Organic aerosol (OA) is a major component of smoke plumes
from open biomass burning (BB). Therefore, adequate representation of the
atmospheric transformations of BB OA in chemistry-transport and climate
models is an important prerequisite for accurate estimates of the impact of
BB emissions on air quality and climate. However, field and laboratory
studies of atmospheric transformations (aging) of BB OA have yielded a wide
diversity of observed effects. This diversity is still not sufficiently
understood and thus not addressed in models. As OA evolution is governed by
complex nonlinear processes, it is likely that at least a part of the
observed variability in the BB OA aging effects is due to the factors
associated with the intrinsic nonlinearity of the OA system. In this study,
we performed a numerical analysis in order to gain a deeper understanding of
these factors. We employ a microphysical dynamic model that represents
gas–particle partitioning and OA oxidation chemistry within the volatility
basis set (VBS) framework and includes a schematic parameterization of BB OA
dilution due to dispersion of an isolated smoke plume. Several VBS schemes
of different complexity, which have been suggested in the literature to
represent BB OA aging in regional and global chemistry-transport models, are
applied to simulate BB OA evolution over a 5 d period representative of
the BB aerosol lifetime in the dry atmosphere. We consider the BB OA mass
enhancement ratio (EnR), which is defined as the ratio of the mass
concentration of BB OA to that of an inert tracer and allows us to eliminate
the linear part of the dilution effects. We also analyze the behavior of the
hygroscopicity parameter, κ, that was simulated in a part of our
numerical experiments. As a result, five qualitatively different regimes of
OA evolution are identified, which comprise (1) a monotonic saturating
increase in EnR, (2) an increase in EnR followed by a decrease, (3) an
initial rapid decrease in EnR followed by a gradual increase, (4) an EnR
increase between two intermittent stages of its decrease, or (5) a gradual
decrease in EnR. We find that the EnR for BB aerosol aged from a few hours
to a few tens of hours typically increases for larger initial sizes of the
smoke plume (and therefore smaller dilution rates) or for lower initial OA
concentrations (and thus more organic gases available to form secondary OA – SOA).
However, these dependencies can be weakened or even reversed, depending on
the BB OA age and on the ratio between the fragmentation and
functionalization oxidation pathways. Nonlinear behavior of BB OA is also
exhibited in the dependencies of κ on the parameters of the plume.
Application of the different VBS schemes results in large quantitative and
qualitative differences between the simulations, although our analysis
suggests also that the main qualitative features of OA evolution simulated
with a complex two-dimensional VBS scheme can also be reproduced with a much
simpler scheme. Overall, this study indicates that the BB aerosol evolution
may strongly depend on parameters of the individual BB smoke plumes (such as
the initial organic aerosol concentration and plume size) that are typically
not resolved in chemistry-transport models.
Introduction
Atmospheric aerosol is known to play an important role as a climate driver
on global and regional scales and to adversely affect human health. A large
fraction of the aerosol mass is typically represented by organic components
forming liquid, amorphous, or glassy particulate matter, which here is
referred to as organic aerosol (OA). As a climate forcer, OA scatters solar
radiation and provides cloud condensation nuclei, thus directly and
indirectly contributing to cooling of the atmosphere on the global scale
(IPCC, 2013; Lelieveld et al., 2019), although part of it, so-called brown
carbon, can absorb sunlight, thus contributing to warming (see, for example,
Andreae and Gelencsér, 2006; Feng et al., 2013; Jo et al., 2016). On a
regional scale, of particular significance is the cooling effect of OA on
climate in the Arctic (Sand et al., 2015), opposing the rapid increase in
surface temperature that has been observed in recent decades (Bekryaev et
al., 2010). As an agent of air pollution, OA constitutes a considerable
fraction of fine particulates (PM2.5; Jimenez et al., 2009) that cause
human health disorders and premature deaths (Pope et al., 2009; Burnett et
al., 2018; Lelieveld et al., 2019). However, as evidenced by the large
differences between the OA atmospheric budgets evaluated with different
models and also by considerable discrepancies between simulations and
observations of OA (see, for example, Tsigaridis et al., 2014; Bessagnet et al.,
2016; Tsigaridis and Kanakidou, 2018), the current knowledge of the sources
and atmospheric transformations of OA is still deficient, and corresponding
modeling representations are very imperfect.
Open biomass burning (BB), i.e., vegetation fires and agricultural waste
burning, provides a major source of OA on the global scale. Specifically, it
has been estimated that BB emissions of primary OA (POA), which typically
constitutes the predominant fraction of BB aerosol, contribute about 70 %
of total POA emissions (Bond et al., 2013). In recent years, numerous
studies have been aimed at investigating and modeling sources (e.g., May et
al., 2013; Jathar et al., 2014; Konovalov et al., 2015; van der Werf et al.,
2017), radiative effects (e.g., Saleh et al., 2013, 2015; Archer-Nicholls et
al., 2016; Pokhrel et al., 2017; Yao et al., 2017), and atmospheric
transformations (e.g., Cubison et al., 2011; Jolleys et al.,
2012; Forrister et al., 2015; Shrivastava et al., 2015; Konovalov
et al., 2015, 2017; Tsimpidi et al., 2018; Theodoritsi and Pandis, 2019) of
BB OA and its components.
Both laboratory experiments and ambient observations suggest that the mass
concentration of BB OA may undergo major, yet highly diverse, changes as a
result of its aging under typical atmospheric conditions. These changes are
commonly evaluated by means of the BB OA mass enhancement ratio (EnR), which
is usually defined as the normalized ratio of the BB OA mass concentration
to the concentration of an inert BB tracer. This normalization makes the EnR
for freshly emitted aerosol equal to unity. In particular, considerable
increases (in many cases exceeding a factor of 2) in EnR were found in smog
chamber experiments after a few hours of photochemical aging of smoke from
wood or grass burning (e.g., Grieshop et al., 2009; Hennigan et al., 2011;
Tiitta et al., 2016; Ciarelli et al., 2017a; Fang et al., 2017; Ahern et
al., 2019), although there has been a large diversity between results of
individual chamber experiments. As a result of aircraft experiments
conducted in North America around Mexico City and on the Yucatan Peninsula,
significant increases in EnR have been reported by DeCarlo et al. (2008) and
Yokelson et al. (2009) for aging BB plumes (from about 30 % up to a
factor of 2). Konovalov et al. (2015) identified a substantial increase (by
a factor of 2) in the enhancement ratio for mass concentration of
particulate matter in smoke plumes after 1–2 d of transport over regions
of eastern Europe. A similar major increase in the enhancement ratio for BB
aerosol mass concentration, but over about 15 h of photochemical
oxidation of BB plumes, was deduced from an analysis of satellite
measurements of aerosol optical depth (AOD) over Siberia (Konovalov et al.,
2017). Based on several years of continuous measurements of BB OA in an
African savannah, Vakkari et al. (2018) found that EnR more than doubles on
average after 3 h of daytime aging. However, there is also evidence
that EnR may decrease or remain nearly constant in aging smoke plumes. For
example, based on aircraft measurements, Akagi et al. (2012) identified a
sharp decrease in EnR during the first hour after emissions. Using data from
several field campaigns conducted in Australia, North America, and western
Africa, Jolleys et al. (2012, 2015) found that the BB OA enhancement ratios
in highly aged BB plumes (typically transported between 3 and 6 d
before the measurements were taken) were consistently smaller than those in
the fresh plumes. The aforementioned analysis of satellite data (Konovalov
et al., 2017) suggested evidence for a gradual decrease in EnR after its
initial strong increase. At the same time, several observational studies
(e.g., Capes et al., 2008; Brito et al., 2014; Sakamoto et al., 2015; May et
al., 2015; Zhou et al., 2017) did not reveal any significant net changes of
EnR in aged BB plumes.
Numerous studies reported major changes in the chemical composition of BB OA
due to its aging (e.g., DeCarlo et al., 2008; Cubison et al., 2011; Pratt et
al., 2011; Jolleys et al., 2012, 2015; Brito et al., 2014; May et al., 2015;
Bertrand et al., 2018; Lim et al., 2019) regardless of whether
significant net changes were detected in the BB OA mass concentration or not. On
the one hand, BB OA aging has been found to be typically associated with a
rapid decay (over a period of a few hours under typical atmospheric
conditions) of some key chemical compounds contributing to POA (such as,
for example, levoglucosan): using aerosol mass spectrometry, such a decay, can be
inferred from a decrease in the mass fragment signatures at m/z 60 (e.g., May
et al., 2015) as well as from a more comprehensive analysis at the molecular
level (Bertrand et al., 2018). On the other hand, atmospheric processing of
BB OA has been reported to result in strong enhancements of the oxidation
state of the organic matter. The increases in the O:C ratio (due to addition
of, for example, alcohol and carbonyl groups) are usually inferred from increases
in the mass spectrometric signal at m/z 44 (e.g., Brito et al., 2014; May et
al., 2015; Fang et al., 2017) and can be indicative of secondary organic
aerosol (SOA) formation. Note that changes in the chemical composition and
oxidation state of OA particles can affect their hygroscopic and optical
properties (e.g., Lambe et al., 2011; Adler et al., 2011; Akagi et al.,
2012; Fan et al., 2019), which need to be adequately specified in
chemistry-transport and climate models.
Useful insights into the possible reasons behind the large variability in
the EnR trends reported earlier for aging BB aerosol have been provided by
recent smog chamber experiments (Ahern et al., 2019; Lim et al., 2019) that
revealed a strong dependence of SOA formation on variable initial
concentrations of organic gases. These experiments, however, do not rule out
the possibility that there are some other factors contributing significantly
to the observed diversity of changes in EnR during the atmospheric lifetime
of BB aerosol. In view of significant nonlinear interactions of the
processes affecting properties, formation, and evolution of SOA (Shrivastava
et al., 2017), it seems reasonable to expect that the diversity of
observational findings concerning BB aerosol atmospheric aging can in part
be due to nonlinear behavior of OA transformations. By nonlinear behavior,
we mean here any manifestations of a dependence of the relative rate of
change of OA mass concentration at a given moment of time on the mass
concentration of OA itself at the current moment or previous moments.
In this study, we investigate qualitative nonlinear features of the behavior
of OA within an isolated BB plume and attempt to reconcile some of the
diverse observational findings concerning BB aerosol aging effects from a
theoretical viewpoint. To this end, using some routines and interfaces of
the CHIMERE chemistry-transport model (Menut et al., 2013), we developed and
employed a microphysical dynamic (box) model of organic aerosol (MDMOA).
While three-dimensional chemistry-transport models are intended to provide
the best possible quantitative representation of the evolution of OA and its
gaseous precursors from various anthropogenic and natural sources, the
principal purpose of MDMOA is to isolate and simulate, under fixed ambient
conditions, the effects of key processes responsible for chemical and
physical transformations of OA from other complex processes affecting
evolution of OA in the real atmosphere (such as mixing with aerosols and
their gaseous precursors from multiple sources, vertical advection, dry and
wet deposition, in-cloud processing, etc.). In this sense, our study is
similar to several previous studies employing box models to study OA
processes (e.g., Camredon et al., 2007; Lee-Taylor et al., 2011, 2015;
Lannuque et al., 2018). Note that while the spatial scales representative of
isolated BB plumes are typically not resolved by chemistry-transport models,
simulations of a single BB plume with a box model can provide useful
insights into possible uncertainties introduced by neglecting the spatial
inhomogeneity of BB OA emissions in chemistry-transport models at the
sub-grid scales. Furthermore, compared to smog chamber and dedicated field
studies, a box-model analysis enables a much more comprehensive examination
of the parameter space of the BB OA system.
It has been proposed that complex atmospheric transformations of OA
(regardless of its origin), including SOA formation, can be adequately
represented in chemistry-transport models within the volatility basis set
(VBS) modeling framework (Donahue et al., 2006, 2011, 2012a; Robinson et
al., 2007). This framework has been implemented in MDMOA. The VBS method
involves splitting semi-volatile organic compounds (SVOCs), and the more
volatile intermediate-volatility organic compounds (IVOCS), into several
classes with respect to volatility and applying the absorptive partitioning
theory (Pankow, 1994) to distribute the organic compounds between gas
phase and particles. The SVOCs and IVOCs can also be distributed between
several model types, depending, for example, on their oxidation state (O:C ratio),
origin (e.g., primary or secondary, anthropogenic or biogenic, etc.), and
photochemical age (Donahue et al., 2012a, b; Shrivastava et al., 2013;
Tsimpidi et al., 2018). Representing the processes involving SVOCs and IVOCs
within the VBS framework has been shown to allow improving the performance
of simulations of OA from vegetation fires with respect to simulations using
the “conventional” OA modeling framework, in which these processes are
basically disregarded and only specific volatile organic compounds (VOCs)
are considered to be precursors of SOA (Hodzic et al., 2010; Shrivastava et
al., 2015; Konovalov et al., 2015, 2017). Based on simulations using the VBS
method, it has also been argued (Konovalov et al., 2015, 2017) that
disregard for the BB OA aging processes might be one of the main reasons for
a strong underestimation of aerosol optical depth in BB plumes by chemistry-transport models using the conventional representation of OA evolution
(e.g., Tosca et al., 2013; Konovalov et al., 2014, 2018; Reddington et al.,
2016; Petrenko et al., 2017). It should be noted, however, that the
representation of the BB OA evolution within the VBS framework in chemistry-transport models is still associated with major uncertainties: while a
variety of VBS schemes of different complexities have been suggested for BB
OA modeling (e.g., Grieshop et al., 2009; Koo et al., 2014; Shrivastava et
al., 2015; Ciarelli et al., 2017a; Tsimpidi et al., 2018), any of these
schemes has only partially been constrained by laboratory or ambient
measurements. In view of these uncertainties, we performed our analysis
using several different available VBS schemes.
Based on simulations of the first few hours of BB OA evolution with a
similar microphysical box model, Bian et al. (2017) and Hodshire et al. (2019) showed that apart from oxidation, evaporation and condensation of
SVOCs, BB OA dynamics is strongly affected by the dilution process. Hodshire
et al. (2019) also pointed out a significant impact of the background
aerosol on near-field BB OA aging processes. Accordingly, both dilution and
entrainment of background aerosol are also taken into account in our model
(although investigating the role of the latter process is not the focus of
this study). Following Bian et al. (2017) and Hodshire et al. (2019), we
approximate the dilution rate as a function of the initial plume size by
using the formulations of the stationary Gaussian dispersion model and
analyze the dependence of the BB OA mass enhancement ratio on the initial
plume size, which controls the dilution rate. However, we considerably
extend the period of analysis (up to 5 d), and instead of a simple
single-step oxidation scheme that was used by Bian et al. (2017) to analyze
atmospheric implications of short-run smoke chamber experiments and by
Hodshire et al. (2019) to investigate the near-source relationships between
parameters of diluting BB plumes and BB OA physical properties, we use
several multi-step oxidation schemes that have been suggested for modeling
of BB OA specifically within chemistry-transport models. Along with the
dynamics of EnR, we consider the evolution of the hygroscopicity parameter,
κ, which is commonly used to characterize the uptake of water by
aerosol particles and their cloud condensation nucleus (CCN) activity (e.g.,
Petters and Kreidenweis, 2007; Chang et al., 2010; Mikhailov et al., 2015).
The evolution of the hygroscopicity parameter, however, is not the main
focus of this study.
