Quantifying the Impacts of Wildfires and Dust Events along the Wasatch
Front: Description of Dissertation Research
Derek V. Mallia
PhD Candidate, Department of Atmospheric Sciences, University of Utah
XX August 2015
1. Introduction
Wildfires and dust storm events are responsible for emitting large quantities of criteria
pollutants as defined by the U.S. Environmental Protection Agency (EPA) [EPA, 2011;
Steenburgh et al., 2012] while wildfires also emit large quantities of greenhouse gases [Galanter
et al., 2000; IPCC, 2013]. For wildfire, the species emitted include CO2, CO, water vapor,
particulate matter, nitric oxides (NOx), CH4, volatile organic compounds (VOCs), organic
carbon, hydrocarbons, and elemental carbon [Yokelson et al., 2009]. Dust storms primarily emit
dust particles than can range from less than < 2.5μm to > than 10 μm in diameter [Fu et al.,
2014]. In addition, suspended dust particles can act as transporters and reaction sites for harmful
species, such as NOx, lead, fungal spores, and microorganisms [Fu et al., 2014; Huang et al.,
2010; Lee et al., 2009]. Emissions from both wildfires and dust events can be transported
thousands of kilometers from the emission point, which can have significant impacts on regional
air quality, visibility [Debell et al., 2004; Spracklen et al., 2009; Wotowa and Trainer, 2000; Val
Martin et al., 2010; Fu et al., 2014; Liu et al., 2006; Carlson 1979, Steenburgh et al., 2012].
The western U.S. is the primary source of wildfire emissions in the U.S., due to arid
conditions, the abundance of needleleaf forests, and a dry season [Westerling et al., 2006,
Wiedinmyer and Neff, 2007]. The greatest wildfire emissions occur between the months of June
and October, with maximum emissions occurring during the month of August. Dennison et al.,
[2014] noted that there was a general increase in wildfire activity (>405 ha) across the western
U.S. over the past 3 decades for most regions (Fig. 1). Littell et al., [2009] also observed an
increase in wildfire activity starting as early as 1970 according to observations and a
reconstructed database that spanned from 1916 through 2004.
Increases in wildfire activity across the western U.S. has attributed to higher annual mean
temperatures and droughts that results in earlier snowmelt, which prolongs the wildfire season
[Westerling et al., 2006; Dennison et al., 2014; Riley et al., 2013; Bowman et al., 2009]. Landuse changes have also considered being a culprit to an increase in wildfire activity [Westerling et
al., 2006]. Under the IPCC’s moderate emissions scenario, climate models have indicated that
climate change will occur over the next century as a result of increased concentrations of
greenhouse gases such as CO2 and CH4 [IPCC, 2013]. General circulation model simulations in
Spracklen et al., [2009] showed that wildfire burned areas across the western U.S. would double
by 2050 resulting in a doubling of carbonaceous aerosols as a result of climate change. Due to
the potential consequences of wildfires on air quality, visibility and climate, its prudent to
1
understand the magnitude and transport of wildfire smoke emissions when trying to identify the
regional and global impacts of wildfire emissions.
Dust that is generated from wind erosion is the largest single contributor to particulate
matter found in the atmosphere [Forster et al., 2007; Rind et al., 2009]. Most dust erosion occurs
over arid and semiarid environments where dust can be easily transported into the atmosphere
Liu et al., 2006]. Particulate matter from dust erosion can be transported over large distances
from its original source, with human health and ecological impacts at downwind sites
[Steenburgh et al., 2012; Goudie and Middleton, 2001; Prospero and Lamb, 2003]. Severe dust
storm events can also have a significant impact in visibility resulting in a disruption in
transportation and daily human activities [Liu et al., 2006]. Previous work has shown that the
number of people with lung respiratory disease and stroke increases dramatically during dust
storm events [Ichinose et al., 2008; Kang et al., 2013]. In addition, transported dust can also
impact the surface radiation budget by absorption and scattering [Sokolik et al., 2001;
Steenburgh et al., 2012]. Another harmful side effect of dust particles can occur during dry and
wet deposition of particulate matter that can result in acidification [Doney et al., 2007; Shi et al.,
2012].
2. Background
a) Wildfire emissions
Biomass burning has been suggested to account for as much as 15-30% of global CO
emissions and 20% of anthropogenic CO2 emissions [van der Werf et al., 2010]. During active
western U.S. wildfire seasons, CO emissions from fires can account for up to 20% of the total
annual emissions in the United States while CO2 wildfire emissions comprises between 4-6% of
anthropogenic emissions [Urbanski et al., 2011; Wiedinmyer and Neff, 2007]. Understanding
CO2 emissions from wildfires is of significant interest due to the fact that CO2 is considered the
major greenhouse gas responsible for climate change [IPCC, 2013]. On the other hand, CO is a
criteria pollutant that is regulated by the EPA [EPA, 2011]. In addition to gaseous species such as
CO2 and CO, fires also release large quantities of aerosols/particulate matter [Davies and Unam,
1994; Sapkota et al., 2002; Park et al., 2007].
Recent work has been carried out to develop emission inventories for wildfires across
North America [Wiedinmyer and Neff, 2007; Urbanski et al., 2011; van Der Werf et al., 2010].
The Global Fire Emissions Database (GFED) [van Der Werf et al., 2010] is a daily biomass
burning emissions product gridded spatially at 0.5°. However, wildfire activity is highly variable
in both time and space, potentially rendering a 0.5° grid insufficient for resolving individual
wildfires [Wiedinmyer and Neff, 2007; Urbanski et al., 2011]. Recent work by Mu et al., [2010]
has developed global diurnal scale factors for wildfires which accounts for the diurnal cycle seen
in wildfires. Research has pointed to significant uncertainty in estimated fire activity in regards
to the burn area, fire intensity and combustion completeness [Urbanski et al., 2011; Bowman et
al., 2009].
2
Urbanski et al. [2011] recently developed the Wildland Fire Emissions Inventory (WFEI)
for 2003-2008, which is a high-resolution fire emissions inventory for the western United States.
This inventory incorporates many different data sets such as MODIS, NCEP FNL, fuel loading
maps, and emission factors in order to reduce the uncertainty associated with CO and CO2
emissions for non-agricultural biomass burning [Urbanski et al., 2011]. The WFEI has been
recently updated to include PM2.5 along with the fire emissions for the 2012, which was a large
wildfire season for the western U.S [Dennison et al., 2014]. The updated WFEI includes new CO
and PM2.5 emission factors for forest fires [Urbanski, 2013]. For 2012, the MODIS-DB based
burned area used in WFEI was unavailable and alternate sources of burned area maps were
employed. Daily burned area was based on a combination of fire perimeter polygons collected
by the USGS GeoMAC (http://wildfire.usgs.gov/geomac/index.shtml) and a daily MODIS burn
scar product produced by the US Forest Service Remote Sensing Applications Center
(http://activefiremaps.fs.fed.us/burnscar.php) using the algorithm of Giglio et al. [2009]. Fire
perimeter area not mapped by the daily MODIS burn scar product was assigned an estimated
burn date using active fire detections from the MODIS MXD14 product [Giglio et al., 2003] and
NOAA’s HMS (http://www.ssd.noaa.gov/PS/FIRE/hms.html). The 2012 emission product also
integrated significant updates for vegetation maps and fuel loading. Forest vegetation type and
fuel loading were assigned based on a Forest Type Group map [Ruefenacht et al., 2008] and the
forest surface fuel classification of Keane et al. [2013]. The surface fuel loading was augmented
with fuel loading estimates of understory fuels [Wilson et al., 2013] and canopy fuels, the latter
of which was estimated from canopy spatial data layers from the LANDFIRE project
[LANDFIRE, 2014]. Area mapped as non-forest in the Forest Type Group map was assigned fuel
loading from a MODIS NDVI based rangeland biomass product [Reeves et al., In prep.]. Forest
canopy fuel consumption was taken as 50% while consumption of other fuel components was
estimated using the First Order Fire Effects Model assuming “dry” conditions (see Urbanski et
al., 2011 for details). As in the 2007 emission dataset, CO and PM2.5 emission factors for forest
fires were taken from Urbanski [2013]. For both 2007 and 2012 the heat flux was estimated
using a heat of combustion of 18.6 MJ kg-1 biomass [Susott et al., 1975; Klass, 1998]. The
WFEI is available daily at 500-m grid spacing for years between 2003 and 2008, as well as 2012.
