Air Quality Impacts from Prescribed Burning Sangil Lee1, Karsten Baumann2, Mei Zheng2, Fu Wang2 1Sch

1
Air Quality Impacts from Prescribed
BurningSangil Lee1, Karsten Baumann2, Mei
Zheng2, Fu Wang21School of Civil and
Environmental Engineering, 2School of Earth and
Atmospheric SciencesGeorgia Institute of
Technology, Atlanta, Georgia

• ? Objective
• Guided by the Endangered Species Act (ESA), the
  DOI through the Fish and
• Wildlife Service mandates that most army and air
  force bases in the South-
• Eastern US use prescribed burning to maintain the
  health of its native long
• leaf pine forest and thus protecting the habitat
  of the endangered red-
• cockaded woodpecker. In recognition of the
  conflicting requirements between
• the ESA and the Clean Air Act (CAA) statutes,
  the Study of Air Quality
• Impacts Resulting from Prescribed Burning on
  Military Facilities was initiated
• and sponsored by the DOA/CERL in support of the
  DOD Pollution prevention
• Partnership.
• ? Prescribed Burning
• ? Develops, maintains, and enhances wildlife
  habitat.
• ? Protects endangered plants and animals.
• ? Preserves and protects cultural resources and
  wilderness.
• ? Minimizes potentials of catastrophic wildfires
  that could result from heavy
• accumulations of vegetative fuels.
• ? Air Quality Issues associated with Prescribed
  Burning
• Emissions from prescribed burning are important
  primary sources of gases
• and particulate matter (PM) to the atmosphere.

Measurement Site (Oxbow Learning Center) and
Prescribed Burning Site (Fort Benning) near
Columbus, GA
Time Series Plot of Measured Gases and
Particulate Matter
January
February
March
April
May
4067 acres
504 acres
251 acres
3770 acres
937 acres
1256 acres
Background measurement without prescribed
burning.
Low pressure system with showers and T-storm from
3rd to 4th. Cold air moved into the region at
early morning of the burn day.
Highest PM2.5 event associated with relatively
calm and clear conditions on 25th.
Decrease in temperature and PAR plus increase in
wind speed in the wake of a front lead to
decreasing O3 maxima and PM2.5 levels.
Build-up in PM2.5 and O3 maxima after record
rainfalls earlier in May strong daytime winds.
Relatively stagnant conditions all week.
Increasing fractions of Unidentified Mass (grey)
with time progressing into warmer season,
indicates likely larger OOE (turquoise)
fraction, reflecting an organic mass to organic
carbon ratio (OM/OC) increasingly larger than
1.4 possibly due to higher oxygenated PM.
Quantification of water-soluble OC fraction
pending!
Distinct increase of PM concentration at Columbus
is associated with air masses coming from
southeast during Nov-Apr, which is the
prescribed burning period.
Weekly Average Diurnal Cycles of Meteorological
Parameters and Air Pollutant Concentrations in
March
Monthly Diurnal Patterns and CO/NOy Relationships
Wind Rose Plots of Major pollutants
Impact on Regional CO Background
Nighttime (1800 - 1100) Daytime (1100 - 1800)
Dec02
Measurement site is characterized by weak
easterly component flow at night, and stronger
westerly component flow at daytime. Weak SE
flow carries pollutant emissions from
prescribed burning at military installation
during active burn period (Dec-Apr), however, for
May June (little burn activity) no distinct
increase of pollutants in air masses coming from
SE.
Jan03
Feb03
Non-burn week
Non- burn week
Burn week
Burn week
Mar03
Background CO concentration increases with
increased prescribed burning activities. Peak
values stay about three months and decrease a
month after burns have largely
ceased. Background CO levels are determined from
linear CO vs NOy regressions. Standard errors
vary between 1.5 and 2.8 ppbv. Slopes indicate
mixed influence from mobile, small industrial
and PB sources.
Apr03
Larger differences between the daily minimum and
maximum air T and RH were observed on two burn
weeks in March, indicating overall less cloud
coverage, more stably stratified nocturnal BL and
generally drier conditions. An overall
increasing warming trend leads to more intense
atmospheric photochemical activity and higher
daytime O3 maxima. The two burn weeks were also
characterized by elevated nighttime PM2.5, CO and
NOy concentrations associated with calmer
conditions under weak easterly component winds.
Continued emissions from smoldering fuel of the
prescribed burnings likely account for this
nighttime increase, since the relatively cold
plumes (compared to daytime flaming condition)
are being mixed into the shallow nocturnal
boundary layer.
May03
Jun03
WD Frequency ()
WS (m/s)
PM (mg/m3)
CO (ppbv)
NOy (ppbv)
Organic Compounds of PM2.5 and Source
Apportionment from Chemical Mass Balance Approach
VOC Emissions at Prescribed Burning Site