By using MDMOA, we address the following questions. What are the
manifestations of the OA system's nonlinearity in the dependencies of EnR
and κ on the initial size and initial density of a smoke plume? Can
variability in the parameters of the plume lead to qualitatively different
types of BB OA evolution? Can differences between available VBS schemes be
associated with qualitatively different responses of EnR to variations in
the plume's parameters? It should be emphasized that our simple model and
its application in this study are not intended to reproduce any realistic
scenarios of atmospheric evolution of BB OA in a quantitatively accurate
way. Instead, we focus our analysis on identification of possible
qualitative features of the BB OA behavior, which may have a sufficiently
general character. We believe that the results of this kind of analysis can
be useful as theoretical guidance for future experimental studies and for
improving parameterizations of BB OA processes in chemistry-transport
models.
Model and method descriptionMicrophysical dynamic model of organic aerosol (MDMOA): dynamic
equations
The CHIMERE-based box model, MDMOA, is intended to represent the following
processes: (1) growth and evaporation of multi-disperse particles of OA due
to partitioning of SVOCs between gas phase and particles; (2) gas-phase
oxidation of VOCs, IVOCs, and SVOCs; and (3) atmospheric dilution of OA. The
model also includes a representation of coagulation, but this process has
not been taken into account in the present study. MDMOA has been developed
by adopting and modifying several modules of the CHIMERE chemistry-transport
model (Menut et al., 2013), including the routines implementing the
Gauss–Seidel iteration scheme (Verwer et al., 1994) to solve a set of
dynamic equations, a sectional representation of the OA mass absorption and
evaporation processes (Gelbard and Seinfeld, 1980), and some model
interfaces facilitating modifications of the simulation configuration as
well as providing simulation outputs in a convenient NetCDF format. Dynamic
mass transfer equations for a semi-volatile species, s, in a particle size
section, l, are formulated as follows:
dCsldt=23πdplcλFNplCsg-KCseq+Il-1l+Ill+1,
where Csl is the condensed-phase mass concentration, dpl is
the particle diameter, c and λ are the mean velocity and free path
of the organic molecule in the air, F is the Fuchs–Sutugin correction factor,
Npl is the number of particles in the size bin l, Csg is the
instantaneous gas-phase concentration, K is the Kelvin effect
factor, Cseq is the equilibrium gas-phase concentration, and
Il-1l and Ill+1 are the intersectional fluxes between the
bins l-1 and l and between the bins l and l+1.
The molecular mean free path, the Fuchs–Sutugin correction factor, and the
Kelvin effect factor are evaluated using the conventional formulations
(Seinfeld and Pandis, 2016):
2λ=3Dc-1,3F=1+Kn1+0.3773Kn+1.33Kn1+Knα,4K=exp4σpMWsRTρpdpl,
where D is the molecular diffusion coefficient, Kn is the Knudsen number
(Kn=2λ/dpl), α is the mass accommodation
coefficient (which is assumed to be unity in all our simulations), σp and ρp are the surface tension and density of the
particle material, MWs is the molecular weight, R is the ideal gas
constant, and T is temperature. Following the basic formulations for the VBS
framework, the gas-phase equilibrium concentration is expressed through the
total mass concentration of SVOCs, Ctot, the mass fraction of a given
species, fs, the total mass concentration of OA particles, COA, and
the saturation concentration Cs∗:
Cseq=Cs∗fsCtotCOA1+Cs∗COA-1.
Note that the formulation of the mass flux term for transfer of SVOCs from
and into the gas phase in Eq. (1) is essentially the same as that in the
kinetic model used by May et al. (2013) to derive the volatility
distributions for BB POA. Following May et al. (2013), we also assumed, for
definiteness, that the diffusion coefficient, surface tension, and the
particle bulk density are equal to 5×10-6 m2 s-1,
0.05 N m-1 and 1.2×103 kg m-3, respectively.
The intersectional fluxes are calculated according to Gelbard and Seinfeld (1980) as a combination of the weighed mass fluxes between the gas and
particle phases for the bins l-1, l, and l+1. A concrete representation of
the intersectional fluxes is not of significance in this study, since they
cannot, by definition, contribute to the mass balance (their sum over the
all particle size bins equals zero), and we do not consider here the
evolution of the particle size distribution. Furthermore, as argued below,
the equilibration timescales determined by Eq. (1) are typically much
smaller than the timescales associated with oxidation of SVOCs, and so our
simulations are not sensitive to the particle size distribution.
The dynamics of the total concentration (both in the gas phase and in
particles), Cstot, of a given SVOCs species is driven by the
following mass balance equations:
dCstotdt=-1VPdVPdtCstot-kOHsOHCsg+Ps,
where kOHs is the oxidation reaction rate, [OH] is concentration of
hydroxyl radical, VP is the volume of a BB plume, and Ps is the
chemical production rate of s. The reaction rates and chemical processes
specified in the model are described in the next section (Sect. 2.2). Note
that Eq. (6) for several different species composes an essentially nonlinear
system. In particular, not only does Csg in thermodynamic
equilibrium depend nonlinearly on the total aerosol concentration, COA,
in accordance with Eq. (5), but COA itself also depends in a complex
nonlinear manner on the total concentrations, Cstot, of all SVOCs.
Furthermore, Ps is determined by the gas-phase concentrations of SVOCs,
too, and therefore depends nonlinearly on both Cstot and COA.
Representation of the dilution process (described in Eq. 6 by the term
proportional to dVPdt) in our simulations largely follows
Bian et al. (2017). Specifically, we assume that all the species considered
are uniformly distributed within a box with a half width of 2σy
across the wind direction and a half thickness of 2σz in the
vertical. The thickness of the box in the wind direction does not need to be
explicitly specified in our simulations (but just for definiteness, it can
be assumed to be equal to 1 m). The evolution of σy and σz is represented by the power law expressions according to Klug et al. (1969; see also Seinfeld and Pandis, 2016) for the neutral (D) Pasquill
atmospheric stability class. The plume is assumed to be transported along
the wind direction with a constant speed of 5 m s-1. The initial width
of the plume (4σy) is considered to be a control parameter,
Sp, in our simulations. The initial plume width, Sp, can also be
interpreted as the across-wind width of the area affected by the fire. The
initial value of σz is expressed as a function of σy. It is assumed that the plume's thickness in the vertical direction
(4σz) cannot exceed the mixed layer height, which is fixed at
2500 m: that is, once σz calculated according to Klug et al. (1969) reaches 625 m, the plume is allowed to disperse only in the
horizontal direction. Such a simple representation of the plume's evolution
is by no means intended to be quantitatively accurate under any real
conditions but is used mainly to roughly characterize a dependence of the
temporal scale of the dilution process on the horizontal spatial scale of a
BB plume, especially during the first few hours of evolution.
Representations of BB OA oxidation processes and gas–particle
partitioning in MDMOA
To take into account the existing ambiguity associated with the
representation of the oxidation of organic matter within the VBS framework,
we performed our simulations using several VBS schemes of varying
complexity. A summary of the main features of the schemes and their
reference codes is provided in Table 1. We also used a conventional OA
scheme that assumes that POA is composed of nonvolatile species. Below we
describe the schemes in more detail.
Reference codes and main features of the BB OA modeling
schemes used in the simulations. SVOC: semi-volatile organic compound. VOC:
volatile organic compound. POG: primary organic gas. SOG: secondary organic
gas. POA: primary organic aerosol. SOA: secondary organic aerosol.
C*: saturation mass concentration.
βfrag: fragmentation branching ratio.
Oxidation schemeKey featuresReferencesC17 (a hybrid 1.5-D-VBS scheme)Three sets of SVOCs to model oxidation of organics; five volatility classes (0.1 µg m-3≤C*≤ 103µg m-3) to model gas–particle partitioning; an implicit representation of functionalization and fragmentation reaction pathways; one-bin shift in volatility of a product of oxidation reactions of POG or SOG with OH (kOH=4×10-11 cm3 s-1) with respect to those of a reactantCiarelli et al. (2017a, b)K15 (a 1-D-VBS scheme)An explicit representation of the functionalization and fragmentation branches (βfrag=0.5) of each reaction of POG or SOG with OH (kOH=2×10-11 cm3 s-1); a two-bin shift in volatility and 40 % increase in the molecular weight for a product of the functionalization reaction pathway; seven volatility classes (0.01 µg m-3≤C*≤104µg m-3) for both primary and secondary SVOCsKonovalov et al. (2015)S15 (a quasi-2-D-VBS scheme)An explicit representation of the functionalization and fragmentation branches with distinction between oxidation of POGs and “fresh” SOGs (which undergo only functionalization reactions) and “aged” SOGs (which undergo both functionalization and fragmentation reactions, βfrag=0.85); a one-bin shift in volatility and 15 % increase in the molecular weight for a product of the functionalization pathway of each reaction of POG or SOG with OH (kOH=4×10-11 cm3 s-1); sevena volatility classes (10-2µg m-3≤C*≤104µg m-3) for both primary and secondary SVOCsShrivastava et al. (2013, 2015)T18 (a 2-D-VBS scheme without fragmentation)A representation of SVOCs on a 2-D grid space covering four volatility classes (C*={10-2;100;102;104}µg m-3) and 11 linearly spaced oxygen content bins (O:C={0.2,0.3,…1.2}); the SVOC molecular mass defined as function of C* and O:C; no explicit representation of fragmentation; a one-bin shift in volatility for a product of any oxidation reaction (kOH=2×10-11 cm3 s-1)Tsimpidi et al. (2018)T18f (a modified T18 scheme with fragmentation)The same as the T18 scheme but with fragmentation reactions (βfrag=(O:C)1/4, a uniform probability of fragmentation across the backbone of an organic molecule)Tsimpidi et al. (2018); Donahue et al. (2012b); Murphy et al. (2012)LIN (a “linear” OA scheme)POA and SOA composed of nonvolatile species; SOA formation from oxidation of several specific VOCsPun et al. (2006); Bessagnet et al. (2008); Menut et al. (2013)
a Note that the original VBS scheme (Shrivastava et al., 2015)
involves only five volatility classes.
The scheme “C17” has been described and evaluated by Ciarelli et al. (2017a, b). It is a relatively simple scheme which has been referred to as a
hybrid 1.5-dimensional VBS (1.5-D-VBS) and is based on a similar scheme proposed
by Koo et al. (2014). The idea behind this scheme is to characterize
sources, volatilities, chemical transformation, and oxidation state of a
complex mixture of SVOCs by considering several surrogate species which are
given average molecular compositions and molecular weights. The scheme
distinguishes between three sets of the surrogate species, i.e., the POA set
(set 1), the set containing oxidation products from reactions of hydroxyl
radical with semi-volatile gases from POA (set 2), and the SOA set (set 3)
representing products of reactions of OH with any VOCs (and, implicitly,
also IVOCs). Semi-volatile gases from sets 2 and 3 are also allowed to react
with OH, with the volatility of the product being an order of magnitude
lower than that of the reactant. The different reaction products have
different molecular weights and are assumed to represent the net effects of
the actual functionalization and fragmentation reactions (neither of which
are specified explicitly). All SVOCs are represented using five volatility
classes covering the volatility range from 10-1 to 103µg m-3. The reaction rate (kOH) is fixed at 4.0×10-11 cm3 molec-1 s-1 for all the oxidation reactions.
Some parameters of the scheme have been optimized by fitting box-model
simulations to the data from several aging experiments with BB aerosol from
stove wood combustion (Ciarelli et al., 2017a). Based on the optimization
results, the average ratio of the initial total mass concentrations of VOCs
and SVOCs was set at 4.75. The scheme was then implemented into a
chemistry-transport model and successfully evaluated against ambient
measurements performed with an aerosol mass spectrometer (AMS) across
Europe, specifically in situations where a considerable part of OA
originated from residential wood burning (Ciarelli et al., 2017b). Note,
however, that the composition and conditions of atmospheric aging of OA from
residential wood burning are not necessary representative of those of OA
from vegetation fires.
The scheme “K15” is a one-dimensional (1-D) scheme that was introduced
by Konovalov et al. (2015) in an air pollution case study to represent
atmospheric aging of BB OA from the 2010 Russian fires. It combines one of
the simplest 1-D schemes (Grieshop et al., 2009), in which only
functionalization reactions have been taken into account, and a more complex
scheme in which both functionalization and fragmentation processes are taken
into account (albeit in a very simplified manner) and which is referred
below as the scheme “S15” (Shrivastava et al., 2013, 2015). The oxidation
processes are described using a volatility grid that includes seven bins
(10-2≤C*≤104µg m-3). The scheme
distinguishes between oxidation of primary organic gases (POGs), which is
assumed to result only in functionalization, and oxidation of secondary
organic gases (SOGs), which is assumed to include both functionalization and
fragmentation branches. The products of the functionalization branch get
their mass increased by 40 % and the volatility reduced by 2 orders
magnitude with respect to those of the reactants. Specifically, oxidation of
POG and SOG from volatility bin i is represented as follows:
7POGi>2+OH→1.4SOGi-2,8SOGi>2+OH→0.5×1.4SOGi-2+0.4SOGi=7+0.1LCN,
where LCN denotes the highly volatile low carbon-number species that are the
products of the fragmentation branch, and all the species are assumed to
have the same molecular weight (250 g mol-1). Along with LCN, the
fragmentation branch yields SOG in the highest volatility bin. While LCN
species are not allowed to participate in any reactions, SOG species can be
reprocessed according to Eq. (8). Note that oxidation of SOGs results in a
net increase in the organic mass, although Eq. (8) formally corresponds to a
fragmentation branching ratio (Jimenez et al., 2009) of 0.5. Note also that
the simulations reported by Konovalov et al. (2015) included the
transformation of condensed-phase SOA into nonvolatile SOA (NVSOA) and
indicated that this process had only a small impact on the simulated
evolution of BB aerosol in the case considered. In view of the lack of
robust knowledge about the condensed-phase processes (see also Sect. 4)
and for consistency with the other numerical experiments performed in the
present study, the transformation of SOA into NVSOA has been disregarded in
our simulations.
The scheme “S15” is a slightly modified version of the VBS scheme that was
proposed by Shrivastava et al. (2013) and adopted, as part of a global
chemistry-transport model, for BB aerosol modeling in a subsequent study
(Shrivastava et al., 2015). Unlike the original VBS scheme by Shrivastava et
al. (2013, 2015), where only five volatility classes are used for
computational reasons, the volatility basis set in the S15 scheme is
specified using the same seven volatility bins as in the K15 scheme for the
sake of easier interpretation of differences between the respective
simulation results. The S15 scheme can be regarded as a quasi-two-dimensional
scheme, as it realizes a computationally efficient way to account for the
increasing probability of fragmentation reactions with BB OA aging (and,
implicitly, with increasing oxidation state) of BB OA by distinguishing
between different generations, n, of SOA precursors. Specifically, while POGs
and the first two generations of SOGs are assumed to undergo only
functionalization reactions,
9POGi>1+OH→1.15SOGi-1,n=1,10SOGi>1,n≤2+OH→1.15SOGi-1,n+1,
the third and further generations undergo both functionalization and
fragmentation reactions:
SOGi,n≥3+OH→(1-βfr)×1.15×SOGi-1,n+1+βfr×(0.88×SOGi=7,n+1+0.12×LCN),
where βfr is the fragmentation branching ratio, which in this
case is assumed to be equal to 0.85.