Annual, domain-wide uncertainties within the WFEI range from 28-51% for CO emissions and
40-65% for PM2.5 emissions [Urbanski et al., 2011].
b) Impacts of wildfire emissions on air quality
Particulate matter with an aerodynamic dynamic diameter < 2.5μm (PM2.5) is a criteria
pollutant that is regulated by the EPA [EPA, 2011]. EPA has established National Ambient Air
Quality Standards (NAAQS) for both short term and long term exposure to PM2.5. Compliance
with the short term standard of 35 μg/m3 is evaluated as the 3-year average of the 98th percentile
of the daily maximum 24-hour average concentration, and compliance with the long term
standard of 12 μg/m3 is evaluated as the 3 year average of the annual mean PM2.5 concentration
[EPA, 2011]. High concentrations of PM2.5 can have adverse effects on human health, as these
3
particulates can be easily inhaled enabling them to penetrate deep into the lungs, which can cause
respiratory problems [EPA, 2011]. Young children, the elderly, along with people with long and
heart disease are the most susceptible to elevated concentrations of PM2.5 [EPA, 2011; Beard et
al., 2012]. Increased concentrations of aerosols and particulate matter from wildfires can have
impacts on regional visibility, which can pose a hazard to transportation in the affected areas
[Val Martin et al., 2013]
Previous studies have investigated the impacts of wildfire emissions to enhance the
atmospheric concentrations of pollutants regulated by the EPA NAAQS for downwind regions
[Clinton et al., 2006; Bravo et al., 2002; Davies and Unam, 1994; Debell et al., 2004; Sapkota et
al., 2002; Dempsey, 2013]. During the summer of 2002, wildfires in Quebec injected large
concentrations of CO and PM2.5 into the free troposphere, which were later transported by midlevel winds to the northeastern U.S [Debell et al., 2004; Sapkota et al., 2002]. Using
observations coupled with trajectory analyses, satellite images and LIDAR, Sapkota et al.,
[2002] concluded that smoke from the Quebec fires were the source of the enhancements seen in
PM2.5 and PM10 observations in Baltimore on July 7th 2002. Debell et al., [2004] noted similar
observations across much of the Northeast for O3, CO, K+, and PM2.5 and also concluded that
these enhancements came from wildfires based on satellite images and 850-hPa streamline
analyses.
A similar situation was observed in 2010 when smoke from wildfires in Saskatchewan
was transported over Toronto, resulting in elevated levels of PM2.5 and O3 [Dempsey, 2013].
Dempsey, [2013] made this conclusion based on NOAA’s Hazardous Mapping System (HMS)
smoke analysis product and measured total column CO concentrations from Atmospheric
Infrared Sounder (AIRS) instrument on the NASA Aqua satellite. While these studies
determined the influences of wildfire emissions on downwind locations using qualitative
methodologies, these studies were unable to quantify the direct contributions of atmospheric
concentrations by these fires.
c) Impact of dust storm events on air quality
Previous research has shown that dust storm events can have significant impacts on
human health and visibility [Pope et al., 1995; Grebhart et al., 2001]. The African Sahara desert
is one of the world’s largest dust emission sources and is responsible for providing dust to
Caribbean Sea causing significant impacts on human health and ecology [Carlson 1979; Goudie
and Middleton, 2001]. Asian dust storms over the Taklimakan and Gobi Deserts can occasionally
transport fine dust across the Pacific which has been responsible for enhancing aerosol
concentrations that exceed the NAAQS [Jaffe et al., 1999; Husar et al., 2001; VanCuren and
Cahill, 2002; Fairlie et al., 2007].
Satellite observations and modeling studies have suggested the western U.S. is a
significant contributor to particulate matter in the atmosphere [Ginoux et al., 2001; Woodward,
2001]. For western U.S., the Great Basin, Colorado Plateau, and the Mojave Desert are the major
sources of dust emissions [Reynolds et al., 2001]. Previous work has indicated that the Great
4
Basin is the primary source region of dust for the Wasatch Front as it has an abundance of
erodible dust from paleolakes from the last glacial maximum along with flat lowlands, burn
scars, ephemeral streams, silt pans and playas [Gill, 1996; Washington et al., 2006; Jewell and
Nicoll, 2011; Steenburgh et al., 2012; Hahnenberger et al., 2012]. Other favorable conditions
that makes the Great Basin an idea source is the lack of surface crusting of sediments, minimal
vegetation cover, and large surface areas with silt-sized particles. As a result, these factors
reduce the threshold friction velocity, which makes it easier for dust to be transported from the
ground into the atmosphere [Park et al., 2009].
On average, areas across the Wasatch Front experience 4-5 dust storms events annually,
which can have regional impacts across much of northern Utah [Hahnenberger et al., 2012].
PM2.5 concentrations generally average around 12 μg/m3 for the duration of the dust storm event
with strong southerly winds out ahead of a strong cold front. As a result, PM2.5 concentrations
generally stay below the NAAQS 24-hour standard unless the dust event is exceptionally severe
[Hahnenberger et al., 2012; Steenburgh et al., 2012]. On average, most dust storm events across
the Wasatch Front occur during the spring and fall when strong cold fronts are most prevalent
[Hahnenberger et al., 2012; Steenburgh et al., 2012]. These events are often the most intense
during the late afternoon when winds are the strongest as a result of a growing PBL height and
turbulent mixing [Hahnenberger et al., 2012]. Steenburgh et al., [2012] also noted that the dust
storm events are generally the strong during the frontal passage. This is likely the result of strong
northwesterly winds behind the front and convergence along the frontal boundary, which will act
to concentrate particulate matter.
Using HYSPLIT backward trajectories, MODIS and GOES satellite images,
Hahnenberger et al., [2012] and [Steenburgh et al., 2012] were able to determine that the eastern
Great Basin does not uniformly emit dust and that specific hotspots were responsible for
contributing the majority of the dust. The hotspots that were identified across the eastern Great
Basin were the Sevier Dry Lake, Tule Dry Lake, the Great Salt Lake Desert, and the Milford Flat
burn scar from a large wildfire back in 2007 [Hahnenberger et al., 2012; Steenburgh et al.,
2012]. It should be noted that these findings were based on subjective analyses [Steenburgh et
al., 2012], thus calls for the need to quantitatively determine the contributions from these source
regions.
d) Stochastic Time-Inverted Lagrangian Transport (STILT) model
In order to determine the impacts of wildfire emissions on CO2, CO, CH4, and PM2.5
concentrations across North America, I will be utilizing the STILT model [Lin et al., 2003].
STILT is a Lagrangian Particle Dispersion Model, which has the ability to simulate atmospheric
motions with an ensemble of particles that represent parcels of air. Turbulent motions within the
STILT model are calculate by a Markov chain process, which has been tested to adhere to the 2nd
Law of Thermodynamics and time-reversibility [Lin et al., 2003]. Time-reversibility aspect of
the STILT model ensures that a single backward-time simulation reveals the same information as
forward-time simulation, but at a much cheaper computational cost [Lin et al., 2003]. A
5
Lagrangian framework like STILT also offers several benefits over Eulerian tracer models in the
Lagrangian formulation’s physical realism, numerical stability, lack of numerical diffusion,
adherence to mass conservation, and computational efficiency [Lin et al., 2012; Wholtmann and
Rex, 2009; Shin and Reich, 2009; Smolarkiewicz and Pudykiewicz, 1992; McKenna et al., 2002].
Using information from the STILT backward trajectories, the STILT model has the
ability to determine the upwind source regions that affect atmospheric concentrations for the area
of interest. This makes STILT a valuable tool for interpreting atmospheric concentrations at
observation sites. In order to determine the upstream source regions, the STILT calculates the
surface flux footprints f(xr, tr|xi, yj, tm) for a receptor at location xr and time tr to an upwind
source at (xi, yj) and prior time tm can be estimated from the WRF-STILT backward trajectories
[Lin et al., 2003; Nehrkorn et al., 2010; Skamarock et al., 2008]. The footprint is simply the
measure of the upwind surface influences for a receptor as determined by the STILT backward
trajectories. The footprint is a function of the number of Lagrangian particles within the
planetary boundary layer (PBL) for some upwind location and has units of mixing ratio per unit
surface flux as seen in the equation below:
𝑁𝑡𝑜𝑡
𝑚𝑎𝑖𝑟
1
𝑓(𝒙𝒓 , 𝒕𝒓 |𝑥𝑖 , 𝑦𝑗 , 𝑡𝑚 ) =
∑ ∆𝑡𝑝,𝑖,𝑗,𝑘
ℎ𝜌̅ (𝑥𝑖 , 𝑦𝑗 , 𝑡𝑚 ) 𝑁𝑡𝑜𝑡
(Eq. 1)
𝑝=1
where mair is the molecular weight of air, h is the height of the volume in which the surface
fluxes are diluted over (surface influence volume), ρ is the average density for all particles, Ntot is
the total number of particles, and Δtp,i,j,k is the amount of time a particle p spends within the
surface influence volume at location (xi, yj) and time tm [Lin et al., 2003; Wen et al., 2012; Kim et
al., 2013; Lin et al., 2013]. Any surface fluxes that occur within the PBL are assumed to be
rapidly mixed within the surface influence volume, which is taken to extend from the surface to a
height of 0.5 zi (one half of the PBL height). Previous studies have indicated that simulated
STILT footprints were insensitive to the exact value of the column height “h” as long as h was
between 10 and 100% of the PBL height [Lin et al., 2003; Gerbig et al., 2003]. By combining
the footprint with a flux source, the STILT can determine the atmospheric concentration
contribution of that source.