• Measurements of particle-phase organic
  compounds (POC) have been made by a High Volume
  Sampler with pre-baked quartz filters. Sampled
  quartz filters were extracted by organic solvent
  and then analyzed by Gas Chromatography/Mass
  Spectrometry (GC/MS).
• Total 105 POCs are identified n-alkanes (20),
  hopanes (10), steranes (4), polycyclic aromatic
  hydrocarbons (19), resin acids (9), aromatic
  carboxylic acids (3), branched alkanes (3),
  n-alkanoic acids (17), alkenoic acids (3),
  alkanedioic acids (19).
• Cellulose, which provides structural strength
  to plants, constitutes 40-50 dry weight of
  wood. Thermal decomposition of cellulose
  produces mainly levoglucosan, a good marker for
  biomass burning. Burning conifers containing
  resin produce resin acids.
• Samples for the February event have been
  analyzed so far. Levoglucosan is dominant among
  organic compounds identified by GC/MS. Two
  biomass tracers have very similar trend of
  concentrations. Their concentrations 5 hours
  after burning is a factor of 2 to 5 higher than
  those of background and right after burning,
  respectively. This increase is associated with
  relatively calm conditions at nighttime.
• The contribution of wood smoke to total organic
  carbon increases from 11 - 2 before to 53 - 5
  after the flaming phase of the prescribed burn,
  conducted ca. 28 km to the east of the sampling
  site. A wind shift from strong westerly to
  weaker easterly component causes to impact the
  samples collected after 1700.

VOC sampling was conducted at different stages of
burning (i.e., pre-ignition, flaming,
smoldering) and locations (i.e., upwind,
downwind, burn unit) at prescribed burning sites
for each burning event. The collected samples
were analyzed at the University of California,
Irvine, for total 47 gas species including CO,
CO2, and CH4. Mixing ratio enhancements were
calculated by subtracting mixing ratio of upwind
(background) from that of flaming or smoldering
stage. As seen by higher CO2/CO ratio, flaming
is more efficient combustion than smoldering,
leading to larger mixing ratio enhancement.
However, less efficient combustion (smoldering
), which has longer duration (8-12 h) than
flaming (2-3 h), makes higher emission factors.
The prolonged emissions at higher relative rates
during smoldering potentially causes the
observed increase of pollutant concentrations at
night under calm conditions in a stable
nocturnal BL.
Acknowledgement This work was sponsored by the
Department of the Army/ U.S. Army Construction
Engineering Research Laboratories (CERL) via
subcontract with the University of South-Carolina
(USC), Grant No. DACA42-02-2-0052, in support of
the DOD Pollution Prevention Partnership. The
authors gratefully acknowledge the collaboration
and field support received by Jill Whiting, Jim
Trostle, and Becky Champion (CSU-OLC), Jack
Greenlee, Hugh Westbury, Polly Gustafson, and
John Brent (Ft. Benning), Frank Burch and Steven
Davis (Columbus Water Works), Allen Braswell and
Stephen Willard (Ft. Gordon), Venus Dookwah, Wes
Younger and Michael Chang (GIT-EAS).
PB start at noon
Wind shift