According to Eqs. (9)–(11), the functionalization reactions of both POGs and
SOGs yield SOG species in the next lower-volatility bin and result in an
increase in the molecular weight by 15 %. Similar to the K15 scheme, all
the VBS species are assumed to have the same molar mass of
250 g mol-1. On the basis of the OA oxidation scheme described
above, Shrivastava et al. (2015) defined the two modeling configurations,
FragSVSOA and FragNVSOA, with SOA treated as semi-volatile or nonvolatile,
respectively. The simulations performed in this study with the S15 scheme
involved only the FragSVSOA configuration. This configuration enables better
consistency of the S15 scheme with the other VBS schemes considered here,
and thus any differences between the simulations performed with the S15
scheme and the other schemes are easier to interpret. When choosing the
FragSVSOA configuration, we also took into account that global-model
simulations involving this configuration were found by Shrivastava et al. (2015) to agree better with both surface observations of OA at a South
Africa measurement site positioned to investigate BB aerosol and with
satellite observations of aerosol optical depth on the global scale than the
simulations with the FragNVSOA configuration. Possible formation of NVSOA
due to particle-phase reactions (e.g., Barsanti and Pankow, 2004; Jang et al.,
2002; Shiraiwa et al., 2013b) is one of the factors (see Sect. 4) that can
affect the real BB OA evolution, but these were not analyzed in this study, which focused
on identification of major qualitative nonlinear effects in the BB OA
behavior due to gas-phase oxidation reactions in BB plumes.
Scheme “T18” is a two-dimensional (2-D) VBS scheme that is adopted (with
minor modifications) from Tsimpidi et al. (2018), where it has been
introduced as part of the ORACLE v2.0 aerosol module of the ECHAM/MESSy
Atmospheric Chemistry (EMAC) global model. The scheme represents the
oxidation of BB OA on the 2-D-VBS grid constructed in the space of the
volatility and the oxygen content (O:C ratio). The volatility dimension is
discretized into four bins, C∗={10-2;100;102;104}µg m-3, and the oxygen content dimension is divided into 11 O:C
bins covering the O:C values from 0.2 to 1.2, with a constant step of 0.1.
The smallest value of the O:C ratio is assumed to be representative of fresh
BB emissions and is attributed to POA. Each species is identified with a
representative number of carbon atoms per molecule, nc, and with a
molecular weight, MW, evaluated using structure–activity relationships (Pankow
and Asher, 2008; Donahue et al., 2011) and an approximation of the hydrogen-to-carbon atomic ratio (Heald et al., 2010) as follows:
12nc=11.875-log10C∗0.475+2.3O:C-0.6O:C1+O:C-1,13MW=15O:C+14nc.
Oxidation of POG (or SOG) species is assumed to result in addition of two or
three oxygen atoms to their molecules with an equal probability. The O:C
ratio of the products is therefore evaluated as follows:
Δ(O:C)product=(O:C)reactant+O(nc)reactant,
where ΔO (equal to 2 or 3) is an assumed increment of the oxygen
atomic content. The product is assumed to belong to the next lower-volatility class with respect to the volatility class of the reactant.
Similar oxidation reactions also apply to POG and SOG species from the
lowest-volatility class except that the products of these reactions keep
the volatility of the reactants. Fragmentation reactions are not explicitly
taken into account. As noted by Tsimpidi et al. (2018), neglecting
fragmentation may result in overestimation of OA concentration at long aging
timescales. However, it should also be noted that since nc can decrease as a result of an oxidation step in accordance with
Eq. (12), the fragmentation pathway is, to some extent, taken into account
in the T18 scheme implicitly.
Along with similar 2-D-VBS schemes for anthropogenic and biogenic OA, the T18
scheme described above has been used for multi-year simulations of OA with
the EMAC model, and the simulation results were compared against AMS
measurements at urban downwind and rural environments in the Northern
Hemisphere (Tsimpidi et al., 2018). However, the comparison results did not
provide enough information about the performance of the T18 scheme in
simulations of OA specifically from BB sources.
The last VBS scheme that we used, “T18f”, is our modification of the
original T18 scheme. It has been obtained for this study by adding explicit
fragmentation pathways to the original T18 scheme so that any POG or SOG
species considered in the T18 scheme is assumed to participate in both
functionalization and fragmentation reactions. The reactions originally
included in the T18 scheme are assumed (for simplicity) to represent only
the functionalization pathways. The probability of a given pathway is
controlled by the fragmentation branching ratio, βfrag, which is
parameterized as follows (Jimenez et al., 2009; Donahue et al., 2012b):
βfrag=(O:C)1/4.
Following Murphy et al. (2012), we assume that splitting of an organic
molecule as a result of fragmentation reactions occurs with a uniform
probability at any site throughout its carbon backbone. Accordingly, a
fragmentation reaction of any species containing (according to Eq. 12)
nc carbon atoms can potentially yield (nc-1)/2+1 (if nc is an
odd number) or nc/2 (if nc is an even number) different (with
respect to the atomic carbon content) products. Furthermore, consistently
with the assumptions underlying the original T18 scheme, we assume that as a
result of any oxidation reaction, one of the two fragments receives two or
three additional oxygen atoms, and so its O:C ratio increases in accordance
with Eq. (14) except that the atomic carbon number of the reactant in the
right-hand part of the equation should be substituted for that of the
product. The O:C ratio of the other fragment is kept the same as that of the
reactant. If the calculated O:C ratio of a product exceeds 1.2 (that is, the
maximum value covered by the O:C grid considered), this product is assumed
to be irreversibly lost in the gas phase. All possible pairs of
fragmentation products described above are introduced in the corresponding
mass balance equations (see Eq. 6). The stoichiometric coefficients for the
products in the functionalization, ksfn, and fragmentation,
ksfr, pathways are evaluated as follows:
16ksfn=(1-βfrag)/Npsfn,17ksfr=βfrag/Npsfr,
where Npsfn and Npsfr are the total numbers of possible
products in the functionalization and fragmentation pathways, respectively.
Finally, we also consider a linear analogue to the above nonlinear
representations of BB OA evolution. The scheme “LIN” is based on a simple
oxidation scheme (Pun et al., 2006; Bessagnet et al., 2008) designed in the
framework of the conventional approach to representation of OA evolution and
SOA formation and implemented in the CHIMERE model. The key assumptions
underlying this scheme are that POA is composed of nonvolatile species and
that SOA is formed from oxidation of several specific (“traditional”)
volatile precursors that were identified earlier in smog chamber experiments
(Odum et al., 1997). In this study, the original scheme described in detail
by Menut et al. (2013) was simplified (linearized) by assuming that all
oxidation products are nonvolatile. The emission factors from Andreae (2019)
for the boreal forest were used to specify initial concentrations of POA
precursors as a function of the initial concentration of BB OA. Note that
the original SOA formation scheme from the CHIMERE model was earlier used in
the simulations of evolution of BB aerosol from Russian fires (Konovalov et
al., 2015, 2017, 2018) and was found to produce rather negligible amounts of
SOA in BB plumes.
It should be stressed that the different VBS schemes outlined above
certainly do not comprise all known mechanisms and pathways of oxidation of
organic compounds composing BB OA. Nonetheless, consideration of even a
limited spectrum of available representations of the BB OA aging processes
allows us to gain a useful insight into the uncertainty associated with
simulations of BB OA aerosol evolution using the VBS framework. Some
processes which could further enhance the diversity of our simulations of BB
OA evolution are briefly discussed in Sect. 4.
Configuration of the numerical experiments and processing of output data
MDMOA was run using each VBS scheme described above for a period of 120 h. This period has been chosen to be within the range of typical
atmospheric lifetimes of submicron aerosol particles emitted from open
vegetation fires in the major BB regions worldwide, as indicated, for example, by a
measurement-based estimate (5.1 d) of the lifetime of black carbon (BC)
in Siberia (Paris et al., 2009) and global-model estimates of the BC
lifetimes for open fires in northern Africa (5.6 d) and northern South
America (3.1 d; Wang et al., 2016). A part of the simulation period (7 h d-1) was assumed to correspond to nighttime conditions when any
oxidation processes were disabled (OH concentration in Eq. 6 was set to be
zero); such a nighttime duration is typical, for instance, for central
Siberia in summer.
The initial conditions for SVOCs in particles and in the gas phase
correspond to the gas–particle equilibrium determined in accordance with the
partitioning theory. Specifically, the gas-phase initial concentration for a
species s was calculated using Eq. (5), where the OA mass concentration was
assumed to include, along with the BB fraction, a background OA
concentration, Cbg, of 5 µg m-3. The same background OA
concentration had been specified in the box-model simulations performed by
Bian et al. (2017). For comparison, particulate matter (PM10) in a
boreal environment of central Siberia under background conditions (that is,
without the detectable influence of local or regional pollution sources,
including fires) was found by Mikhailov et al. (2017) to have concentration
ranging from about 2 to 10 µg m-3 in summer, being composed
mostly of organic material. Therefore, the background OA concentration in
our simulations can be assumed to be representative of typical background
conditions in Siberian boreal forest in summer. Note that specifying a much
larger or much smaller value of Cbg would likely result in noticeable
quantitative changes of the simulated BB OA behavior, since entrainment of
background aerosol affects evaporation rates and gas–particle partitioning
in a BB plume (Hodshire et al., 2019). The total mass concentration of
SVOCs, Ctot, that is involved in Eq. (5) was evaluated as follows:
Ctot=C0/∑ifi1+Ci∗C0+Cbg,
where fi defines the mass fraction of all species in bin i of the
volatility distribution, and C0 is the initial BB OA concentration.
C0 is considered – along with the initial plume size, Sp – to be a
control parameter in our simulations. Test experiments (see Sect. 3.1) have
shown that the characteristic timescales for the adjustment to the
“local” thermodynamic equilibrium are short (seconds or minutes) compared
to the timescales associated with the oxidation processes (hours). The
background aerosol concentration was not affected by any process except for
the intersectional fluxes and so was basically kept constant in all the
simulations.
The volatility distribution for all our experiments with the K15 and S15
schemes (in which the volatility grid includes seven bins) was adopted from
the study by Konovalov et al. (2015): f={0.1;0;0.05;0.05;0.2;0.15;0.45} at a temperature of
298 K. This distribution is consistent (within the range of uncertainties)
with the data from thermodenuder measurements of BB emissions (May et al.,
2013). The volatility distributions for the C17, T18, and T18f schemes were
obtained from the same distribution by disregarding or aggregating the
corresponding volatility bins.
In the experiments with the C17, K15, S15, and LIN schemes, the aerosol size
distribution was modeled using nine size bins covering the range from 20 nm to
10 µm and following a geometric progression with the common ratio of
5001/9 (≈2.0). To limit the computational time, the
experiments with the T18 and T18f schemes were conducted using only three size
bins that were defined to cover the same range (from 20 nm to 10 µm)
using a geometric progression with the common ratio of 5001/3 (≈7.9). In all cases, we used a log-normal distribution with a mass mean
diameter of 0.3 µm and a geometric standard deviation of 1.6 (Reid et
al., 2005). While the size distribution can affect the timescales for
evaporation and growth of particles, these timescales, as noted above, are
much smaller than those associated with chemical aging. Accordingly, in all
our simulations, the results of the simulations have been practically
independent of the assumed particle size distribution (except for a very
minor influence of the Kelvin effect). Note that the representation of
nonequilibrium processes in accordance with Eq. (1) has been included in
MDMOA mainly to enable simulations of the evolution of BB aerosol optical
properties in prospective studies.
In accordance with Bian et al. (2017) and Hodshire et al. (2019), we assume
that the OH concentration does not depend on the initial parameters of the
plume and does not change as the plume evolves. Based on the in-plume
measurements by Akagi et al. (2012), its value was set to 5×106 cm-3 in all our simulations, which is larger than the value
(1.08×106 cm-3) specified in the simulations by Bian et
al. (2017) and Hodshire et al. (2019). Assuming a lower or higher constant
value of the OH concentration would slow down or speed up chemical evolution
of BB OA in our simulations but is not expected to alter major qualitative
nonlinear features of the BB OA behavior, which are the focus of this study.
Like Bian et al. (2017) and Hodshire et al. (2019), we assumed a constant
temperature of 298 K within the plume, thus making any
assumptions regarding enthalpies of vaporization of SVOCs unnecessary. The concentration
of OH within real BB plumes can be affected by many factors (such as, for example,
the UV flux, the concentrations of nitrogen oxides, and VOCs within the
plume) which cannot be unambiguously simulated within our box model.
Variability in these factors can cause variability in the OH concentration
levels across different plumes as well as temporal and spatial fluctuations
of the OH concentration within a given plume. In particular, attenuation of
the downwelling UV flux within dense plumes (Hobbs et al., 2003) can
suppress both the OH levels and SOA formation rates (Konovalov et al.,
2016). Temperature is also likely to vary, both spatially and temporally,
within real-world BB plumes. In particular, it is likely to be lower in the
upper part of a plume than near the surface (Hodshire et al., 2019). All
possible variability and inhomogeneities of the OH concentration and
temperature are disregarded in our simulations. Our study is therefore
limited in this respect, but this limitation allowed us to isolate and
investigate the internal dynamics of the BB OA system under fixed
predefined conditions.
Along with aerosol species, MDMOA has been configured to simulate the
evolution of a special tracer which is intended to represent the evolution
of the BB OA mass concentration in a hypothetical situation where BB aerosol
is composed of chemically inert and nonvolatile components. Accordingly,
the tracer has been introduced into our model as a chemically inert species
that can be affected only by dilution (since dry and wet deposition
processes were not considered in our simulations), and the behavior of the
tracer's mass concentration was controlled only by the parameter Sp
(initial plume size). The tracer's initial mass concentration in all our
simulations was the same as that of BB OA (that is, C0). The molecular
weight of the tracer has been set to be the same as that of carbon monoxide (CO).
Note that the concept of analyzing the evolution of BB aerosol versus the
evolution of an inert tracer (usually represented by CO) has been fruitfully
exploited in many previous experimental and modeling studies of BB aerosol
(e.g., Akagi et al., 2012; Konovalov et al., 2015; Bian et al., 2017;
Vakkari et al., 2018). Figure 1 shows the dynamics of the mass concentration
of the inert tracer for several different values of Sp according to our
simulations. Consistent with the tracer dynamics demonstrated by Bian et al. (2017), our simulations indicate that the initial plume size may have a
major impact on the subsequent plume evolution (and thus on the BB OA mass
concentration). In particular, whereas the tracer concentration drops by
4 orders of magnitudes during the evolution of a plume with the smallest
value of Sp (100 m), it keeps a practically constant value (decreasing
by less than 3 %) within the largest plume considered (with Sp of 100 km). Note that for the plumes with Sp smaller than 3 km, the largest
changes of tracer concentration occur during the first 5 h of evolution;
afterwards, the changes are relatively small and are not of significance for
the results of this study. It should be emphasized that our simulations of
the BB plume dispersion were not intended to be fully realistic and
quantitatively accurate. Rather, we used a simple plume dispersion model
(which is formally not applicable at timescales exceeding a few hours and
has many other limitations) just to specify several definite scenarios for
the dilution process, with strongly different dilution rates during the
initial stage of the BB plume evolution.
Simulated evolution of the mass concentration of an inert
tracer for different values of the initial plume size,
Sp. The simulations were done with
an initial tracer concentration of 1000 µg m3.
Using the simulated tracer mass concentration (Tr), we evaluated the BB OA
mass enhancement ratio (abbreviated as EnR throughout this paper and also
denoted as γa below) at a given time t as follows:
γat=COAt-CbgTrt,
where COA(t) is the total OA mass concentration. Note that γa is analogous to the “inert OA mass enhancement ratio” introduced
by Bian et al. (2017). Since the initial concentration of the tracer was set
to be the same as the initial concentration of BB OA, the initial value of
γa in our simulations was always equal to 1 without using any
additional normalization (which is usually involved in similar definitions
of EnR used in chamber and field studies). The simulations were performed
with a sufficiently small nominal time step of 1 s. Based on some test
simulations performed with smaller time steps, we estimate that the
numerical error in γa does not exceed 10 % in any case
considered (but is typically much smaller).