The STILT modeling framework offers the ability to quantify the uncertainty impacts of
transport error, which is unique among Lagrangian models [Lin and Gerbig, 2005]. A stochastic
error velocity is added to the parcel motion that is based on error statistics derived from
comparisons of observations versus model wind fields. Errors are then propagated through the
STILT trajectories in a Monte Carlo Fashion. Using uncertainties from atmospheric transport,
simulated concentrations can be matched up with observed values calculated from a formal
Bayesian inversion system in order to optimize wildfire and dust emissions fields. A similar
methodology has been developed by Gerbig et al., [2008] in order to account for uncertainties is
vertical mixing.
6
e) Weather Analysis and Forecasting (WRF) model
Backward trajectory ensembles generated by STILT model can be driven by wind fields
from the Advanced Research version of the WRF model (ARW, version 3.4.1) model
[Skamarock et al., 2008]. WRF is a Eulerian non-hydrostatic atmospheric model equipped with a
large suite of physical parameterizations [Skamarock et al., 2008]. Time-averaged, mass coupled
winds from the WRF model haven been shown to improve mass conservation and the temporal
representation of wind variation [Nehrkorn et al., 2010; Hegarty et al., 2013]. In addition, the
native vertical levels within STILT were selected to closely match the WRF vertical levels to
further improve mass continuity [Nehrkorn et al., 2010].
f) Wildfire plume rise heights
A potential limitation of the Lagrangian particle dispersion modeling framework is the
absence of a plume rise parameterization geared specifically for wildfires [Mallia et al., 2015].
In its current formulation, the STILT model footprint assumes that emissions only occur near the
surface, therefore these emissions are only diluted through the PBL depth. However, studies have
shown that major wildfires can inject emissions into the free troposphere (FT) [Val Martin et al.,
2010; Sofiev et al., 2012] and in some cases, the lower stratosphere [Fromm et al., 2005].
Emitted gases that escape into the FT can have longer lifetimes and can be transported much
further distances (>1000 km) [Debell et al., 2004; Val Martin et al., 2010]. The height in which
the wildfire plume reaches is a complex function of the wildfire radiative power and the local
meteorology [Val Martin et al., 2010].
Labonne et al., [2007] first utilized satellite observations of aerosols from the CloudAerosol Lidar with Orthogonal Polarization in order the estimate the number of wildfire plumes
that lofted emissions in the FT. It was estimated that the vast majority of plume rises stay within
the PBL and that the lofting of emissions into the FT is a rare occurrence. More recently, Khan et
al., [2008] used data from the Multi-angle Imaging SpectroRadiometer (MISR) instrument on
board of the NASA Terra satellite, which indicated that 5-18% of the observed wildfire plumes
across Alaska and the Yukon territories reached the FT. The contradicting results between these
two studies are thought to be the result of spatial sampling and sensitivity differences between
the two instruments [Kahn et al., 2008]. A similar study to Kahn et al., [2008] was performed for
all of North America using digitized MISR data, which showed similar percentage of wildfire
plumes that made it into the FT [Val Martin et al., 2010]. A recent campaign led by Clements
and Lareau [2015] utilized LIDAR measurements in order to determine the plume characteristics
of wildfire smoke plumes in 8 different fires across California for the 2014 wildfire season.
Initial results suggest that the injection of wildfire emissions into the FT occurs much more
frequently than what previous literature has suggested [Clements and Lareau, personal
communication]. It was suggested that this may be the result of satellite observations only having
one scan of fires per day which generally doesn’t occur during the peak of the wildfire diurnal
cycle [Clements and Lareau, personal communication]. Another hypothesis is that the smoke
intrusions into the FT may also be too fine scale in size to be depicted in satellite observations
7
[Clements, personal communication]. Work from previous transport modeling studies that
utilized the chemical transport model (CTM) seem to agree with these finding as 50% of the
wildfire emissions had to be prescribed above the PBL in order to match observations [Hyer et
al., 2007; Generoso et al., 2007].
Plume rise models from buoyant sources were originally formulated in the 1960 and
1970s by Briggs [1969] and Guldberg [1975]:
(Eq. 2)
where Hc is the final rise of the plume, F is the buoyancy flux parameter, g is the acceleration
due to gravity, x* is the downwind distance from the source, U is the mean horizontal wind
speed, and s is the buoyancy parameter. These semiformulas were later refined in the 1980s in
order the better represent the meteorology along with a more detailed description of the source
characteristics [Briggs, 1984; Weil, 1988]:
(Eq. 3)
where N is the Brunt-Vaisala frequency, u* is the friction velocity, w* is the convective velocity
scale, zi is the mixed layer height, ϕ is the Froude number, and vs is the stack exit velocity. A
major drawback to the Briggs plume rise models is that these formulas were originally designed
for smoke stack tops which generally have a well defined emission diameter where emission
temperature and velocity can be assumed to be constant (spatially) [Sofiev et al., 2012].
However, wildfires often have a non-circular shape with strongly varying temperature over time
and space, along with no definitive stack release height [Briggs, 1984; Sofiev et al., 2012]. It
should also be noted that these formulas also assume a vertically homogenous atmosphere. While
this assumption may be valid for plume rises that stay within the PBL, this could be a poor
assumption for plume rises that make it into the FT [Briggs, 1984]. Some of these limitations
may have manifested itself in Raffuse et al., [2012] who found that the Briggs plume rise models
consistently underestimated wildfire plume rise heights when compared to MISR and CALIOP
8
observations. Additionally, the correlation between the modeled and observed plume rise heights
was weak [Raffuse et al., 2012; Val Martin et al., 2012].
Recently, a more advanced approach was taken in order to calculate wildfire plume rises
using 1-D plume rise models. These models include BOUYANT [Martin et al., 1997], BUOFMI [Nikmo et al., 1999], FIREPLUME [Liu et al., 2010], and the Freitas et al., [2007]. These
models explicitly integrate 1-D equations for energy, mass and momentum of the plume rise
centerline. Using this methodology, the varying vertical structure of the atmosphere can also be
accounted for instead of assuming a 2-layer model (PBL and FT) that has prescribed temperature
and wind speeds for each model layer [Frietas et al., 2007; Sofiev at el., 2012]. Some limitations
include that these 1-D models assume horizontal symmetry of the smoke plume, require accurate
knowledge of the emission characteristics, and that these models are computationally expensive
[Sofiev et al., 2012].
A new methodology for estimating wildfire smoke plume heights was presented in
[Sofiev et al., 2012]. Similar to the Briggs models [Briggs, 1969; Briggs, 1984], this
methodology uses a semi-empirical formula. However, the Sofiev et al., [2012] method also
incorporates physical process that control uplift that is specific towards wildfires, which is absent
in the Briggs plume rise models. Following this framework, Sofiev et al., [2012] introduced the
following formula:
𝐹𝑅𝑃
𝛾
𝐻𝑝 = 𝛼𝑧𝑖 + 𝛽 ( 𝑃 ) 𝑒𝑥𝑝 (
𝑓0
2
−𝛿𝑁𝐹𝑇
𝑁02
)
(Eq. 4)
where 𝛼 is part of the PBL passed freely, 𝛽 weights the contribution of fire intensity, 𝛾
determines the power-law depends on the fire radiative power (FRP), 𝛿 weights the dependence
of the stability of the FT on the plume rise height (Hp), 𝑃𝑓0 is the reference fire power (Pfo = 106
2
W), 𝑁02 is the Brunt-Vaisala frequency reference number 𝑁02 = 2.5 x 10-4 s-2, and 𝑁𝐹𝑇
is the
Brunt-Vaisala frequency of the FT. A leaning subset of MISR fire observations were then used to
determine the value of the empirical calibration constants (𝛼, 𝛽, 𝛾, 𝛿) which were estimated as
the following:
𝛼 = 0.15; 𝛽 = 102𝑚; 𝛾 = 0.49; 𝛿 = 0
(Eq. 5)
Results in Sofiev at el., [2012] found that their methodology significantly outperformed both the
Briggs and 1-D BOUYANT plume rise models when applied to 2000 fire plumes from the MISR
database across North America and Siberia. The Sofiev at el., [2012] plume rise model accurately
predicted two thirds of all plume rises within 500m of the observed values. In addition, it was
found that 82% of the plume rises that made it into the FT were accurately predicted within
500m when compared to MISR data (Fig. 2) [Sofiev at el., 2012].
