• Achievability of EPA default estimates is
  compromised by transboundary pollution influences
• Transboundary sulfate influence from Asia is
  comparable in magnitude to that from Canada
  Mexico

32
Another Option.

• Couple the GEOS-CHEM global circulation model
  with the CMAQ regional air quality model
• Annual simulation for 2002
• Time varying boundary conditions from GEOS-CHEM
• Zero out U.S. anthropogenic emissions in CMAQ
• Highly dependent on accurate inventory for
  biogenic and other natural emission sources
• Provides site-specific natural background values

33
Caveat and Outlook

• These results and conclusions are rough and
  tentative
• They are based on limited review of literature
• Our ability to provide site-specific, time
  varying estimates of natural levels should
  improve with further study of the literature
• More relevant research results are being
  published regularly

34
Recommendations

• Understand technical issues and policy
  implications (e.g., this conference call)
• Interpret further data analyses by Air Resource
  Specialists (especially new coastal IMPROVE
  sites)
• Decide on which default values to refine and
  approach for doing so
• Cooperate with other RPOs on new natural
  background project (longer term)