Analysis of γa allows us to identify changes of BB OA
concentration due to the combined effects of oxidation and gas–particle
partitioning processes. Furthermore, any kind of dependence of γa on the initial BB plume size (Sp) or the initial concentration of
BB OA (C0) can be considered to be a manifestation of a nonlinear behavior
of the BB OA mass concentration. Indeed, it is easy to show using Eq. (6)
that if the gas-phase concentration of each SVOC, Csg, were a linear
function of the total concentration, Cstot, of the same species,
then the dynamics of γa would not depend on C0 and could
not be affected by dilution (and accordingly would not depend on Sp
either).
Interaction of SVOCs with water is not taken into account in the OA schemes
described above; thus any known effects of humidity on evolution of BB OA,
such as, for example, formation of SOA from oxidation of water-soluble organic
compounds in the liquid phase (Brege et al., 2018), have been disregarded.
However, as suggested by Tsimpidi et al. (2018), we used the calculations of
the O:C ratio to evaluate the hygroscopicity parameter κ (Petters
and Kreidenweis, 2007), which characterizes hygroscopicity and CCN activity of BB OA. Specifically, we expressed the
hygroscopicity parameter for any organic species s, κs, as a
linear function of the O:C ratio by using a parameterization proposed by
Lambe et al. (2011):
κs=0:18(O:C)+0.03.
The overall value of κ for BB OA, κorg, was obtained
using a mixing rule (Petters and Kreidenweis, 2007):
κorg=∑sεsκs,
where εs is a volume fraction of a given species. The
volume fractions were estimated by assuming a constant volumetric mass
density of 1.2×103 kg m-3 for any OA species
considered. Note that the hygroscopicity parameter was calculated only with
the C17, T18, and T18f schemes, as the other oxidation schemes (K15, S15,
and LIN) considered in this study are not designed to evaluate the O:C
ratio.
ResultsDynamical regimes of the BB OA evolution
The evolution of EnR (γa) according to our simulations
performed with the different OA schemes is presented in Figs. 2 and 3. The
simulations were done with several different values of the initial plume
size, Sp (Fig. 2), and initial BB OA concentration, C0 (Fig. 3). The
parameter range considered in our simulations is intended to represent the
highly variable characteristics of typical smoke plumes from any kind of
vegetation fires. The fixed values of C0 (103µg m3)
and Sp (5 km) in the simulations shown in, respectively, Figs. 2
and 3 are chosen to approximately represent midrange values of the
corresponding parameters (on a logarithmic scale). The simulation results
allow us to identify five distinctive dynamical regimes of the BB OA
evolution. The simulations corresponding to the specific regimes are marked
in the figures by the numbers from 1 to 5. The first regime (1)
corresponds to a monotonic saturating increase in γa. This
regime is found in the simulations with the C17, K15, and LIN schemes. An
increase in γa is followed by its decrease in the second regime
(2), which is typical for the S15, T18, and T18f schemes. The third
regime (3) features a sharp initial decrease in γa
followed by a slow monotonic increase. This regime is found with the C17,
K15, and T18 schemes but only when Sp is relatively small (in
particular, when Sp equals to 100 m; see Fig. 2c and d). The most complex
behavior of γa corresponds to the fourth regime (4) and is
found with the S15, T18, and T18f schemes (Figs. 2c, 2d, and 3d) for specific
C0 and Sp values. In this regime, γa first decreases,
then increases, and finally decreases again. Finally, the fifth regime
(5) corresponds to the monotonic decrease in γa. It is
found only in a simulation with the T18f scheme (Fig. 3d).
Evolution of the BB OA mass enhancement ratio (γa) according to the simulations performed with the
different OA schemes and with different values of the initial plume size:
(a)Sp=100 km, (b)Sp=3 km, and (c, d)Sp=0.1 km. The
initial BB OA concentration, C0,
was fixed at 103µg m3
in all the simulations. Panel (d) shows a zoomed fragment of the simulations
presented in panel (c). Shaded bands depict nighttime periods when no
oxidation reactions were allowed to occur. The numbers on the curves denote
the different dynamical regimes of BB OA evolution according to the
definitions in Sect. 3.1. A horizontal dashed–dotted line in each panel
indicates the situation where no BB OA mass enhancement occurs (γa=1).
Evolution of the BB OA mass enhancement ratio (γa) according to the simulations performed with the
different OA schemes and with different values of the initial mass
concentration of BB OA: (a)C0=102µg m3, (b)C0=5×102µg m3, (c)C0=103µg m3, and (d)C0=104µg m3. The initial BB plume size,
Sp, was fixed at 5 km in all the
simulations shown. Shaded bands depict nighttime periods when no oxidation
reactions were allowed to occur. The numbers on the curves denote the
different dynamical regimes of BB OA evolution according to the definitions
in Sect. 3.1. A horizontal dashed–dotted line in each panel indicates the
situation where no BB OA mass enhancement occurs (γa=1).
The results shown in Figs. 2 and 3 demonstrate, on the one hand, that the
differences in the considered representations of the OA evolution are
associated with both major quantitative and qualitative differences in the
BB OA behavior even when the control parameters are the same. For example,
if Sp=5 km and C0=103µg m3 (Fig. 3c),
the simulation with the K15 scheme predicts a strong monotonic increase in
γa up to a factor of 3 after a 120 h evolution. At the same
time, the T18f scheme predicts a slight net decrease in γa
even though it also predicts an increase in γa at the initial
stage of the evolution. Even larger differences (exceeding a factor of 8)
between the simulations are evident if Sp equals 0.1 km.
On the other hand, changing parameter values in the simulations with the
same VBS scheme can result in “switching” between different types of the
BB OA evolution. In particular, depending on Sp, the simulations with
the C17 and K15 schemes can demonstrate both the regimes 1 (Figs. 2a, 2b,
and 3) and 3 (Fig. 2c and d). Both the regimes 2 (e.g., Fig. 2a and b) and
4 (e.g., Fig. 2c) are found in the simulations performed with the S15
and T18f schemes. The simulations with the T18 scheme are found to manifest
four regimes, from 1 to 4 (Figs. 3a, 3c, 3d, and 2c). In contrast to
the simulations using the VBS framework, the simulations with the LIN scheme
manifest a single dynamic regime (regime 1). Note that the simulation
results shown in Figs. 2 and 3 are not meant to provide an exhaustive
analysis of the parameter space with regard to possible dynamical regimes,
but are rather intended to illustrate the diversity of the BB OA behavior
simulated with different OA schemes and different parameter values.
Figures 2 and 3 also show that using the same OA scheme with different
values of Sp and C0 can be associated with not only
qualitative differences in the EnR behavior (as indicated above) but also
with considerable quantitative differences between the simulated EnR values.
For example, γa for the BB OA at an age of 120 h with the
K15 scheme and C0=1000µg m3 increases from about 2.7 to 4.2 as Sp decreases from 100 to 0.1 km (cf. Fig. 2a and c). In
contrast, the same change of Sp in the simulation with the S15 scheme
results in a dramatic drop in γa, from about 1.5 to 0.5. The
simulations with the C17 and T18 are relatively insensitive to changes of
Sp. An increase in C0 is typically associated with a decrease in
γa. The sensitivity of γa to C0 is very
significant in the simulations with the K15 scheme but is weak in the
simulations with the S15 and T18f schemes.
As noted in Sect. 2.3, any effects of changes in the parameters
Sp and C0 on γa can be considered to be a
manifestation of a nonlinear behavior of BB OA. In contrast to the nonlinear
behavior demonstrated by BB OA in the simulations with the VBS schemes, the
simulations with the LIN scheme manifest a simple “linear” behavior, being
quite insensitive to changes of Sp and C0: γa
monotonically increases by ∼5 % after 120 h irrespective
of the parameter values. This insensitivity is a result of the assumptions
that POA is not affected by any processes except for dilution and that the
initial concentration of SOA precursors is proportional to C0. The
dependencies of γa on the control parameters of our
simulations are analyzed in more detail in the next section (Sect. 3.2).
The existence of the different dynamical regimes of BB OA evolution can be
interpreted by considering the interplay between several competing
processes, such as functionalization, fragmentation, and evaporation with
dilution. In particular, regime 1, featuring a saturating enhancement of
γa, can be explained by SOA formation dominated by
functionalization reactions associated with a decrease in volatility and an
increase in the molecular weight of SVOCs. A steady state is reached in this
regime when all SVOCs belong to the two lowest-volatility bins. Rapid
fragmentation reactions in the case of S15 and T18f schemes or the decrease in the molecular weight of the products of the oxidation reactions
(effectively combining the effects of the functionalization and
fragmentation oxidation pathways) in the case of the T18 scheme can cause a
depletion of the SOA amounts that have initially been formed from oxidation
of POGs, leading to regime 2. A rapid POA evaporation caused by dilution
at the initial stage of the BB plume evolution can result in a depletion of
the bulk OA amounts and give rise to the regimes 3, 4, or 5,
depending on the interplay between fragmentation and functionalization after
the fast dilution.
Unlike the BB OA enhancement ratio, the hygroscopicity parameter κorg exhibits a rather simple behavior (Fig. 4), similar to that of
γa in the regime 1. The monotonic growth of κorg
over time is hardly surprising, since each reaction is assumed to result in
an increase in the O:C ratio (according to, for example, Eq. 14) for at least one
of the products with respect to that of the reactant (and the O:C ratio of
the other products is assumed to remain the same as that of the reactant).
The initial rapid increase in κorg eventually slows down
because the semi-volatile POGs and SOGs participating in the oxidation
processes are eventually transformed either into low-volatile products (as
in the simulations with the C17 and T18 schemes) or, on the contrary, into
volatile gases (as in the simulations with the T18f scheme). Note again that
the hygroscopicity parameter was calculated using only the C17, T18, and T18f
schemes, since the other oxidation schemes (K15, S15, and LIN) considered in
this study are not designed to track changes of the O:C ratio.
Evolution of the hygroscopicity parameter κorg simulated as a function of time with different
values of the initial BB OA mass concentration, C0, and the initial BB plume size, Sp, using the (a) C17, (b) T18, and
(c) T18f schemes. Shaded bands depict nighttime periods when no oxidation
reactions were allowed to occur.
Nonetheless, in spite of a relatively simple behavior, the simulations of
κorg are still rather sensitive to the parameters of the BB
plume. In particular, the value of κorg after the 120 h
evolution is found, according to the simulation with the T18 scheme, to be
about a factor of 2 bigger in the cases with a small plume (Sp=0.1 km), irrespective of the initial aerosol concentration, than in the cases
with a large and dense plume (Sp=100 km; C0=104µg m3). Values of κorg are higher for smaller plumes
because a larger fraction of POA species evaporates due to more rapid
dilution of smaller plumes, followed by gas-phase formation of highly
oxidized SOA species. The sensitivity of the simulations of κorg with the T18f scheme to the control parameter values is smaller
but is still considerable. The fact that κorg is sensitive to
the parameters of a BB plume can be regarded as another manifestation of the
nonlinear behavior of BB OA. Note that only simulations for the “extreme”
values of C0 and Sp (among those considered in this study) are shown
in Fig. 4 for all of the three oxidation schemes allowing an evaluation of
the hygroscopicity parameter. Simulations performed using a given scheme
with other (intermediate) parameter values would fall between the brown and
blue curves in a corresponding plot. The dependencies of κorg
on C0 and Sp are analyzed in some more detail in the next section.
Note also that quantitatively, the hygroscopicity parameter values ranging
from 0.08 to 0.18 for aged aerosol in our simulations look rather
reasonable. For example, Mikhailov et al. (2015) reported a volume-based
hygroscopicity parameter value of about 0.1 (on average) for the
accumulation mode of aerosol in Siberia during the period of intense fires.
As noted above (Sect. 2.3), the initial conditions for our BB OA simulations
correspond to instantaneous thermodynamic equilibrium. Additional test
simulations performed using the S15 scheme with perturbed initial conditions
and two different numbers of size bins (nine, as in the simulations with the
C17, K15, and S15 schemes, or three, as in the simulations with the T18 and
T18f schemes) confirm (Fig. 5) that thermodynamic equilibrium between POGs
and particles is typically reached before any noticeable changes occur due
to chemical processes. The equilibration timescales are found to range from
just a few seconds (when the balance between gas and particle phases is
shifted toward the particle phase) to about 10 min (when all the BB
emissions are assumed to be in the gas phase and the bin number equals
3). In accordance with these results, the BB OA mass concentration after
the 5 d evolution was not found to be sensitive in our additional test
simulations (which are not presented here) to either the initial
perturbations of the gas–particle equilibrium or to the change in the number
of bins in the particle size distribution in the range of the considered
values of the plume parameters even though when OA is aged or strongly diluted, it can feature much longer equilibration timescales (Riipinen et al.,
2010).
Adjustment of mass concentration of fresh BB OA to
thermodynamic equilibrium according to test simulations performed with the
S15 scheme without dilution. The simulations were initiated from different
initial conditions: (a)C0=0µg m3 and (b)C0=2×102µg m3. A horizontal
dashed–dotted line in each panel indicates the BB OA concentration
(103µg m3) at
thermodynamic equilibrium. The time step in the test simulations was set at
0.1 s.
Dependencies of the BB OA enhancement ratio and hygroscopicity parameter
on the parameters of the smoke plume
Figures 6 and 7 present the simulated dependencies of EnR on the initial BB
OA concentration, C0, after 5 and 120 h of evolution, γa(5 h) and γa(120 h), respectively. These dependencies have
been calculated using the different OA schemes with varying values of the
initial plume size, Sp. Figure 6 also shows (by dashed lines) the
initial ratio of the total SVOCs mass concentration (Ctot) to the total
BB OA mass concentration (C0) at thermodynamic equilibrium according to
Eq. (18). The Ctot/C0 ratio depends on the parameters of the
assumed volatility distribution and, for this reason, is different for
different schemes. Our simulations show that γa(5 h) follows, in
most cases, a slow and monotonic inverse dependence on C0. This
dependence is qualitatively (and in some cases, but not always, even
quantitatively) rather similar to that of Ctot/C0: evidently, the
decreasing character of the dependence of γa on C0 can be
due to the fact that larger values of C0 correspond to a smaller
SVOC gas-phase fraction available for SOA formation. Larger values of
Sp typically correspond to larger values of
γa(5 h); such a dependence is evidently due to evaporation of POA
with dilution. However, the dependences of γa(5 h) on C0 and Sp are not always monotonic. In particular, the dependence of
γa on C0 has a “weak” minimum in the simulations with the
T18 scheme and Sp of 1 km when C0 is around 5×103µg m3. A similar minimum (but around 2×103µg m3) can be seen in the simulations with the T18f scheme. When C0
equals 5×103 or 104µg m3 in the simulations
with the T18 scheme, γa calculated with Sp=5 km is
larger than that calculated with Sp=100 km, although an inverse
relation exists between the corresponding values of γa
simulated with Sp=0.1 km and Sp=0.5 km. These
irregularities cannot be easily explained but are certainly not due to
numerical errors (as has been confirmed in additional test simulations with
a reduced time step), emphasizing the essentially nonlinear character of the
BB OA behavior, especially when it is modeled using such complex VBS schemes
as T18 or T18f.
Simulated dependencies of the BB OA enhancement ratio
after 5 h of evolution on the initial BB OA mass concentration,
C0. The dependencies are obtained
with different OA schemes (Table 1) and with different values of the initial
plume size, Sp. Dashed lines show
the ratio of the total initial mass concentration of SVOCs
(Ctot) to the initial BB OA
concentration (C0) at thermodynamic
equilibrium according to Eq. (18). A horizontal dashed–dotted line in each
panel indicates the situation where no BB OA mass enhancement occurs
(γa=1).