3. Research questions and results
9
a. What are the impacts of wildfires on air quality along the Wasatch Front
My main authored publication focused on quantifying the impacts of upwind wildfires on
air quality in Salt Lake City. From this study we were able to conclude that the STILT model can
(1) reasonably resolve CO and CO2 concentrations within the Salt Lake Valley, (2) that wildfires
can have large, episodic impacts on PM2.5 and CO concentrations, and (3) that WRF-STILT
model can capture the timing and magnitude of wildfire influences when compared to
observations. This work also lays the groundwork for future back-trajectory wildfire studies,
which will need to include the impacts of wildfire plume rises and chemistry for more reactive
chemical species such as ozone. This study also lays a modeling groundwork for state agencies
to prove exceptional events to EPA when National Ambient Air Quality Standards (NAAQS) are
violated.
In addition to Mallia et al., [2015], the impacts of wildfire emissions on observed PM2.5
was extended beyond SLC to include sites all along the Wasatch Front (Fig. 3, 5, 7, 8, and 9).
These sites included Logan, Provo, Brigham City, Ogden, and Tooele. Simulations for PM2.5
included wet and dry deposition, and emissions from wildfires only (i.e no anthropogenic
emissions or background values included). The secondary formation of PM2.5 was also not
included in these simulations. Similar to the results at SLC, wildfires also had significant,
episodic impacts across all the sites for the 2012 western U.S. wildfire season, which matched
enhancement seen in the observed concentrations of PM2.5.
Logan had the largest enhancements, especially during the months of August and
September (Fig. 3). This is likely the result of Logan’s closer proximity to the Idaho fires, which
was the dominants area of fire activity during these months (Fig. 4). It should be noted that the
observed enhancements seen at Logan match well with the modeled wildfire contributed PM2.5
concentrations. While the timing of the observed and modeled enhancements of PM2.5 match
well, there is a slight underestimation of PM2.5 concentrations during the middle of September.
As of right now its unclear whether this discrepancy is the result of transport error, lack of plume
rise formulation, and/or an underestimation in the emission fields.
Provo, which was the southern most site in our study, showed the largest enhancements
of PM2.5 concentrations during the month of June and July from wildfires in central Utah (Fig. 5
and 6). This is likely a result of Provo being closer to the central Utah wildfires, which peaked in
intensity during the months of June and July. Our three remaining sites (Tooele, Brigham City,
Ogden) showed similar enhancements in the observed concentrations of PM2.5 when compared to
the observed values at SLC (Fig. 7, 8, & 9). Enhancements of the PM2.5 concentrations were
maximized during mid August and September, which matched the timing of modeled wildfire
contributed PM2.5 concentrations. The magnitude of the observed PM2.5 enhancements roughly
matched with modeled wildfire contributed PM2.5 concentrations, though there was an
underestimation of 10 μg/m3 of PM2.5 during the middle of September, which was present at all
of the observation sites. As a whole, the extended work presented here was consistent with the
results found in Mallia et al., [2015]. Wildfires in Idaho during the months of August and
September had significant impacts on quality on a regional scale, which may have been
10
responsible for the NAAQS violations across the Wasatch Front.
b. How does wildfire plume rises affect the dispersion of wildfire emissions
In its current formulation, STILT doesn’t have a way to account for plume rises from
wildfires that make it into the FT. In its currently formulation, emissions as assumed to occur at
the surface and are diluted through half of the PBL [Lin et al., 2003; Gerbig et al., 2002].
However, previous literature has suggested that up to 15% of wildfire smoke plumes make it into
the FT [Kahn et al., 2008; Val Martin et al., 2010], while other studies have observed more
frequent occurrences of smoke intrusions into the FT [Clements and Lareau, personal
communication].
In an effort to account for a fire plume height (FPH) that exceeds the PBL height,
wildfire plume rises will be calculated using a semi-empirical model derived by Sofiev et al.,
[2012], which requires inputs for fire radiative power, the PBL height, and the local atmospheric
stability. Both the PBL height and atmospheric stability can be obtained from WRF model
output, while the fire radiative power can be obtained from the WFEI and RxCADRE data sets.
The calculated FPH would then instead be used to calculate the footprint, as follows (Fig. 10):
𝑁𝑡𝑜𝑡
𝑚𝑎𝑖𝑟
1
𝑓(𝒙𝒓 , 𝒕𝒓 |𝑥𝑖 , 𝑦𝑖 , 𝑡𝑚 ) =
∑ ∆𝑡𝜌,𝑖,𝑗,𝑘
𝐹𝑃𝐻𝜌̅ (𝑥𝑖 , 𝑦𝑖 , 𝑡𝑚 ) 𝑁𝑡𝑜𝑡
(Eq. 6)
𝜌=1
The goal will be to apply this formulation to our STILT PM2.5 simulations and determine the
sensitivity of our model results by comparing STILT simulations that use wildfire plume rises
with the simulations that do not. It has been hypothesized in Mallia et al., [2015] that
underestimations of modeled PM2.5 concentrations at SLC during middle of September 2012
could be due to transport errors related to wildfire plume rise heights.
This framework will also be extended into our work involving the Joint Fire Science
Program’s (JFSP) Prescribed Fire Combustion and Atmospheric Dynamics Research Experiment
(RxCADRE) field campaign at Eglin Air Force Base, Florida, which has a full suite of airborne
smoke measurements (downwind pollutant concentrations, plume rise, and emission factors),
fuel consumption, spatially resolve fire intensity, and local meteorology [Urbanksi, 2014]. The
purpose of this work will be to demonstrate that the WRF-STILT modeling framework can
provide a rigorous constraint on biomass fire emissions by comparing modeled concentrations of
CO and CH4 with high precision aircraft measurements of these quantities. Understanding how
fire emissions are distributed throughout the lower atmosphere will be crucial due to the close
proximity of the prescribed burn to the aircraft measurements.
As an additional step, we may attempt to couple STILT with WRF-Sfire [Mandel et al.,
2011], which would allow us to explicitly calculate the prescribed burn plume height during the
RxCADRE field campaign. WRF-Sfire modeling framework allows for the coupling of local
meteorology with fire characteristics which determines the amount of fire spread, along with the
feedback between the fire and atmosphere [Mandel et al., 2014]. WRF-Sfire also has the ability
11
to determine the fuel moisture based on the local meteorology within WRF. In addition, WRFSfire also has the capability to explicitly model fire smoke plumes based on the meteorology
from WRF and the associated heat fluxes from the fire [Mandel et al., 2014]. WRF can then be
coupled with STILT, without the modification to the footprint definition, since the fire plume
will already be explicitly resolved within the WRF meteorological fields. Simulations of CO and
CH4 will then be compared to the previous WRF-STILT runs along with the airborne
measurements.
c. What are the impacts of dust events along the Wasatch Front?
Air quality and its impacts on health is a growing concern across the Wasatch Front as
this region is densely populated with a rapidly growing population [U.S. Census Bureau, 2010].
The Wasatch front is heavily impacted by particulate matter from anthropogenic sources and
from wildfires, which are particularly prevalent during the summer months [Mallia et al., 2015].
Previous work has shown that dust storm events are fairly common across the Wasatch Front
with an annual occurrence of 3-5 times per year [Hahnenberger et al., 2012]. However, only
~5% of these events are considered moderate/severe and has a return frequency around once
every 5 year [Steenburgh et al., 2012]. However, when moderate to severe dust events occur,
visibility can be reduced to >1-km with PM2.5 and PM10 concentrations that exceed the NAAQS
for these species [DAQ 2010a; DAQ 2010b]. However, dust storms are considered exceptional
events as they are not reasonably controllable or preventable, not caused by human activity that
is unlikely to recur and a particular location, and are a natural event. As a result, there is a need
for research that has the ability to quantify the total contribution from dust sources during these
events.
Until recently, it has been difficult to quantify the effects of these events on Utah’s air
quality as it requires the ability to separate their effects from other sources of pollution through
observational studies. Due to recent advances in atmospheric modeling such as WRF-STILT, a
modeling frame exists that can separate the effects of natural dust sources from anthropogenic
sources that are not considered exceptional. A similar methodology that was employed in Mallia
et al., [2015] will be used to separate the anthropogenic contributions towards PM2.5 and PM10
concentrations from natural dust sources during dust storm events. Recent dust storm events
from May 21st 2008, March 30th 2010, April 28th 2010, and April 14th 2015 will be the case
studies used to evaluate the performance of WRF-STILT modeling framework. These events
were considered severe with PM2.5 concentrations exceeding > 200 μg/m3 and visibility that was
reduced to less than ¼ 1-km [DAQ 2010a; DAQ 2010b].