The same as in Fig. 6 but for the BB OA enhancement ratio
after 120 h of evolution and except that the ratio of
Ctot to
C0 (which is not supposed to
characterize the state of BB OA after 5 d aging) is not shown. Note
that the y axis scale in the panel (b) is not the same as in other panels. A
horizontal dashed–dotted line in each panel indicates the situation where no
BB OA mass enhancement occurs (γa=1).
The value of γa(120 h) also tends to decrease as C0
increases in most cases, but with differences among the schemes: C17, K15,
and T18 schemes with a weaker degree of fragmentation show a pronounced
decrease, while schemes with a stronger degree of fragmentation are less
sensitive to C0 (Fig. 7). This is presumably due to the fact that if
functionalization is more important than fragmentation, then the
availability of more SOGs at lower C0 leads to enhancements in SOA
formation.
However, different schemes feature more complex and diverse dependencies on
Sp (Fig. 7). For example, while the simulations with the C17 and T18
schemes, where the fragmentation and functionalization processes are treated
implicitly, do not reveal any strong sensitivity of γa(120 h) on
Sp, the scheme K15 where functionalization strongly dominates over
fragmentation yields a rather strong but inverse dependence of the same
parameter (that is, unlike γa(5 h) calculated with the same
scheme, γa(120 h) decreases with increasing Sp). Presumably,
evaporation of POA due to dilution provides more “fuel” for
functionalization reactions, leading to a larger increase in γa
when Sp is smaller (as found in the simulations with K15 scheme). In
contrast, the schemes with strong fragmentation reactions, S15 and T18f,
reveal a strong increasing dependence of γa(120 h) on Sp,
while they are less sensitive on C0 relative to the other schemes.
This dependence can mainly be due to the fact that the evaporated organic
compounds are exposed to fragmentation reactions: accordingly, a decrease in
Sp results in larger losses of SVOCs and smaller SOA concentrations.
A comparison of the results shown in Figs. 6 and 7 reveals that the
differences between the different VBS schemes are associated with mostly
only quantitative differences in the dependencies of γa(5 h) on
C0 and Sp, but with both quantitative and some qualitative
differences in the corresponding dependencies of γa(120 h). This
observation indicates that the mass concentration of aged BB OA is likely to
be much more strongly affected in the simulations by uncertainties in
available representations of the BB OA evolution than the mass concentration
of relatively fresh BB OA. One of the reasons is that as the amount of SOGs
susceptible to fragmentation reactions increases with time, the competition
between functionalization and fragmentation reactions plays an increasingly
important role, creating a more complex dependence on the plume parameters.
As a result of this competition, the outcome of the BB OA evolution becomes
strongly dependent on the fragmentation branching ratio associated with a
given OA scheme. One of the most significant findings of our analysis in
this respect is that the sensitivity of γa to the initial plume
size (or, similarly, to the dilution rate) may be positive or negative,
depending on the VBS scheme and the BB OA age. Factors controlling the
dependence of γa on Sp are examined more in detail in the
next section (Sect. 3.3).
The simulated dependencies of the values of the hygroscopicity parameter
after 5 and 120 h of evolution, κorg(5 h) and κorg(120 h), respectively, on the initial BB OA concentration,
C0, and the plume size, Sp, are presented in Fig. 8. Evidently,
similar to γa(5 h) and γa(120 h), κorg(5 h) and κorg(120 h) are affected by changes of the
control parameters in a nonlinear way, although the nonlinearities are not
as pronounced as in the case of the corresponding dependencies for EnR. In
particular, while the values of κorg(5 h) and κorg(120 h) decrease with increasing C0 for the largest plumes
(Sp=100 km), they are nearly constant and even slightly increase
with increasing C0 for the smallest plumes (Sp=0.1 km).
According to the simulations with all three of the schemes enabling
simulations of the hygroscopicity parameter, both κorg(5 h) and
κorg(120 h) are more sensitive to the changes of Sp than to
the changes of C0. Furthermore, the sensitivity of κorg(5 h)
and κorg(120 h) to Sp increases with increasing C0. Note
that as C0 increases, a larger fraction of the total particulate organic matter concentration,
Ctot, belongs, under thermodynamic equilibrium, to the condensed phase.
This fraction is oxidized (as a result of mass transfer between the
particles and gas phase) slowly compared to the gas-phase fraction, which is
why both κorg(5 h) and κorg(120 h) mostly tend to
decrease with C0. However, the particulate fraction of Ctot can
readily evaporate and become accessible for fast gas-phase oxidation as a
result of evaporation in the rapidly diluted small plumes: this may explain
why κorg(5 h) and κorg(120 h) are nearly insensitive
to C0 in the simulations with Sp=0.1 km.
Dependencies of the hygroscopicity parameter, κorg, after (a, c, e) 5 h and (b, d, f) 120 h of BB OA
evolution, simulated using the (a, b) C17, (c, d) T18, and (e, f) T18f
schemes.
The simulations of κorg performed with the different schemes
are noticeably different both quantitatively and qualitatively.
Quantitatively, the C17 and T18 schemes typically yield larger values of
κorg than the T18f scheme, and, for the smaller plumes (with
Sp≤1 km), the T18 scheme predicts bigger values of κorg(120 h) than the C17 scheme. Qualitatively, while aerosol aging is
associated with increasing sensitivity of κorg to Sp in the
simulations with the T18 scheme (cf. Fig. 8c and d), the sensitivity of
κorg either does not change significantly or diminishes
(depending on C0) as the aerosol ages in the simulations with the C17
(cf. Fig. 8a and b) and T18f schemes (cf. Fig. 8e and f). The likely
reasons for the differences between the simulations of κorg
with, on the one hand, the C17 and T18 schemes, and, on the other hand, the
T18f scheme, have already been mentioned above (see Sect. 3.1). That is, on
the one hand, oxidation reactions in the C17 and T18 schemes yield less
volatile products, which eventually accumulate in particles and increase the
overall O:C ratio for BB OA. On the other hand, fragmentation reactions in
the T18f scheme result in irreversible loss of the condensable organic
matter into the gas phase, thus limiting an increase in the O:C ratio for
the condensed phase. The significant differences between the κorg(120 h) values simulated with the C17 and T18 schemes for the
smaller plumes (with Sp≤1 km) are likely due to the fact that
semi-volatile products from oxidation of OA can achieve a higher degree of
oxidation in the T18 scheme than in the C17 scheme.
The effects of fragmentation reactions
As noted in the previous section, our simulations performed with the S15 (or
T18f) and K15 schemes reveal qualitatively different dependencies of the BB
OA enhancement ratio after 120 h of evolution, γa(120 h), on
the initial concentration, C0, but even more on the initial plume
size, Sp. Specifically, while the S15 and T18f schemes predict that
γa(120 h) increases with an increase in Sp, the K15 scheme
yields an inverse dependence. Meanwhile, the main difference between the S15
and K15 schemes is that the fragmentation pathway of the oxidation reactions
has a larger weight relative to the functionalization pathway in the S15
scheme than in the K15 scheme. The effect of fragmentation reactions is also
quite visible in simulations with the T18f scheme compared to the simulations
with the T18 scheme. Taking these considerations into account, we performed
additional simulations in which we varied the fragmentation branching ratio
in the same scheme. For definiteness, we used the S15 scheme which is rather
“transparent” and flexible but not oversimplified.
Figure 9 shows the dependencies of γa(120 h) on both
C0 and Sp according to our simulations performed with
four different values of the fragmentation branching ratio, βfrag, for the SOA oxidation reactions (see Eq. 11): 0, 0.25, 0.5, and
1.0. Figure 7c shows similar calculations for the case with βfrag=0.85. The simulations with the extreme values of βfrag (0 and 1) do not correspond to any realistic situation and are
provided only for reference purposes, but taking into account that the
fragmentation branching ratio is a highly uncertain parameter of VBS schemes
(Donahue et al., 2012b; Shrivastava et al., 2015), neither of the
simulations with the other values of βfrag is intended to be
more or less realistic. It is noticeable that the dependencies obtained
using the S15 scheme with βfrag=0 (see Fig. 9a) are very
similar to those obtained with the K15 scheme (Fig. 7b). Specifically,
according to the simulation with both schemes, γa(120 h) is
inversely dependent on Sp. An increase in βfrag (Fig. 9b) to
0.25 results in a stronger sensitivity of γa(120 h) to Sp
and in an increase in maximum values of γa(120 h) (reached with
Sp=0.1 km): apparently, this is due to the fact that the effect of
fragmentation of SOGs is counteracted by oxidation of their products which
are reprocessed according to Eq. (11). However, as βfrag increases further, the reprocessing of SOGs cannot compensate for a loss
of SVOCs in the fragmentation branch: the values of γa(120 h)
calculated with βfrag=0.5 (Fig. 9c) are typically lower (and
sometimes very considerably) than those calculated with βfrag=0.25. Furthermore, when βfrag equals 0.5, the dependence of
γa(120 h) on Sp “collapses”. That is, an approximate
balance between the fragmentation and functionalization branches makes the
BB OA enhancement ratio after 5 d of evolution nearly insensitive to
dilution. Note that there is an obvious similarity between the dependences
obtained using, on the one hand, the S15 scheme with the branching ratio of
0.5 (Fig. 9c), and, on the other hand, the C17 and T18 schemes. This
similarity implies that the fragmentation and functionalization processes
are, effectively, nearly balanced in the C17 and T18 schemes. The dominance
of fragmentation over functionalization is again associated with a rather
strong sensitivity of γa(120 h) to Sp (Figs. 7c and 9d).
However, unlike the cases with βfrag<0.5, γa(120 h) increases with increasing Sp. Another noteworthy feature
of the simulations with βfrag=0.85 or βfrag=1.0 is a rather weak sensitivity of γa(120 h) to C0. This
is likely because the oxidation of POGs, which are available in larger
amounts when C0 is smaller, yields much less SOA in the cases with
dominating fragmentation compared to the cases with dominating
functionalization.
Dependencies of the BB OA enhancement ratio after 120 h of evolution on the initial BB OA mass concentration,
C0, according to the simulations
using the S15 scheme with different values of
Sp and varying values of the
fragmentation branching ratio: (a)βfrag=0, (b)βfrag=0.25,
(c)βfrag=0.5, and (d)βfrag=1.0. A horizontal dashed–dotted line
in each panel indicates the situation where no BB OA mass enhancement occurs
(γa=1).
The effects of fragmentation and functionalization reactions on the
composition of BB OA in the simulation performed with the S15 scheme on the
composition of BB OA are further visualized in Fig. 10, which illustrates
the BB OA evolution by distinguishing primary and secondary fractions in
particles (POA and SOA) and in the gas phase (POG and SOG) as well as
between “fresh” (n≤2, see Sect. 2.2) and “aged” (n>2) fractions of SOA
(SOA-f and SOA-a, respectively). The simulations shown in Fig. 10 were
performed for the four cases with different values of C0 (102 and 103µg m-3) and βfrag (0 and
1). The initial plume size was taken to be 100 km in all these simulations;
that is, the BB OA evolution was practically not affected by dilution.
Evolution of the mass concentration of BB OA and its
components, including POA, “fresh” SOA (SOA-f), and “aged” SOA (SOA-a),
according to the simulations performed with the S15 scheme. Also shown are
the total mass concentration of SOA and the mass concentrations of POGs and
SOGs. The simulations were done with two different values of βfrag, (a, b)βfrag=0 and (c, d)βfrag=1, and two different values of
C0, (a, c)C0=102µg m3 and
(b, d)C0=103µg m3. The initial plume size was set at 100 km in all
the simulations. A horizontal dashed–dotted line in each panel indicates the
initial concentration of BB OA aerosol. Shaded bands depict nighttime
periods when no oxidation reactions were allowed to occur.
It is evident that after about 5–10 h of evolution, SOA already provides
a major contribution to the OA mass concentration in all of the cases
considered. Not surprisingly, the SOA fraction is larger in the simulations
with βfrag=0 (in the “functionalization” case) than in
the simulations with the βfrag=1 (in the “fragmentation”
case). In addition, the aged SOA fraction is larger for the
non-fragmentation case. Consistent with the dependence of γa on C0 in Fig. 9a, the SOA fraction calculated with βfrag=0 is also larger in the simulations with a smaller value of
C0 (cf. Fig. 10a and b). A smaller C0 value is also associated
with a larger contribution of SOA-a (and, correspondingly, with a smaller
contribution of SOA-f) to the total concentration of SOA. It is worth noting
that compared with the simulations performed with C0=103µg m-3, the simulations with C0=102µg m-3 feature a faster decrease in SOA-f. This is probably due to an
initially larger fraction of SOGs in the simulations with C0=102µg m-3: rapid oxidation of fresh SOGs causes
evaporation of corresponding species from particles, thus depleting SOA-f. A
major qualitative difference between the simulations performed for the
functionalization and fragmentation cases is that the OA, SOA, and
SOA-a concentrations monotonically increase with time in the former case
(regime 1) but show a “humped” dependence on time in the latter case
(regime 2). That is, the simulations shown in Fig. 10 confirm our initial
suggestion (see Sect. 3.1) that realization of regimes 1 and 2
depends on the ratio between the fragmentation and functionalization
branches of the oxidation reactions.
Discussion
In this section, we briefly discuss possible implications and limitations of
our findings presented in Sect. 3. First of all, our results indicate that
uncertainties associated with the representation of BB OA evolution in
chemistry-transport models (CTMs) may have a major impact on the simulated
behavior of BB aerosol. Indeed, our simulations show that differences
between available OA schemes may entail differences in the calculated EnR
(γa) as large as a factor of 5 (Fig. 3a) or even greater
(Fig. 2c) after 5 d of evolution. Furthermore, there may be not only
quantitative but also qualitative differences in the BB OA simulations, such
as a difference between a gradually increasing EnR (in the regime 1) and
a humped temporal dependence of EnR (in the regime 2; see, for example,
Fig. 2a). Typically, OA VBS schemes are constrained (at least to some
extent) with data from aerosol chamber experiments (e.g., Hennigan et al.,
2011) which are, however, rarely representative of more than a few hours of
BB aerosol aging under atmospheric conditions. Taking into account that
major quantitative and qualitative differences between our simulations with
different OA schemes are manifested at longer timescales, our results
therefore indicate the particular importance of observational constraints at
the timescales ranging from tens of hours to several days. Potentially,
such constraints can be provided by in situ aerosol measurements (e.g.,
Jolleys et al., 2012; Shrivastava et al., 2015) or remote (e.g., satellite)
observations (Konovalov et al., 2017).