Currently, no dust emission inventories exist for the Great Basin for the case studies
discussed earlier. As a result, an existing dust model will be needed in order to generate dust
emissions for the times of interest. Tong et al., [2011] recently developed a dust emission model
called FENGSHA, which was used to estimate dust emissions across the U.S. and is currently
coupled with the latest version of CMAQ (CMAQ 5.0) (http://www.camq-model.org). The
vertical flux of dust (gm-2 s-1) can be calculate bed the following equation:
12
𝐹 = ∑𝑖,𝑗 𝐾 × 𝐴 ×
𝜌
𝑔
2
𝑆𝑖 × 𝑆𝐸𝑃 × 𝑢∗ × (𝑢∗2 − 𝑢∗𝑡𝑖,𝑗
) for u* > u*t
(Eq. 6)
where i is the landuse type, j is the soil type. K represents the ratio of the vertical flux to
horizontal sediment, which is dependent on the clay content (%) and calculated as the following
[Tong et al., 2010; Fu et al., 2014]:
𝐾 ={
10.136[𝑐𝑙𝑎𝑦 %]−6
. 0002
𝑓𝑜𝑟 𝑐𝑙𝑎𝑦% < 20%
𝑓𝑜𝑟 𝑐𝑙𝑎𝑦% ≥ 20%
(Eq. 7)
A is the particle supply limitation, ρ is the air density, g is the gravitational constant, S is the
area of landuse type i. SEP is the soil erodibility factor, which is defined as the following:
𝑆𝐸𝑃 = 0.08 × 𝑐𝑙𝑎𝑦% + 1.00 𝑥 𝑠𝑖𝑙𝑡% + 0.12 × 𝑠𝑎𝑛𝑑%
(Eq. 8)
u* is the friction velocity, and u*t is the threshold friction velocity which determines the intensity
and the onset of dust emissions. u*t can be defined as the following:
′
𝑢∗𝑡 = 𝑢∗𝑡
× 𝑓𝑑 × 𝑓𝑚
(Eq. 9)
′
where 𝑢∗𝑡
is the threshold friction velocity for loose fine-grained soil with low surface
roughness. This variable is set as a constant of .7, which is based on dust studies in the Mojave
Desert [Gillette et al., 1980]. 𝑓𝑑 is the soil moisture while 𝑓𝑚 is the snow cover.
The initial plan is to drive the dust emission model with WRF output, which contains
information for landuse and soil types, landuse area, air density, friction velocity, soil moisture,
and snow cover. It should be noted that previous work has shown that the WRF default USGS
database is insufficient for providing realistic landuse and soil type information across much of
western Utah [Massey et al., 2014] (Fig. 11). To rectify this issue, WRF was recompiled with the
National Landuse Cover 2006 Database (NLCD 2006), which has a much more realistic
representation of landuse and soil type data across western Utah. In additional, additional soil
and landuse categories were added that includes playa.
Once the dust emission model is ran for the case studies listed above, the WRF-STILT
modeling framework will then be used to determine the direct contributions towards observed
PM2.5 and PM10 concentrations by following the methodology used in Mallia et al., [2015].
Preliminary STILT trajectory analyses for the case studies presented above that suggest that
trajectories original from the dust emission hot spots discussed in Hahnenberger et al., [2012]
and Steenburgh et al., [2012] (Fig. 12 and 13). A gravitational settling scheme by Zender et al.,
[2003] will be adopted in addition to including dry and wet deposition for the WRF-STILT PM2.5
and PM10 simulations. More information on dry and wet depositions schemes used in the WRFSTILT modeling framework can be found in Mallia et al., [2015]. Additional work may also
13
include using a formal Bayesian Inversion system that utilizes transport error in order to
constrain derived emissions from the dust emission model described in Eq. 6.
A project was recently proposed that involves the development of Bear River reservoir,
which is one of several rivers that empty into the Great Salt Lake. This has been proposed in
order to keep up with the growing water demands for a rapidly expanding Wasatch Front. As a
result, the Great Salt Lake could shrink in size resulting in more exposed playa across the eastern
Great Basin. Since playa is considered a major source of dust, this could potentially lead to more
frequent and intense dust storm events across the Wasatch Front, which will only exacerbate the
air quality issue further in the future. Sensitivity analyses will be performed on the case studies,
which will change the landuse and soil type information to reflect a shrinking Great Salt Lake as
a result of a Bear River Dam. Model simulations for PM2.5 and PM10 will then be compared to
the control run to determine whether there are any impacts of a Bear River Dam during intense
dust storm episodes.
4. Additional projects
In addition to my work on Mallia et al., [2015], I was directly involved in 2 publications
during 2014, which focused on a stable isotope analysis of Hurricane Sandy (Good et al., 2014a;
Good et al., 2014b) The first two publications used over 600 hundreds precipitation samples
collected during Hurricane Sandy for an oxygen and hydrogen isotopic analysis. Using
Lagrangian back-trajectories coupled with the stable isotope analyses, we were able to
reconstruct the water cycling within Hurricane Sandy allowing for a better understanding of
storm evolution and dynamics along with the hydroclimatical impacts associated with such
storms.
I have also been working on another side project with our collaborators (Jason Davidson
and Hyoun-Tae Hwang) over at the University of Waterloo working on coupling the
HydoGeoSphere (HGS) land-surface model with the WRF model. Previous work has shown that
the built in WRF land-surface models may be overly simplified which results in errors in
evapotranspiration and soil saturation (literature?). In order to remedy this issue, the HGS and
WRF models will be coupled together. In this modeling framework, WRF will provide the HGS
model with potential evapotranspiration and precipitation. HGS will then compute the integrated
water flow with evapotranspiration and feed WRF information about the actual
evapotranspiration and soil saturation. As an initial step, a case study with the coupling of these
two models will be ran over California for several years. My work for this project will
specifically focus on setting up the WRF configuration for this study, and helping with the
validation of the WRF meteorological fields against observations.
Lastly, I’ve been loosely involved in TRAX project with setting up and running WRF
model simulations during times of data collection (August 11 – September 13th 2014) in an effort
to aid the research for this project.
5. Timeline
14
I have by enrolled as graduate student since August of 2012, and fully admitted to the
PhD program as of July of 2014. As of right now I am proposing a target graduation date by the
end of the spring semester of 2017. Currently, my proposed dissertation will comprise of three
sections, which will focus on the research questions outlined above. The first and second topics
are currently associated with our “Exceptional Events” project that is being funded by the Utah
Division of Air Quality along with the Improving Biomass Emissions within CarbonTracker
project, which is being funded by NOAA. The third topic is also being funded under the
Exceptional Events project. Most of the work in the first topic had already been completed and is
published in Mallia et al., [2015]. Additionally, progress has been made in the two remaining
topics thus a target graduation date by the end of 2017 spring semester should be feasible. It is
expected several more lead-authored publications will come out of the remaining topics along
with at least one additional co-authored publications from my additional projects.
15
Figure 1. Western U.S. trends for the number of large wildfires (>405 ha) for each ecoregion
from 1984 - 2012. The black line for each plot presents the Theil-Sen estimate slope for each
ecoregion. Source: Dennison et al., [2014]
16
Figure 2. Modeled plume rise heights versus observed MISR plume rise heights [Sofiev et al.,
2012].
17
Figure 3. Observed concentrations of PM2.5 (a) vs. modeled wildfire contributed PM2.5
concentrations (b) at Logan.
18
Figure 4. Wildfire modeled of PM2.5 contributions towards concentrations at Logan.
19
Figure 5. Observed concentrations of PM2.5 (a) vs. modeled wildfire contributed PM2.5
concentrations (b) at Provo.
20
Figure 6. Wildfire modeled of PM2.5 contributions towards concentrations at Provo.
21
Figure 7. Observed concentrations of PM2.5 (a) vs. modeled wildfire contributed PM2.5
concentrations (b) at Brigham City.
22
Figure 8. Observed concentrations of PM2.5 (a) vs. modeled wildfire contributed PM2.5
concentrations (b) at Ogden.
23
Figure 9. Observed concentrations of PM2.5 (a) vs. modeled wildfire contributed PM2.5
concentrations (b) at Tooele.
24
Figure 10. Visualization of the modification to the footprint definition.
25
Figure 11. Figure of soil types changes within WRF. The left shows the WRF soil categories
using the default USGS data set while the updated WRF (right) uses the NLCD 2006 database,
which now includes a playa category. Source: Massey et al., [2013]
26
Figure 12. STILT trajectories arriving at SLC during the April 28th 2010 dust storm event. The
blue box represents the Sevier Lake Reservoir.
27
Figure 13. STILT trajectories arriving at SLC during the May 21st 2008 dust storm event. The
black box represents the Utah Salt Flats.
6. References
28
Ackerman, A.S., O. B. Toon, D. E. Stevens, A. J. Heymsfeld, V. Ramanathan, E. J. Welton,
(2000), Reduction of tropical cloudiness by soot, Science, 288, 1042–1047.
Beard, J. D., C. Beck, R. Graham, S. C. Packham, M. Traphagan, R. T. Giles, and J. G. Morgan
(2012), Winter temperature inversions and emergency department visits for asthma in Salt Lake
County, Utah, 2003-2008, Environ. Health Perspect., 120(10), 1385–1390,
doi:10.1289/ehp.1104349.
Bowman, M. J. S, and coauthors (2009), Fire in the earth system, Science, 324, 481–48,
doi:10.1126/science.1163886.
Bravo, A. H., E. R. Sosa, A. P. Sánchez, P. M. Jaimes, and R. M. Saaverde (2002), Impacts of
wildfires on the air quality in Mexico City, 1992-1999, Environmental Pollution, 117(2), 243–
253.
Briggs, G. (1969), Plume rise, Tech. Rep. Crit. Rev. Ser. TID-25075, 81pp., Natl. Tech. Inf.
Serv., Springfield, Va.
Briggs, G. A. (1984), Plume rise and buoyancy effects, in: Atmospheric science and power
production, edited by: Anderson, D., US Department of Energy, Germantown, Maryland, p. 855.
Carlson, T. N. 1979, Atmospheric turbidity in Saharan dust out- breaks as determined by
analyses of satellite brightness data, Mon. Wea. Rev., 107, 322–335.
Clinton, N., P. Gong, and K. Scott (2006), Quantification of pollutants emitted from very large
wildland fires in Southern California, USA, Atmos. Environ., 40, 3686–3695,
doi:10.1016/j.atmosenv.2006.02.016.
Davies, S. J., and L. Unam (1999), Smoke-haze from the 1997 Indonesian forest fires: effects on
pollution levels, local climate, atmospheric CO2 concentrations, and tree photosynthesis, Forest
Ecology and Management, 124, 137–144.
Debell, L. J., R. W. Talbot, J. E. Dibb, J. W. Munger, E. V. Fischer and S. E. Frolking (2004), A
major regional air pollution event in the northeastern United States caused by extensive forest
fires in Quebec, Canada, J. Geophys. Res, 109, D19305, doi:10.1029/2004JD004840.
Dempsey, F. (2013), Forest fire effects on air quality in Ontario: evaluation of several recent
examples, Bull. Amer. Meteor. Soc., 94, 1059–1064.
Dennison, P. E., S. C. Brewer, J. D. Arnold, and M. A. Moritz (2014), Large wildfire trends in
the western United States, 1984-2011, Geophysical Research Letters, 41, 2928–2933,
doi:10.1002/2014GL059576.
Division of Air Quality (DAQ) (2010a), PM10 and PM2.5 Exceptional Event –High wind, March
29
30th 2010. [Available online at http://www
http://deq.utah.gov/ProgramsServices/programs/air/exceptionalevents/docs/2013/07Jul/
highwindmarch302010.pdf.].
Division of Air Quality (DAQ) (2010b), PM10 and PM2.5 Exceptional Event –High wind, April
27th, 2010. [Available online at http://www
http://deq.utah.gov/ProgramsServices/programs/air/exceptionalevents/docs/2013/07Jul/highwind
april272010.pdf.].
Doney, S. C., N. Mahowald, I. Lima, R. A. Feely, F. T. Mackenzie, J.-F. Lamarque, and P. J.
Rasch (2007), Impact of anthropogenic atmospheric nitrogen and sulfur deposition on ocean
acidification and the inorganic carbon system, Proc. Natl. Acad. Sci. USA, 104, 14580–14585,
doi:10.1073/pnas.0702218104.
EPA (2011), National Ambient Air Quality Standards (NAAQS), [Available online at
www.epa.gov/air/criteria.html.].
Fairlie, T. D., D. J. Jacob, and R. J. Park (2007), The impact of transpacific transport of mineral
dust in the United States, Atmos. Environ., 41, 1251–1266.
Flannigan, M.D., M. A. Krawchuk, W. J. de Groot, B. M. Wotton, L. M. Gowman (2009),
Implications of changing climate for global wildland fire, Int. J. Wildland Fire, 18, 483–507.
Forster, P., and Coauthors (2007), Changes in atmospheric constituents and in radiative forcing.
In: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the
Fourth Assessment Report of the Inter- governmental Panel on Climate Change, Cambridge
University Press, Cambridge.
Freitas, S. R., and coauthors (2007), Including the sub-grid scale plume rise of veg- etation fires
in low resolution atmospheric transport models, Atmos. Chem. Phys., 7, 3385–3398.
Fromm, M., R. Bevilacqua, R. Servranckx, J. Rosen, J. P. Thayer, J. Herman, and D. Larko
(2005), Pyro-cumulonimbus injection of smoke to the stratosphere: Observations and impact of a
super blowup in northwestern Canada on 3–4 August 1998, J. Geophys. Res., 110, D08205,
doi:10.1029/2004JD005350.
Fu, X., S. X. Wang, Z. Cheng, J. Xing, B. Zhao, J. D. Wang, and J. M. Hao (2014), Source,
transport and impacts of a heavy dust event in the Yangtze River Delta, China, in 2011, Atmos.
Chem. Phys., 14 1239–1254.
30
Galanter, M., H. Levy, and G. R. Carmichael (2010), Impacts of biomass burning on
tropospheric CO, NOx, and O3, J. Geophys. Res., 105(D5), 6633–6653.
Gebhart, K. A., S. M. Kreidenweis, and W. C. Malm (2001), Back-trajectory analyses of fine
particulate matter measured at Big Bend National Park in the historical database and the 1996
scoping study, Sci. Total Environ., 276, 185–204.
Generoso, S., I. Bey, J.-L. Atti, and F.-M Bron (2007), A satellite and model-based assessment
of the 2003 Russian fires: Impact on the Arctic region, J. Geophys. Res., 112, D15302,
doi:10.1029/2006JD008344.
Gerbig, C., and coauthors (2002), Toward constraining regional-scale fluxes of CO2 with
atmospheric observations over a continent: 2. Analysis of COBRA data using a receptororiented framework, J. Geophys. Res., 108(D24), 4757, doi:10.1029/2003JD003770.
Gerbig, C., S. Korner, and J. C. Lin (2008), Vertical mixing in atmospheric tracer transport
models: Error characterization and propagation, Atmos. Chem. Phys., 8, 591–602.
Giglio, L., J. Descloitres, C. O. Justice, and Y. J. Kaufman, Y. J., (2003), An enhanced
contextual fire detection algorithm for MODIS, Remote Sens. Environ., 87, 273–282.
Giglio, L., T. Loboda, D. P. Roy, B. Quayle, and C. O. Justice, (2009), An active-fire based
burned area mapping algorithm for the MODIS sensor, Remote Sens. Environ., 113, 408–420.
Gill, T.E. (1996), Eolian sediments generated by anthropogenic disturbance of playas: human
impacts on the geomorphic system and geomorphic impacts on the human system,
Geomorphology, 17, doi:10.1016/0169-555X(95)00104-D.
Ginoux, P., M. Chin, I. Tegen, J. M. Prospero, B. Holben, O. Dubovik, and S.-J. Lin, ( 2001),
Sources and distributions of dust aerosols simulated with the GOCART model. Journal of
Geophysical Research, 106, 20, doi:10.1029/2000JD000053.
Good S. P., D. V. Mallia, E. H. Denis, K. H. Freeman, X. Feng, S. Li, N. Zegre, J. C. Lin, and G.
J. Bowen (2014), High frequency trends in the isotopic composition of Superstorm Sandy, In
Learning from the Impacts of Superstorm Sandy, Elsevier, edited by J. B. Bennington and E. C.
Farmer.
Good S. P., D. V. Mallia, G. Bowen, and J. C. Lin (2014), Stable Isotope Analysis of
Precipitation Samples Obtained Via Crowdsourcing Reveals the Spatiotemporal Evolution of
Superstorm Sandy, PLoS One, 9(3), doi:10.1371/journal.pone.0091117.g001.
Goudie, A. S., and N. J. Middleton (2001), Saharan dust storms: Nature and consequences. Earth
31
Sci. Rev., 56, 197–204.
Guldberg, P. H (1975), A comparison study of plume rise formulas ap- plied to tall stack data, J.
Appl. Meteor. Climatol., 14, 1402–1405.
Hahnenberger, M., and K. Nicoll (2012), Meteorological characteristics of dust storm events in
the eastern Great Basin of Utah, U.S.A., Atmospheric Environment, 60, 601–612.
Hegarty, J., et al. (2013), Evaluation of Lagrangian Particle Dispersion Models with
Measurements from Controlled Tracer Releases, J. Appl. Meteor. Climatol., 52, 2623–2637.
Huang, K., G. Zhuang, J. Li, Q. Wang, Y. Sun, Y. Lin, and J. S. Fu (2010), Mixing of Asian dust
with pollution aerosol and the transformation of aerosol components during the dust storm over
China in spring 2007, J. Geophys. Res.-Atmos., 115, D13, doi:10.1029/2009jd013145.
Husar, R. B., and Coauthors (2001), Asian dust events of April 1998. J. Geophys. Res., 106,
18317–18330.