Even if one of the VBS schemes considered in this study were perfectly
adequate, its application in the framework of a typical CTM would still
likely entail considerable model biases due to several factors. First, the
bias may be due to a dependence of BB OA evolution on its initial
concentration. Indeed, smearing of BB emissions from an intense but small
(relative to the grid cell size) fire within a model grid cell would result
in smaller initial concentration (C0) of BB OA in the model than in the
real atmosphere. For example, if the model had a horizontal resolution of
50 km×50 km but the fire size was 5 km×5 km, the
initial concentration of BB OA in the model would be 2 orders of magnitude
lower than within the actual plume (e.g., 102 instead of 104µg m-3). As indicated by our analysis for several cases (Figs. 6a, 6b, and 6d and 7a, 7b, and 7d), the EnR simulated for such a situation can be
strongly overestimated due to its inverse dependence on C0. Second, a
model bias in EnR can be caused by a dependence of EnR on the initial plume
size, Sp. For example, due to the effect of rapid dilution of BB OA
plumes originating from relatively small (with a typical size of about 1 km
or less) but multiple fires, a model that does not resolve the plume scales
will likely overestimate EnR and corresponding changes of the BB OA mass
concentration after its aging during several days if the evolution of BB OA
is strongly affected by fragmentation (Fig. 7c and e). However, it is also
possible that a CTM will underestimate EnR if functionalization dominates
over fragmentation (Fig. 7b). If the BB OA evolution is reproduced
adequately with the C17 and T18 schemes, then the same situation (where BB
OA plumes originate from relatively small fires) may be associated with a
negative bias in the hygroscopicity parameter κorg (Fig. 8b and d)
rather than with any significant bias in EnR (Fig. 7a and d). Over shorter timescales (a few hours), a CTM is likely to overestimate EnR and underestimate
κorg in the case of small fires, irrespective of the relative
importance of functionalization and fragmentation reactions (Figs. 6a–e and
8a, c, and e). Note that our findings concerning the impact of the plume size on
EnR after a few initial hours of aging are qualitatively consistent with the
results of numerical experiments conducted by Bian et al. (2017) and
Hodshire et al. (2019). The net effects of the limited horizontal resolution
of a CTM on the simulated BB OA evolution are likely very variable and are
hardly possible to predict accurately with a box model, as they should be
dependent on the spatial structure and intensity of fire emissions as well
as on meteorological conditions. Third, the simulated values of EnR (and to
a lesser extent, those of the hygroscopicity parameter) may be biased as a
result of model errors in the vertical injection profile of fire emissions.
Indeed, available parameterizations can strongly underestimate or
overestimate the smoke injection height, sometimes by an order of magnitude
(Sofiev et al., 2012); these errors can cause biases in EnR as a result of
the sensitivity of the simulations of BB OA to uncertainties in both the
plume's density and dilution rate and also to its temperature, affecting
thermodynamic equilibria. Similar effects of chemical nonlinearities on
simulations of the consequences of atmospheric processing of BB emissions
were first pointed out by Chatfield and Delany (1990).
The results of our analysis may be helpful for understanding the reasons for
the observational diversity of BB OA aging effects. Indeed, our findings
suggest that the occurrence of differences in the observed effects of BB OA
aging, such as an increase (see, for example, Yokelson et al., 2009; Konovalov et
al., 2015; Vakkari et al., 2018) or a decrease (e.g., Jolleys et al., 2012;
2015) in EnR can be driven by differences in the initial parameters of the
BB plumes as well as by the BB aerosol's photochemical age and the ratio
between the fragmentation and functionalization reaction pathways. In
particular, our simulations predict that aging of BB aerosol in a large
plume is more likely to result initially in an increase in EnR, whereas the
initial evolution of a small plume is likely to be associated with an
initial decrease in EnR (Fig. 2a and c). Consistent with these
predictions, there is observational evidence for a growth of EnR in
the situations with large-scale fires (e.g., Konovalov et al., 2015, 2017;
Vakkari et al., 2018) as well as evidence for an initial drop in γa in a relatively small isolated plume (Akagi et al., 2012). It is
noteworthy that in agreement with our simulations corresponding to the
regimes 3 and 4, the observations by Akagi et al. (2012) show an
increase in EnR after the initial decrease. Our analysis also indicates that
if fragmentation dominates over functionalization, an initial growth of EnR
in situations with large fires is likely to be followed by a decreasing
stage (see the simulations for regime 2 in Figs. 2a, 2b, and 3). A similar
humped dependence of EnR on the BB aerosol photochemical age was
identified earlier in the analysis of satellite observations of BB aerosol
in Siberia (Konovalov et al., 2017). The fact that, according to aircraft
observations, EnR can be around or smaller than unity even in relatively
large plumes transported during a few days (Jolleys et al., 2012, 2015) is
consistent with our simulations corresponding to regimes 2, 4, or
5 in Fig. 3.
Our analysis indicates that BB OA evolution is strongly dependent on the
fragmentation branching ratio (Fig. 9), which is known to be a growing
function of the oxidation state of OA (Jimenez et al., 2009). It seems
reasonable to suggest that the oxidation state of BB OA (and thus the
effective fragmentation branching ratio) within a BB plume at any given
moment is dependent on the initial chemical composition (and therefore
oxidation state) of fresh BB OA. There is evidence (e.g., Grieshop et al.,
2009; May et al., 2015; Tiita et al., 2016; Ahern et al., 2019; Lim et al.,
2019) that the oxidation state of fresh BB OA is strongly variable,
depending on conditions of burning and fuel type. Variable effects of
fragmentation reactions have been suggested to be a possible reason for the
diversity of values of the BB OA EnR (ranging from about 0.8 to almost 3) in
BB aerosol aging experiments (Hennigan et al., 2011). Therefore, the
differences between our simulations performed with different oxidation
schemes may reflect, to some extent, the diversity of the observed aging
effects of BB OA due to variability in its initial chemical composition.
It should be especially noted that the VBS schemes employed in our
simulations do not comprise the whole range of uncertainties associated with
the representation of the BB OA aging in models. In particular, all of the
considered schemes share the common assumption that OA particles are
composed of well-mixed liquids, which is also used in the majority of other
modeling studies using the VBS framework. However, based on findings from
some laboratory experiments, Shrivastava et al. (2013, 2015) argued that SOA
should rather be represented as nonvolatile glassy semisolid mass.
Available estimates of the annual-mean SOA glass transition temperature
(Shiraiwa et al., 2017) indicate that in major biomass burning regions (such
as the Amazon basin and Siberia), SOA is typically in the liquid state near
the ground but transits to the semi-glassy state towards the top of the
boundary layer. Taking these estimates into account, it seems reasonable to
expect that the SOA phase state in ambient BB aerosol in these regions is
quite variable, depending on relative humidity and ambient temperature. When
the equilibrium state of SOA is semisolid, the effects of gas-phase
fragmentation reactions are expected to diminish (since SOA, once formed,
cannot evaporate), depending on the highly uncertain timescale of
transformation of semi-volatile SOA into a nonvolatile state (Shrivastava et
al., 2015). Another uncertain factor that affects the evolution of the
ambient BB aerosol but is not taken into account in our simulations is
heterogeneous oxidation of OA particles. Similar to gas-phase fragmentation
reactions, heterogeneous oxidation increases the oxidation state of
particulate carbon, resulting in formation of volatile products escaping
irreversibly to the gas phase. Kroll et al. (2015) estimated carbon loss
from the particles after 1 week of OA aging under typical atmospheric
conditions to be in the range from 3 % to 13 %. Heterogeneous reactions
would be slowed down further if the particles are in a semisolid state
(Shiraiwa et al., 2013a). These estimates mean that heterogeneous oxidation
is unlikely to affect BB OA behavior significantly at the timescales
addressed in this study except that including even a slow heterogeneous
oxidation into our model would inevitably transform regime 1 (a
monotonic saturating increase in EnR, leading to a steady state) into
regime 2 (an increase in EnR, followed by its decrease), as
heterogeneous oxidation would eventually convert all particulate carbon into
CO2 (Donahue et al., 2013). Finally, it should again be noted that none
of the VBS schemes considered take into account interaction of SVOCs with
water. That is, we essentially assume that both POA and SOA are formed of
hydrophobic species. This assumption is not expected to have a strong effect
on BB aerosol evolution in a relatively dry atmosphere (with the relative
humidity below about 60 %) where the water uptake by BB aerosol is small
(Hand et al., 2010). Significant uptake of water can result in important
contributions of aqueous-phase oxidation reactions to transformations of BB
OA chemical composition and to formation of SOA (Brege et al., 2018).
As argued above, the nonlinear behavior of BB OA in our model experiments
stems merely from the well-established semi-volatile nature of organic
compounds composing it. This fact indicates that the nonlinear features of
BB OA evolution, which have been revealed in our simulations, are not a
consequence of any simplifications involved in our model but rather an
inherent property of real BB aerosol. However, the great diversity of the
simulations using different VBS schemes indicates that an accurate
quantitative representation of these features in the presently available
atmospheric models is yet hardly feasible. Taking into account that regional
and global CTMs are not designed to address the scales associated with
individual plumes, the results of our study indicate the need for robust
sub-grid parameterizations of BB OA evolution. Instead of explicitly assuming
variable properties of individual smoke plumes and the BB aerosol they
contain, such parameterizations might rely on some external observable
characteristics of fire emissions, such as the density and size distribution
of fire spots and the fire radiative power, as well as on common
meteorological parameters (e.g., temperature and relative humidity). Useful
observational constraints for robust representations of atmospheric aging of
BB aerosol in CTMs can hopefully be inferred from satellite measurements of
BB aerosol (Konovalov et al., 2017).
Conclusions
In this paper, we analyzed the role of the intrinsic nonlinearity of the
processes driving gas–particle partitioning and oxidation of SVOCs during
the atmospheric evolution of BB organic aerosol. We performed simulations of
BB OA evolution during a 5 d period using a microphysical box model in
which BB OA chemical transformations and SOA formation were represented
within the VBS framework. A simple parameterization based on a Gaussian
dispersion model was used to specify several scenarios for dilution of a BB
plume. The model was run with several VBS schemes of varying complexity,
including 1-D-VBS, 1.5-D-VBS, quasi-2-D-VBS, and 2-D-VBS schemes that had been proposed in
the literature to represent BB OA evolution specifically in the framework of
regional and global chemistry-transport models (CTMs). We analyzed the BB OA
evolution by considering the BB OA mass enhancement ratio (EnR) to be a
function of two control parameters, i.e., the initial horizontal plume size,
Sp, across the wind direction and the initial BB OA concentration,
C0, corresponding to thermodynamic equilibrium. For a part of our
simulations, we also considered the dependence of the hygroscopicity
parameter, κorg, on the same parameters. The initial plume size
controls the dilution rate (the larger Sp, the smaller the dilution
rate), and the initial BB OA concentration determines the partitioning of
SVOCs between the gas phase and the particles (the larger C0, the
smaller the gas-phase fraction of SVOCs).
The simulation results allowed us to identify several qualitatively
different regimes of BB OA evolution, which feature a monotonic saturating
increase in EnR (regime 1), an increasing and then decreasing EnR (regime 2), an increase in EnR after its initial rapid decrease (regime 3),
a stage with an increasing EnR between two intermittent stages of its decrease
(regime 4), or a monotonically decreasing EnR (regime 5). The
manifestations of nonlinear behavior of BB OA are found to include
pronounced dependencies of EnR on both Sp and C0. For relatively
fresh BB aerosol (with the age ranging from a few hours to several tens of
hours), EnR increases as Sp increases or C0 decreases. However,
these kinds of dependencies can be strongly suppressed or even reversed,
depending on the VBS scheme used and the aerosol age. Another interesting
manifestation of nonlinear behavior of BB OA is possible shifts between the
regimes as a result of a change in Sp: for example, the regimes 1
and 2 for slowly diluting large smoke plumes can transform into the
regimes 3 and 4, respectively, for small plumes with fast initial
dilution. Evolution of κorg is also found to be affected by
changes of the control parameters in a nonlinear way, although the
manifestations of the nonlinearities are not as spectacular as in the case
of the corresponding dependencies for EnR.
Differences between the VBS schemes can result in large quantitative
differences between the simulations; they increase with the aerosol age and
can almost be as large as a factor of 10 in the EnR values after a 5 d
evolution under typical conditions in summer midlatitudes. Such large
quantitative differences are usually associated with qualitative differences
between the simulations: specifically, the simulations resulting in the
largest values of EnR correspond to regime 1 of BB OA evolution, while
those yielding relatively small EnR values typically correspond to regimes 2, 4, or 5. Our analysis indicates that one of the major factors
behind the quantitative and qualitative differences between the simulations
with the different VBS schemes is the ratio between fragmentation and
functionalization. Specifically, prevalence of fragmentation over
functionalization (when the effective fragmentation branching ratio exceeds
0.5) gives rise to regimes 2 and 4 and is associated with an
eventual decrease in EnR, while the dominance of functionalization over
fragmentation is associated with regimes 1 and 3, which correspond
to a saturating increase in EnR. A change of the fragmentation branching
ratio also can eventually cause a reversal of the dependence of EnR
calculated after a 5 d evolution on Sp: that is, EnR increases with
increasing Sp if fragmentation is prevailing over functionalization and
decreases otherwise. Future studies involving a microphysical model with
different OA oxidation schemes are needed to examine how the BB OA behavior
depends on varying environmental conditions and on several factors (e.g., a
possibility of SOA being in semi-glassy state and heterogeneous and aqueous-phase oxidation reactions) that have not been taken into account in our
simulations.
We argue that in spite of the inevitable limitations of our study, its
results have important implications for modeling of BB OA in the framework
of CTMs. First, our analysis allows us to point out nonlinear behavior of
the OA system as a possible reason for the observed diversity of effects of
aging of ambient BB aerosol (see, for example, Cubison et al., 2011). A better
understanding of the factors behind the diversity of BB OA aging effects is
essential for ensuring the efficiency of in situ ambient observations of BB
OA as observational constraints to representations of BB OA processes in
CTMs. Second, our findings suggest that uncertainties associated with the
representation of BB OA in CTMs may have a major impact on the simulated
behavior of BB aerosol at the scales of its typical lifetime in the boundary
layer with respect to dry deposition. Note that these uncertainties can be
especially important in the context of modeling rapid climate change in the
Arctic, where BB OA provides a considerable contribution to the radiative
balance (Sand et al., 2015). Third, a rather strong sensitivity of EnR
evolution to the parameters of a BB plume (such as C0 and Sp)
indicates that application of even a perfect VBS scheme in the framework of
available regional or global models would likely entail considerable model
biases in simulations of atmospheric transformations of BB aerosol due to
the fact that the evolution of individual BB plumes with varying parameters
cannot be accurately represented on a typical model grid with a horizontal
resolution ranging from tens to a few hundreds of kilometers. Overall, our
findings call for the development of sub-grid parameterizations of the BB OA
evolution, which could be constrained with available in situ and satellite
measurements but, at the same time, would be sufficiently robust with
respect to nonlinear effects that cannot be properly addressed in typical
CTMs.
Data availability
The data presented in this paper were obtained by integrating the dynamic
equations specified in Sect. 2.1 and are available upon request from the
corresponding author. Numerical codes representing the equations and the
involved chemical and microphysical processes in the framework of the box
model used in this study are based on the routines and interfaces of the
CHIMERE chemistry-transport model. The CHIMERE codes are available at
http://www.lmd.polytechnique.fr/chimere/ (last access: 23 April 2019).
Author contributions
IBK and MB designed the study. IBK
also performed the simulations and data analysis and wrote the paper. NAG
contributed to development of the box model. MOA contributed to the
discussion of the results and to writing the paper.
Competing interests
The authors declare that they have no conflict of interest.
Acknowledgements
The authors thank the anonymous reviewers
for their helpful comments.
Financial support
The simulations and analysis of evolution of the BB OA mass enhancement ratio were supported by the Russian Foundation for Basic Research (grant no. 18-05-00911). The simulations of the hygroscopicity parameter were performed with support from the Russian Science Foundation (grant agreement no. 18-17-00076). Igor B. Konovalov was supported by travel expenses in the framework of the French PARCS (Pollution in the ARCtic System – PARCS) project.
Review statement
This paper was edited by Kostas Tsigaridis and reviewed by two anonymous referees.
ReferencesAdler, G., Flores, J. M., Abo Riziq, A., Borrmann, S., and Rudich, Y.: Chemical, physical, and optical evolution of biomass burning aerosols: a case study, Atmos. Chem. Phys., 11, 1491–1503, 10.5194/acp-11-1491-2011, 2011.Ahern, A. T., Robinson, E. S., Tkacik, D. S., Saleh, R., Hatch, L. E.,
Barsanti, K. C., Stockwell, C. E., Yokelson, R. J., Presto, A. A., Robinson,
A. L., Sullivan, R. C., and Donahue, N. M.: Production of secondary organic
aerosol during aging of biomass burning smoke from fresh fuels and its
relationship to VOC precursors, J. Geophys. Res., 124, 3583–3606,
10.1029/2018JD029068, 2019.Akagi, S. K., Craven, J. S., Taylor, J. W., McMeeking, G. R., Yokelson, R. J., Burling, I. R., Urbanski, S. P., Wold, C. E., Seinfeld, J. H., Coe, H., Alvarado, M. J., and Weise, D. R.: Evolution of trace gases and particles emitted by a chaparral fire in California, Atmos. Chem. Phys., 12, 1397–1421, 10.5194/acp-12-1397-2012, 2012.Andreae, M. O.: Emission of trace gases and aerosols from biomass burning – an updated assessment, Atmos. Chem. Phys., 19, 8523–8546, 10.5194/acp-19-8523-2019, 2019.Andreae, M. O. and Gelencsér, A.: Black carbon or brown carbon? The nature of light-absorbing carbonaceous aerosols, Atmos. Chem. Phys., 6, 3131–3148, 10.5194/acp-6-3131-2006, 2006.Archer-Nicholls, S., Lowe, D., Schultz, D. M., and McFiggans, G.: Aerosol–radiation–cloud interactions in a regional coupled model: the effects of convective parameterisation and resolution, Atmos. Chem. Phys., 16, 5573–5594, 10.5194/acp-16-5573-2016, 2016.Barsanti, K. C. and Pankow, J. F.: Thermodynamics of the formation of
atmospheric organic particulate matter by accretion reactions – Part 1:
aldehydes and ketones, Atmos. Environ., 38, 4371–4382,
10.1016/j.atmosenv.2004.03.035, 2004.Bekryaev, R. V., Polyakov, I. V., and Alexeev, V. A.: Role of polar
amplification in long-term surface air temperature variations and modern
Arctic warming, J. Climate, 23, 3888–3906,
10.1175/2010jcli3297.1, 2010.Bertrand, A., Stefenelli, G., Jen, C. N., Pieber, S. M., Bruns, E. A., Ni, H., Temime-Roussel, B., Slowik, J. G., Goldstein, A. H., El Haddad, I., Baltensperger, U., Prévôt, A. S. H., Wortham, H., and Marchand, N.: Evolution of the chemical fingerprint of biomass burning organic aerosol during aging, Atmos. Chem. Phys., 18, 7607–7624, 10.5194/acp-18-7607-2018, 2018.Bessagnet, B., Menut, L., Curci, G., Hodzic, A., Guillaume, B., Liousse, C.,
Moukhtar, S., Pun, B., Seigneur, C., and Schulz, M.: Regional modeling of
carbonaceous aerosols over Europe – Focus on Secondary Organic Aerosols, J.
Atmos. Chem., 61, 175–202, 10.1007/s10874-009-9129-2, 2008.Bessagnet, B., Pirovano, G., Mircea, M., Cuvelier, C., Aulinger, A., Calori, G., Ciarelli, G., Manders, A., Stern, R., Tsyro, S., García Vivanco, M., Thunis, P., Pay, M.-T., Colette, A., Couvidat, F., Meleux, F., Rouïl, L., Ung, A., Aksoyoglu, S., Baldasano, J. M., Bieser, J., Briganti, G., Cappelletti, A., D'Isidoro, M., Finardi, S., Kranenburg, R., Silibello, C., Carnevale, C., Aas, W., Dupont, J.-C., Fagerli, H., Gonzalez, L., Menut, L., Prévôt, A. S. H., Roberts, P., and White, L.: Presentation of the EURODELTA III intercomparison exercise – evaluation of the chemistry transport models' performance on criteria pollutants and joint analysis with meteorology, Atmos. Chem. Phys., 16, 12667–12701, 10.5194/acp-16-12667-2016, 2016.Bian, Q., Jathar, S. H., Kodros, J. K., Barsanti, K. C., Hatch, L. E., May, A. A., Kreidenweis, S. M., and Pierce, J. R.: Secondary organic aerosol formation in biomass-burning plumes: theoretical analysis of lab studies and ambient plumes, Atmos. Chem. Phys., 17, 5459–5475, 10.5194/acp-17-5459-2017, 2017.Bond, T. C., Doherty, S. J., Fahey, D. W., Forster, P. M., Berntsen, T.,
DeAngelo, B. J., Flanner, M. G., Ghan, S., Kärcher, B., Koch, D., Kinne,
S., Kondo, Y., Quinn, P. K., Sarofim, M. C., Schultz, M. G., Schulz, M.,
Venkataraman, C., Zhang, H., Zhang, S., Bellouin, N., Guttikunda, S. K.,
Hopke, P. K., Jacobson, M. Z., Kaiser, J. W., Klimont, Z., Lohmann, U.,
Schwarz, J. P., Shindell, D., Storelvmo, T., Warren, S. G., and Zender, C.
S.: Bounding the role of black carbon in the climate system: A scientific
assessment, J. Geophys. Res.-Atmos., 118, 5380–5552,
10.1002/jgrd.50171, 2013.Brege, M., Paglione, M., Gilardoni, S., Decesari, S., Facchini, M. C., and Mazzoleni, L. R.: Molecular insights on aging and aqueous-phase processing from ambient biomass burning emissions-influenced Po Valley fog and aerosol, Atmos. Chem. Phys., 18, 13197–13214, 10.5194/acp-18-13197-2018, 2018.Brito, J., Rizzo, L. V., Morgan, W. T., Coe, H., Johnson, B., Haywood, J., Longo, K., Freitas, S., Andreae, M. O., and Artaxo, P.: Ground-based aerosol characterization during the South American Biomass Burning Analysis (SAMBBA) field experiment, Atmos. Chem. Phys., 14, 12069–12083, 10.5194/acp-14-12069-2014, 2014.Burnett, R., Chen, H., Szyszkowicz, M., Fann, N., Hubbell, B., Pope, C. A.,
Apte, J. S., Brauer, M., Cohen, A., Weichenthal, S., Coggins, J., Di, Q.,
Brunekreef, B., Frostad, J., Lim, S. S., Kan, H., Walker, K. D., Thurston,
G. D., Hayes, R. B., Lim, C. C., Turner, M. C., Jerrett, M., Krewski, D.,
Gapstur, S. M., Diver, W. R., Ostro, B., Goldberg, D., Crouse, D. L.,
Martin, R. V., Peters, P., Pinault, L., Tjepkema, M., van Donkelaar, A.,
Villeneuve, P. J., Miller, A. B., Yin, P., Zhou, M., Wang, L., Janssen, N.
A. H., Marra, M., Atkinson, R. W., Tsang, H., Quoc Thach, T., Cannon, J. B.,
Allen, R. T., Hart, J. E., Laden, F., Cesaroni, G., Forastiere, F.,
Weinmayr, G., Jaensch, A., Nagel, G., Concin, H., and Spadaro, J. V.: Global
estimates of mortality associated with long-term exposure to outdoor fine
particulate matter, P. Natl. Acad. Sci., 115, 9592–9597,
10.1073/pnas.1803222115, 2018.Camredon, M., Aumont, B., Lee-Taylor, J., and Madronich, S.: The SOA/VOC/NOx system: an explicit model of secondary organic aerosol formation, Atmos. Chem. Phys., 7, 5599–5610, 10.5194/acp-7-5599-2007, 2007.Capes, G., Johnson, B., McFiggans, G., Williams, P. I., Haywood, J., and
Coe, H.: Aging of biomass burning aerosols over West Africa: Aircraft
measurements of chemical composition, microphysical properties, and emission
ratios, J. Geophys. Res., 113, D00C15, 10.1029/2008jd009845, 2008.Chang, R. Y.-W., Slowik, J. G., Shantz, N. C., Vlasenko, A., Liggio, J., Sjostedt, S. J., Leaitch, W. R., and Abbatt, J. P. D.: The hygroscopicity parameter (κ) of ambient organic aerosol at a field site subject to biogenic and anthropogenic influences: relationship to degree of aerosol oxidation, Atmos. Chem. Phys., 10, 5047–5064, 10.5194/acp-10-5047-2010, 2010.Chatfield, R. B. and Delany, A. C.: Convection links biomass burning to
increased tropical ozone; However, models will tend to over-predict O3,
J. Geophys. Res., 95, 18473–18488, 1990.Ciarelli, G., El Haddad, I., Bruns, E., Aksoyoglu, S., Möhler, O., Baltensperger, U., and Prévôt, A. S. H.: Constraining a hybrid volatility basis-set model for aging of wood-burning emissions using smog chamber experiments: a box-model study based on the VBS scheme of the CAMx model (v5.40), Geosci. Model Dev., 10, 2303–2320, 10.5194/gmd-10-2303-2017, 2017a.Ciarelli, G., Aksoyoglu, S., El Haddad, I., Bruns, E. A., Crippa, M., Poulain, L., Äijälä, M., Carbone, S., Freney, E., O'Dowd, C., Baltensperger, U., and Prévôt, A. S. H.: Modelling winter organic aerosol at the European scale with CAMx: evaluation and source apportionment with a VBS parameterization based on novel wood burning smog chamber experiments, Atmos. Chem. Phys., 17, 7653–7669, 10.5194/acp-17-7653-2017, 2017b.Cubison, M. J., Ortega, A. M., Hayes, P. L., Farmer, D. K., Day, D., Lechner, M. J., Brune, W. H., Apel, E., Diskin, G. S., Fisher, J. A., Fuelberg, H. E., Hecobian, A., Knapp, D. J., Mikoviny, T., Riemer, D., Sachse, G. W., Sessions, W., Weber, R. J., Weinheimer, A. J., Wisthaler, A., and Jimenez, J. L.: Effects of aging on organic aerosol from open biomass burning smoke in aircraft and laboratory studies, Atmos. Chem. Phys., 11, 12049–12064, 10.5194/acp-11-12049-2011, 2011.DeCarlo, P. F., Dunlea, E. J., Kimmel, J. R., Aiken, A. C., Sueper, D., Crounse, J., Wennberg, P. O., Emmons, L., Shinozuka, Y., Clarke, A., Zhou, J., Tomlinson, J., Collins, D. R., Knapp, D., Weinheimer, A. J., Montzka, D. D., Campos, T., and Jimenez, J. L.: Fast airborne aerosol size and chemistry measurements above Mexico City and Central Mexico during the MILAGRO campaign, Atmos. Chem. Phys., 8, 4027–4048, 10.5194/acp-8-4027-2008, 2008.Donahue, N. M., Robinson, A. L., Stanier, C. O., and Pandis, S. N.: Coupled
partitioning, dilution, and chemical aging of semivolatile organics,
Environ. Sci. Technol., 40, 2635–2643, 10.1021/es052297c,
2006.Donahue, N. M., Epstein, S. A., Pandis, S. N., and Robinson, A. L.: A two-dimensional volatility basis set: 1. organic-aerosol mixing thermodynamics, Atmos. Chem. Phys., 11, 3303–3318, 10.5194/acp-11-3303-2011, 2011.Donahue, N. M., Kroll, J. H., Pandis, S. N., and Robinson, A. L.: A two-dimensional volatility basis set – Part 2: Diagnostics of organic-aerosol evolution, Atmos. Chem. Phys., 12, 615–634, 10.5194/acp-12-615-2012, 2012a.Donahue, N. M., Henry, K. M., Mentel, T. F., Kiendler-Scharr, A., Spindler,
C., Bohn, B., Brauers, T., Dorn, H. P., Fuchs, H., Tillmann, R., Wahner, A.,
Saathoff, H., Naumann, K.-H., Möhler, O., Leisner, T., Müller, L.,
Reinnig, M.-C., Hoffmann, T., Salo, K., Hallquist, M., Frosch, M., Bilde,
M., Tritscher, T., Barmet, P., Praplan, A. P., DeCarlo, P. F., Dommen, J.,
Prévôt, A. S. H., and Baltensperger, U.: Aging of biogenic secondary
organic aerosol via gas-phase OH radical reactions, P. Natl. Acad. Sci.,
109, 13503–13508, 10.1073/pnas.1115186109, 2012b.Donahue, N. M., Chuang, W., Epstein, S. A., Kroll, J. H., Worsnop, D. R.,
Robinson, A. L., Adams, P. J., and Pandis, S. N.: Why do organic aerosols
exist? Understanding aerosol lifetimes using the two-dimensional volatility
basis set, Environ. Chem., 10, 151–157, 10.1071/EN13022,
2013.Fan, X., Yu, X., Wang, Y., Xiao, X., Li, F., Xie, Y., Wei, S., Song, J., and
Peng, P.: The aging behaviors of chromophoric biomass burning brown carbon
during dark aqueous hydroxyl radical oxidation processes in laboratory
studies, Atmos. Environ., 205,
9–18, 10.1016/j.atmosenv.2019.02.039, 2019.Fang, Z., Deng, W., Zhang, Y., Ding, X., Tang, M., Liu, T., Hu, Q., Zhu, M., Wang, Z., Yang, W., Huang, Z., Song, W., Bi, X., Chen, J., Sun, Y., George, C., and Wang, X.: Open burning of rice, corn and wheat straws: primary emissions, photochemical aging, and secondary organic aerosol formation, Atmos. Chem. Phys., 17, 14821–14839, 10.5194/acp-17-14821-2017, 2017.Feng, Y., Ramanathan, V., and Kotamarthi, V. R.: Brown carbon: a significant atmospheric absorber of solar radiation?, Atmos. Chem. Phys., 13, 8607–8621, 10.5194/acp-13-8607-2013, 2013.Forrister, H., Liu, J., Scheuer, E., Dibb, J., Ziemba, L., Thornhill, K. L.,
Anderson, B., Diskin, G., Perring, A. E., Schwarz, J. P., Campuzano-Jost,
P., Day, D. A., Palm, B. B., Jimenez, J. L., Nenes, A., and Weber, R. J.:
Evolution of brown carbon in wildfire plumes, Geophys. Res. Lett., 42,
2015GL063897, 10.1002/2015GL063897, 2015.
Gelbard, F. and Seinfeld, J. H.: Simulation of multicomponent aerosol
dynamics, J. Colloid Interf. Sci., 78, 485–501, 1980.Grieshop, A. P., Logue, J. M., Donahue, N. M., and Robinson, A. L.: Laboratory investigation of photochemical oxidation of organic aerosol from wood fires 1: measurement and simulation of organic aerosol evolution, Atmos. Chem. Phys., 9, 1263–1277, 10.5194/acp-9-1263-2009, 2009.Hand, J. L., Day, D. E., McMeeking, G. M., Levin, E. J. T., Carrico, C. M., Kreidenweis, S. M., Malm, W. C., Laskin, A., and Desyaterik, Y.: Measured and modeled humidification factors of fresh smoke particles from biomass burning: role of inorganic constituents, Atmos. Chem. Phys., 10, 6179–6194, 10.5194/acp-10-6179-2010, 2010.Heald, C. L., Kroll, J. H., Jimenez, J. L., Docherty, K. S., DeCarlo, P. F.,
Aiken, A. C., Chen, Q., Martin, S. T., Farmer, D. K., and Artaxo, P.: A
simplified description of the evolution of organic aerosol composition in
the atmosphere, Geophys. Res. Lett., 37, L08803,
10.1029/2010GL042737, 2010.Hennigan, C. J., Miracolo, M. A., Engelhart, G. J., May, A. A., Presto, A. A., Lee, T., Sullivan, A. P., McMeeking, G. R., Coe, H., Wold, C. E., Hao, W.-M., Gilman, J. B., Kuster, W. C., de Gouw, J., Schichtel, B. A., Collett Jr., J. L., Kreidenweis, S. M., and Robinson, A. L.: Chemical and physical transformations of organic aerosol from the photo-oxidation of open biomass burning emissions in an environmental chamber, Atmos. Chem. Phys., 11, 7669–7686, 10.5194/acp-11-7669-2011, 2011.Hobbs, P. V., Sinha, P., Yokelson, R. J., Christian, T. J., Blake, D. R.,
Gao, S., Kirchstetter, T. W., Novakov, T., and Pilewkie, P.: Evolution of
gases and particles from a savanna fire in South Africa, J. Geophys.