Hyer, E. J., D. J. Allen, and E. S. Kasischke (2007), Examining injection properties of boreal
forest fires using surface and satellite measure- ments of CO transport, J. Geophys. Res., 112,
D18307, doi:10.1029/ 2006JD008232.
Ichinose, T., S. Yoshida, K. Hiyoshi, K. Sadakane, H. Takano, M. Nishikawa, I. Mori, R.
Yanagisawa, H. Kawazato, A. Yasuda, T. and Shibamoto (2008), The effects of microbial
materials adhered to Asian sand dust on allergic lung inflammation, Arch. Environ. Cont.
Toxicol., 55, 348–357, doi:10.1007/s00244-007- 9128-8.
Intergovernmental Panel on Climate Change (IPCC), 2013: Climate Change 2013: The Physical
Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the
Intergovern- mental Panel on Climate Change, edited by Stocker, T.F., D. Qin, G.-K. Plattner,
M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley, Cambridge
University Press, Cambridge, United Kingdom and New York, 1535 pp.
Jaffe, D., and Coauthors (1999), Transport of Asian air pollution to North America, Geophys.
Res. Lett., 26, 711–714.
Jewell, P. W., and K. Nicoll (2011), Wind regimes and aeolian transport in the Great Basin,
U.S.A., Geomorphology, 129, 1e13, doi:10.1016/j.geomorph.2011.01.005.
Kahn, R. A., Y. Chen, D. L. Nelson, F.-Y. Leung, Q. Li, D. J. Diner, and J. A. Logan (2008),
Wildfire smoke injection heights: Two perspectives from space, Geophys. Res. Lett., 35, L04809,
32
doi:10.1029/ 2007GL032165.
Kanakidou, M. and Coauthors (2005), Organic aerosol and global climate modelling: a review,
Atmos. Chem. Phys. 5, 1053–1123.
Kang, J. H., T. C. Liu, J. Keller, and H. C. Lin (2013), Asian dust storm events are associated
with an acute increase in stroke hospitalization, J. Epidemiol. Comm. Health, 67, 125–131,
doi:10.1136/jech-2011-200794.
Keane R. E., J. M. Herynk, C. Toney, S. P. Urbanski, D. C. Lutes, R. D. Ottmar, (2013),
Evaluating the performance and mapping of three fuel classification systems using Forest
Inventory and Analysis surface fuel measurements, Forest Ecology and Management, 305, 248–
263.
Kim, S., D. Millet, L. Hu, M. Mohr, T. Griffis, D. Wen, J.C. Lin, S. Miller, and M. Longo,
(2013) Constraints on carbon monoxide emissions in the United States based on tall tower
measurements, Environmental Science and Technology, 47, 8316-8324, doi:10.1021/es4009486.
Klass, D. L. (1998), Biomass for Renewable Energy, Fuels, and Chemicals. Academic Press, San
Diego, CA, pg. 651.
Koren, I., Y. J. Kaufman, L. Remer, J. V. Martins (2004), Measurement of the effect of Amazon
smoke on inhibition of cloud formation, Science, 303, 1342–1345.
Labonne, M., Breon, F.-M., and Chevallier, F (2007), Injection height of biomass burning
aerosols as seen from a spaceborne lidar, Geo- phys. Res. Lett., 34, 1–5,
doi:10.1029/2007GL029311.
LANDFIRE (2014), LANDFIRE, [Available online at http://www.landfire.gov/].
Lee, S., B. Choi, S. M. Yi, and G. Ko (2009), Characterization of microbial community during
Asian dust events in Korea, Sci. Total Environ., 407, 5308–5314,
doi:10.1016/j.scitotenv.2009.06.052.
Lin, J. C., C. Gerbig, S. C. Wofsy, A. E. Andrews, B. C. Daube, K. J. Davis, and C. A. Grainger
(2003), A near-field tool for simulating the upstream influence of atmospheric observations: the
Stochastic Time-Inverted Lagrangian Transport (STILT) model, J. Geophys. Res., 108(4493),
D003161, doi: 10.1029/2002JD003161.
Lin, J. C., D. Brunner, C. Gerbig, A. Stohl, A. Luhar, and P. Webley (2013), Lagrangian
Modeling of the Atmosphere, AGU Geophysical Monograph, 200, 349 pp.
Lin, J. C., and C. Gerbig (2005), Accounting for the effect of transport errors on tracer
33
inversions, Geophys. Res. Lett., 32, L01802, doi:10.1029/ 2004GL021127.
Littell J. S., D. McKenzie, D. L. Peterson, A. L. Westerling (2009), Climate and wildfire area
burned in western US ecoprovinces, 1916–2003, Ecol. Appl., 19, 1003–1021.
Liu, M., D. L. Westphal, A. L. Walker, T. R. Holt, K. A. Richardson, and S. D. Miller (2006),
COAMPS Read-time dust storm forecast during Operation Iraqi Freedom, Weather and
Forecasting, 22, 192–205.
Liu, Y., Goodrick, S., and W. Heilman (2014), Wildland fire emissions, carbon, and
climate: wildfire-climate interactions, For. Ecol. Manage., 317, 80–96.
Liu, Y., G. L. Achtemeier, S. L. Goodrick, and W. A. Jackson (2010), Important parameters for
smoke plume rise simulation with Daysmoke, Atmospheric Pollution Research, 1, 250–259.
Mallia, D. V., J. C. Lin, S. Urbanski, J. Ehleringer, and T. Nehrkorn (2015), Impacts of upwind
wildfire emissions on CO, CO2, and PM2.5 concentrations in Salt Lake City, Utah, Journal of
Geophysical Research-Atmospheres, 120, 147–166.
Mandel J., S. Amram, J. Beezley, G. Kelma, A. Kochanski, V. Kondratenko, B. Lynn, B. Regev,
and M. Vejmelka (2014) Recent advances and applications of WRF-SFIRE, Natural Hazards
and Earth System Science, 14, 2829–2845, doi:10.5194/nhess-14-2829-2014.
Martin, D., D. M. Webber, S. J. Jones, B. Y. Underwood, G. A. Tickle, and S. A. Ramsdale
(1997), Near and intermediate-field dispersion from strongly buoyant sources, AEAT/1388,
Final Report.
Massey, J. D., W. J. Steenburgh, S. W. Hoch, and J. C. Knievel (2014), Sensitivity of NearSurface Temperature Forecasts to Soil Properties over a Sparsely Vegetated Dryland Region. J.
Appl. Meteor. Climatol., 53, 1976–1995.
McKenna, D. S., P. Konopka, J.-U. GrooB, G. Gunther, and R. Muller (2002a), A new Chemical
Lagrangian Model of the Stratosphere (CLaMS) 1. Formulation of advection and mixing,
Journal of Geophysical Research, 107(D16).
Mu, M., J. T. Randerson, G. R. van der Werf, L. Giglio, P. Kasibhatla, D. Morton, G. J. Collatz,
R. S. DeFries, E. J. Hyer, E. M. Prins, D. W. T. Griffith, D. Wunch, G.C. Toon, V. Sherlock, and
P.O. Wennberg (2010), Daily and 3-hourly variability in global fire emissions and consequences
for atmospheric model predictions of carbon monoxide. J. Geophys. Res., 116: D24303,
doi:10.1029/2011JD016245.
Nehrkorn, T., J. Eluszkiewicz, S. C. Wofsy, J. C. Lin, C. Gerbig, M. Longo, and S. Freitas
34
(2010), Coupled Weather Research and Forecasting–Stochastic Time-Inverted Lagrangian
Transport (WRF–STILT) model, Meteorology and Atmospheric Physics, 107, 51–64.
Nikmo, J., J.-P. Tuovinen, J. Kukkonen, J., and I. Valkama (1999), A hybrid plume model for
local-scale atmospheric dispersion, Atmos. Environ., 33, 4389–4399, doi:10.1016/S13522310(99)00223-X.
Park, R. J., D. J. Jacob, and J. A. Logan (2007), Fire and biofuel contributions to the annual
mean aerosol mass concentrations in the United States, Atmos. Environ., 41, 7389–7400,
doi:10.1016/j.atmosenv.2007.05.061.
Park, S. H., S. L. Gong, W. Gong, P. A. Makar, M. D. Moran, C. A. Stroud, and J. Zhang (2009),
Sensitivity of surface characteristics on the simulation of wind-blown- dust source in North
America, Atmospheric Environment, 43, 3122e3129, doi:10.1016/j.atmosenv.2009.02.064.
Pope, C. A., D. V. Bates, and M. E. Raizenne (1995), Health effects of particulate air pollution:
Time for reassessment?, Environ. Health Perspect., 103, 472–480.
Prospero, J. M., and P. J. Lamb (2003), African droughts and dust transport to the Caribbean:
Climate change implications. Science, 302, 1024–1027.
Raffuse, S. M., K. J. Craig, N. K. Larkin, T. T. Strand, D. C. Sullivan, N. J. M. Wheeler, and R.
Solomon (2012), An evaluation of modeled plume injection height with satellite-derived
observed plume height, Atmosphere, 3(1), 103–123, doi:10.3390/atmos3010103.
Reynolds, R., J. Belnap, M. Reheis, P. Lamothe, and F. Luiszer (2001), Aeolian dust in Colorado
Plateau soils: Nutrient inputs and recent change in source, Proc. Natl. Acad. Sci. USA, 98, 7123–
7127.
Riley, K. L., J. T. Abatzoglou, T. G. John, C. Isaac, A. E. Klene, F. A. Heinsch, 2013, The
relationship of large fire occurrence with drought and fire danger indices in the western USA,
1984-2008: the role of temporal scale. International Journal of Wildland Fire, 22(7), 894–909.
Rind, D., M. Chin, G. Feingold, D. Streets, R. A. Kahn, S. E. Schwartz, and H. Yu (2009),
Modeling the effects of aerosols on climate, in atmospheric aerosol properties and climate
impacts, A Report by the US Climate Change Science Program and the Subcommittee on Global
Change Research, National Aeronautics and Space Administration, Washington, DC.
Ruefenacht, B., M. V. Finco, M. D. Nelson, R. Czaplewski, E. H. Helmer, J. A. Blackard, G. R.
Holden, A. J. Lister, D. Salaanu, D. Weyermann, and K. Winterberger, (2008), Conterminous US
and Alaska Forest Type Mapping Using Forest Inventory and Analysis Data, Photogramm. Eng.
Remote Sens., 74, 1379–1388.
35
Sapkota, A., J. M. Symons, J. Kleissl, L. Wang, M. B. Parlange, J. Ondov, P. N. Breysse, G. B.
Diette, P. A. Eggleston, and T. Buckley (2005), Impacts of the 2002 Canadian forest fires on
particulate matter air quality in Baltimore City, Environ. Sci. Technol., 39(1), 24–32.
Skamarock, W., J. B. Klemp, J. Dudhia, D. O. Gill, D. M. Barker, M. G. Duda, X. Y. Huang, W.
Wang, and J. G. Powers (2008), A description of the Advanced Research WRF Version 3, NCAR
Technical Note, 126-126.
Shi, J. H., H. W. Gao, J. Zhang, S. C. Tan, J. L. Ren, C. G. Liu, Y. Liu, and X. Yao (2012),
Examination of causative link be- tween a spring bloom and dry/wet deposition of Asian dust in
the Yellow Sea, China, J. Geophys. Res.-Atmos., 117, doi:10.1029/2012jd017983.
Shin, S., and S. Reich (2009), Hamiltonian particle-mesh simulations for a non- hydrostatic
vertical slice model, Atmospheric Science Letters, 10, 233–240.
Smolarkiewicz, P., and J. A. Pudykiewicz (1992), A class of semi-Lagrangian approximations
for fluids, Journal of the Atmospheric Sciences, 49, 2082–2096.
Sofiev, M., T. Ermakova, and R. Vankevich (2012), Evaluation of the smoke-injection height
from wild-land fires using remote-sensing data, Atmos. Chem. Phys., 12, 1995–2006.
Sokolik, I. N., D. M. Winker, G. Bergametti, D. A. Gillette, G. Carmichael, Y. J. Kaufman, L.
Gomes, L. Schuetz, and J. E. Penner (2001), Introduction to special section: Outstanding
problems in quantifying the radiative impacts of mineral dust, J. Geophys. Res.-Atmos., 106,
18015–18027, doi:10.1029/2000jd900498.
Spracklen, D. V., L. J. Mickley, J. A. Logan, R. C. Hudman, R. Yevich, M. D. Flannigan, A. L.
Westerling (2009), Impacts of climate change from 2000 to 2050 on wildfire activity and
carbonaceous aerosol concentrations in the western United States, J. Geophys. Res., 114(D20),
1984–2012.
Steenburgh, W. J., J. D. Massey, and T. H. Painter (2012), Episodic dust events of Utah’s
Wasatch Front and adjoining region, Journal of Applied Meteorology, 51, 1654–1669.
Susott, R. A., W. F. DeGroot, and Shafizadeh (1975), Heat content of natural fuels, J. Fire
Flammability, 6, 311–325.
Tong, D. Q., G. E. Bowker, S. He, D. W. Byun, R. Mathur, and D. A. Gillette (2011),
Development of a Windblown Dust Module within the Community Multi-scale Air Quality
(CMAQ) Model: Description and Preliminary Applications in the Continental United States,
submitted J. Geophys. Res.
36
Urbanski, S. P., W. M. Hao, and B. Nordgren (2011), The wildland fire emission inventory:
western United States emission estimates and an evaluation of uncertainty, Atmospheric
Chemistry and Physics, 11, 12973-13000, doi:10.5194/acp-11-12973-2011.
Urbanski, S. P. 2013. Combustion efficiency and emission factors for wildfire-season fires in
mixed conifer forests of the northern Rocky Mountains, US. Atmospheric Chemistry and
Physics, 13, 7241–7262.
U.S. Census Bureau, (2010), State and County QuickFacts, http://quickfacts.census.
gov/qfd/states/49/49035.html.
Val Martin, M., J. A. Logan, R. A. Kahn, F.-Y., Leung, D. L. Nelson, and D. J. Diner (2010),
Smoke injection heights from fires in North America: analysis of 5 years of satellite
observations, Atmos. Chem. Phys., 10, 1491-1510, doi:10.5194/acp-10-1491-2010.
Val Martin, M., C. L. Heald, B. Ford, A. J. Prenni, and C. Wiedinmyer (2013), A decadal
satellite analysis of the origins and impacts of smoke in Colorado, Atmos. Chem. Phys., 13,
7429-7439, doi:10.5194/acp-13-7429-2013.
VanCuren, R. A., and T. A. Cahill (2002), Asian aerosols in North America: Frequency and
concentration of fine dust, J. Geophys. Res., 107, D12, 4804, doi:10.1029/2002JD002204.
van der Werf, G. R., J. T. Randerson, L. Giglio, G. J. Collatz, M. Mu, P. S. Kasibhatla, D. C.
Morton, R. S. DeFries, Y. Jin, and T. T. van Leeuwen (2010), Global fire emissions, and the
contribution of deforestation, savanna, forest, agricultural, and peat fires (1997-2009),
Atmospheric Chemistry and Physics, 10, 11707-11735, doi:10.5194/acp-10-11707-2010.
Washington, R., M. C. Todd, G. Lizcano, I. Tegen, C. Flamant, I. Koren, P. Ginoux, S.
Engelstaedter, C. S. Bristow, C. S. Zender, A. S. Goudie, A. Warren, and J. M Prospero (2006),
Links between topography, wind, deflation, lakes and dust: the case of the Bodele Depression,
Geophysical Research Letters, 3, L09401, doi:10.1029/2006GL025827.
Weil, J. (1988), Plume rise, in: Lectures on Air Pollution Modelling, edited by: Venkatram, A.
and Wingaard, J. C., American Meteorological Soc., Boston, p. 390.
Wen, D., J. C. Lin, D. Millet, A. Stein, and R. Draxler (2012), A backward-time stochastic
Lagrangian air quality model, Atmospheric Environment, 54, 373–386,
doi:10.1016/j.atmosenv.2012.02.042.
Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam (2006), Warming and earlier
spring increases western U.S. forest fire activity, Science, 313, 940–943.
37
Wiedinmyer, C., and J. C. Neff (2007), Estimates of CO2 from fires in the United States:
implications for carbon management, Carbon Balance and Management, 2(10),
doi:10.1186/1750-0680-2-10.
Wilson, B. T., C. W. Woodall, and D. M. Griffith (2013), Imputing forest carbon stock estimates
from inventory plots to a nationally continuous coverage, Carbon Balance and Management, 8,
doi:10.1186/1750-0680-8-1.
Wohltmann, I., and M. Rex (2009), The Lagrangian chemistry and transport model ATLAS:
validation of advective transport and mixing, Geoscientific Model Development, 2, 153–173
Woodward, S., 2001, Modeling the atmospheric life cycle and radiative impact of mineral dust in
the Hadley Centre climate model. Journal of Geophysical Research, 106, D16,
doi10.1029/2000JD900795.
Wotawa, G. and Trainer, M. (2000), The influence of Canadian forest fires on pollutant
concentrations in the United States, Science, 288(5464), 324–328.
Yokelson, R. J., and coauthors (2009), Emissions from biomass burning in the Yucatan, Atmos.
Chem. Phys., 9(15), 5785–5812, doi:10.5194/acp-9-5785-2009.
Zender, C. S., H. S. Bian, and D. Newman (2003), Mineral Dust Entrainment and Deposition
(DEAD) model: Description and 1990s dust climatology, J. Geophys. Res.-Atmos., 108, 4416,
doi:10.1029/2002JD002775.
38