Res., 108, 8485, 10.1029/2002JD002392, 2003.Hodzic, A., Jimenez, J. L., Madronich, S., Canagaratna, M. R., DeCarlo, P. F., Kleinman, L., and Fast, J.: Modeling organic aerosols in a megacity: potential contribution of semi-volatile and intermediate volatility primary organic compounds to secondary organic aerosol formation, Atmos. Chem. Phys., 10, 5491–5514, 10.5194/acp-10-5491-2010, 2010.Hodshire, A. L., Bian, Q., Ramnarine, E., Lonsdale, C. R., Alvarado, M. J.,
Kreidenweis, S. M., Jathar, S. H., and Pierce, J. R.: More than emissions
and chemistry: Fire size, dilution, and background aerosol also greatly
influence near-field biomass burning aerosol aging, J. Geophys. Res.-Atmos.,
124, 5589–5611, 10.1029/2018JD029674, 2019.
IPCC: Summary for Policymakers, in: Climate Change 2013: The Physical
Science Basis. Contribution of Working Group I to the Fifth Assessment Report
of the Intergovernmental Panel on Climate Change, edited by: Stocker, T. F.,
Qin, D., Plattner, G.-K., Tignor, M., Allen S. K., Boschung, J., Nauels, A.,
Xia, Y., Bex, V., and Midgley, P. M., Cambridge University Press, Cambridge,
UK and New York, NY, USA, 2013.Jang, M. S., Czoschke, N. M., Lee S., and Kamens, R. M.: Heterogeneous
atmospheric aerosol production by acid-catalyzed particle-phase reactions,
Science, 298, 5594, 814–817,
10.1126/science.1075798, 2002.Jathar, S. H., Gordon, T. D., Hennigan, C. J., Pye, H. O. T., Pouliot, G.,
Adams, P. J., Donahue, N. M., and Robinson, A. L.: Unspeciated organic
emissions from combustion sources and their influence on the secondary
organic aerosol budget in the United States, P. Natl. Acad. Sci. USA, 111,
10473–10478, 10.1073/pnas.1323740111, 2014.Jimenez, J., Canagaratna, M., Donahue, N., Prevot, A., Zhang, Q., Kroll, J.
H., DeCarlo, P. F., Allan, J. D., Coe, H., Ng, N. L., Aiken, A. C.,
Docherty, K. S., Ulbrich, I. M., Grieshop, A. P., Robinson, A. L., Duplissy,
J., Smith, J. D., Wilson, K. R., Lanz, V. A., Hueglin, C., Sun, Y. L., Tian,
J., Laaksonen, A., Raatikaine, T., Rautiainen, J., Vaattovaara, P., Ehn, M.,
Kulmala, M., Tomlinson, J. M., Collins, D. R., Cubison, M. J., Dunlea, E.
J., Huffman, J. A., Onasch, T. B., Alfarra, M. R., Williams, P. I., Bower,
K., Kondo, Y., Schneider, J., Drewnick, F., Borrmann, S., Weimer, S.,
Demerjian, K., Salcedo, D., Cottrell, L., Griffin, R., Takami, A., Miyoshi,
T., Hatakeyama, S., Shimono, A., Sun, J. Y., Zhang, Y. M., Dzepina, K.,
Kimmel, J. R., Sueper, D., Jayne, J. T., Herndon, S. C., Trimborn, A. M.,
Williams, L. R., Wood, E. C., Middlebrook, A. M., Kolb, C. E.,
Baltensperger, U., and Worsnop, D. R.: Evolution of organic aerosols in the
atmosphere, Science, 326, 1525–1529,
10.1126/science.1180353, 2009.Jo, D. S., Park, R. J., Lee, S., Kim, S.-W., and Zhang, X.: A global simulation of brown carbon: implications for photochemistry and direct radiative effect, Atmos. Chem. Phys., 16, 3413–3432, 10.5194/acp-16-3413-2016, 2016.Jolleys, M. D., Coe, H., McFiggans, G., Capes, G., Allan, J. D., Crosier,
J., Williams, P. I., Allen, G., Bower, K. N., Jimenez, J. L., Russell, L.
M., Grutter, M., and Baumgardner, D.: Characterizing the aging of biomass
burning organic aerosol by use of mixing ratios: a meta-analysis of four
regions, Environ. Sci. Technol., 46, 13093–13102,
10.1021/es302386v, 2012.Jolleys, M. D., Coe, H., McFiggans, G., Taylor, J. W., O'Shea, S. J., Le Breton, M., Bauguitte, S. J.-B., Moller, S., Di Carlo, P., Aruffo, E., Palmer, P. I., Lee, J. D., Percival, C. J., and Gallagher, M. W.: Properties and evolution of biomass burning organic aerosol from Canadian boreal forest fires, Atmos. Chem. Phys., 15, 3077–3095, 10.5194/acp-15-3077-2015, 2015.
Klug, W.: A method for determining diffusion conditions from synoptic
observations, Staub-Reinhalt. Luft, 29, 14–20, 1969.Konovalov, I. B., Berezin, E. V., Ciais, P., Broquet, G., Beekmann, M., Hadji-Lazaro, J., Clerbaux, C., Andreae, M. O., Kaiser, J. W., and Schulze, E.-D.: Constraining CO2 emissions from open biomass burning by satellite observations of co-emitted species: a method and its application to wildfires in Siberia, Atmos. Chem. Phys., 14, 10383–10410, 10.5194/acp-14-10383-2014, 2014.Konovalov, I. B., Beekmann, M., Berezin, E. V., Petetin, H., Mielonen, T., Kuznetsova, I. N., and Andreae, M. O.: The role of semi-volatile organic compounds in the mesoscale evolution of biomass burning aerosol: a modeling case study of the 2010 mega-fire event in Russia, Atmos. Chem. Phys., 15, 13269–13297, 10.5194/acp-15-13269-2015, 2015.Konovalov, I. B., Berezin, E. V., and Beekmann, M.: Effect of photochemical
self-action of carbon-containing aerosol: wildfires, Izv. Atmos. Ocean.
Phy., 52, 263–270, 10.1134/S0001433816030063, 2016.Konovalov, I. B., Beekmann, M., Berezin, E. V., Formenti, P., and Andreae, M. O.: Probing into the aging dynamics of biomass burning aerosol by using satellite measurements of aerosol optical depth and carbon monoxide, Atmos. Chem. Phys., 17, 4513–4537, 10.5194/acp-17-4513-2017, 2017.Konovalov, I. B., Lvova, D. A., Beekmann, M., Jethva, H., Mikhailov, E. F., Paris, J.-D., Belan, B. D., Kozlov, V. S., Ciais, P., and Andreae, M. O.: Estimation of black carbon emissions from Siberian fires using satellite observations of absorption and extinction optical depths, Atmos. Chem. Phys., 18, 14889–14924, 10.5194/acp-18-14889-2018, 2018.Koo, B., Knipping, E., and Yarwood, G.: 1.5-Dimensional volatility basis set
approach for modeling organic aerosol in CAMx and CMAQ, Atmos. Environ., 95,
158–164, 10.1016/j.atmosenv.2014.06.031, 2014.Kroll, J. H., Lim, C. Y., Kessler, S. H., and Wilson, K. R.: Heterogeneous
oxidation of atmospheric organic aerosol: Kinetics of changes to the amount
and oxidation state of particlephase organic carbon, J. Phys. Chem. A, 119,
10767–10783, 10.1021/acs.jpca.5b06946, 2015.Lambe, A. T., Onasch, T. B., Massoli, P., Croasdale, D. R., Wright, J. P., Ahern, A. T., Williams, L. R., Worsnop, D. R., Brune, W. H., and Davidovits, P.: Laboratory studies of the chemical composition and cloud condensation nuclei (CCN) activity of secondary organic aerosol (SOA) and oxidized primary organic aerosol (OPOA), Atmos. Chem. Phys., 11, 8913–8928, 10.5194/acp-11-8913-2011, 2011.Lannuque, V., Camredon, M., Couvidat, F., Hodzic, A., Valorso, R., Madronich, S., Bessagnet, B., and Aumont, B.: Exploration of the influence of environmental conditions on secondary organic aerosol formation and organic species properties using explicit simulations: development of the VBS-GECKO parameterization, Atmos. Chem. Phys., 18, 13411–13428, 10.5194/acp-18-13411-2018, 2018.Lee-Taylor, J., Madronich, S., Aumont, B., Baker, A., Camredon, M., Hodzic, A., Tyndall, G. S., Apel, E., and Zaveri, R. A.: Explicit modeling of organic chemistry and secondary organic aerosol partitioning for Mexico City and its outflow plume, Atmos. Chem. Phys., 11, 13219–13241, 10.5194/acp-11-13219-2011, 2011.Lee-Taylor, J., Hodzic, A., Madronich, S., Aumont, B., Camredon, M., and Valorso, R.: Multiday production of condensing organic aerosol mass in urban and forest outflow, Atmos. Chem. Phys., 15, 595–615, 10.5194/acp-15-595-2015, 2015.Lelieveld, J., Klingmüller, K., Pozzer, A., Burnett, R. T., Haines, A.,
and Ramanathan, V.: Effects of fossil fuel and total anthropogenic emission
removal on public health and climate, P. Natl. Acad. Sci., 116, 7192–7197,
10.1073/pnas.1819989116, 2019.Lim, C. Y., Hagan, D. H., Coggon, M. M., Koss, A. R., Sekimoto, K., de Gouw, J., Warneke, C., Cappa, C. D., and Kroll, J. H.: Secondary organic aerosol formation from biomass burning emissions, Atmos. Chem. Phys. Discuss., 10.5194/acp-2019-326, in review, 2019.May, A. A., Levin, E. J. T., Hennigan, C. J., Riipinen, I., Lee, T., Collett
Jr., J. L., Jimenez, J. L., Kreidenweis, S. M., and Robinson, A. L.:
Gas-particle partitioning of primary organic aerosol emissions: 3. Biomass
burning, J. Geophys. Res.-Atmos., 118, 11327–11338,
10.1002/jgrd.50828, 2013.May, A. A., Lee, T., McMeeking, G. R., Akagi, S., Sullivan, A. P., Urbanski, S., Yokelson, R. J., and Kreidenweis, S. M.: Observations and analysis of organic aerosol evolution in some prescribed fire smoke plumes, Atmos. Chem. Phys., 15, 6323–6335, 10.5194/acp-15-6323-2015, 2015.Menut, L., Bessagnet, B., Khvorostyanov, D., Beekmann, M., Blond, N., Colette, A., Coll, I., Curci, G., Foret, G., Hodzic, A., Mailler, S., Meleux, F., Monge, J.-L., Pison, I., Siour, G., Turquety, S., Valari, M., Vautard, R., and Vivanco, M. G.: CHIMERE 2013: a model for regional atmospheric composition modelling, Geosci. Model Dev., 6, 981–1028, 10.5194/gmd-6-981-2013, 2013.Mikhailov, E. F., Mironov, G. N., Pöhlker, C., Chi, X., Krüger, M. L., Shiraiwa, M., Förster, J.-D., Pöschl, U., Vlasenko, S. S., Ryshkevich, T. I., Weigand, M., Kilcoyne, A. L. D., and Andreae, M. O.: Chemical composition, microstructure, and hygroscopic properties of aerosol particles at the Zotino Tall Tower Observatory (ZOTTO), Siberia, during a summer campaign, Atmos. Chem. Phys., 15, 8847–8869, 10.5194/acp-15-8847-2015, 2015.Mikhailov, E. F., Mironova, S., Mironov, G., Vlasenko, S., Panov, A., Chi, X., Walter, D., Carbone, S., Artaxo, P., Heimann, M., Lavric, J., Pöschl, U., and Andreae, M. O.: Long-term measurements (2010–2014) of carbonaceous aerosol and carbon monoxide at the Zotino Tall Tower Observatory (ZOTTO) in central Siberia, Atmos. Chem. Phys., 17, 14365–14392, 10.5194/acp-17-14365-2017, 2017.Murphy, B. N., Donahue, N. M., Fountoukis, C., Dall'Osto, M., O'Dowd, C., Kiendler-Scharr, A., and Pandis, S. N.: Functionalization and fragmentation during ambient organic aerosol aging: application of the 2-D volatility basis set to field studies, Atmos. Chem. Phys., 12, 10797–10816, 10.5194/acp-12-10797-2012, 2012.
Odum, J. R., Jungkamp, T. P.W., Griffin, R. J., Forster, H. J. L., Flagan,
R. C., and Seinfeld, J. H.: Aromatics, reformulated gasoline and atmospheric
organic aerosol formation, Environ. Sci. Technol., 31, 1890–1897, 1997.Pankow, J. F.: An absorption model of gas/particle partitioning of organic
compounds in the atmosphere, Atmos. Environ., 28, 185–188,
10.1016/1352-2310(94)90093-0, 1994.Pankow, J. F. and Asher, W. E.: SIMPOL.1: a simple group contribution method for predicting vapor pressures and enthalpies of vaporization of multifunctional organic compounds, Atmos. Chem. Phys., 8, 2773–2796, 10.5194/acp-8-2773-2008, 2008.Paris, J.-D., Stohl, A., Nédélec, P., Arshinov, M. Yu., Panchenko, M. V., Shmargunov, V. P., Law, K. S., Belan, B. D., and Ciais, P.: Wildfire smoke in the Siberian Arctic in summer: source characterization and plume evolution from airborne measurements, Atmos. Chem. Phys., 9, 9315–9327, 10.5194/acp-9-9315-2009, 2009.Petrenko, M., Kahn, R., Chin, M., and Limbacher, J.: Refined use of
satellite aerosol optical depth snapshots to constrain biomass burning
emissions in the GOCART model, J. Geophys. Res.-Atmos., 122, 10983–11004,
10.1002/2017JD026693, 2017.Petters, M. D. and Kreidenweis, S. M.: A single parameter representation of hygroscopic growth and cloud condensation nucleus activity, Atmos. Chem. Phys., 7, 1961–1971, 10.5194/acp-7-1961-2007, 2007.Pokhrel, R. P., Beamesderfer, E. R., Wagner, N. L., Langridge, J. M., Lack, D. A., Jayarathne, T., Stone, E. A., Stockwell, C. E., Yokelson, R. J., and Murphy, S. M.: Relative importance of black carbon, brown carbon, and absorption enhancement from clear coatings in biomass burning emissions, Atmos. Chem. Phys., 17, 5063–5078, 10.5194/acp-17-5063-2017, 2017.Pope, C. A., Ezzati, M., and Dockery, D. W.: Fine-Particulate Air Pollution
and Life Expectancy in the United States, N. Engl. J. Med., 360, 376–386,
10.1056/NEJMsa0805646, 2009.Pratt, K. A., Murphy, S. M., Subramanian, R., DeMott, P. J., Kok, G. L., Campos, T., Rogers, D. C., Prenni, A. J., Heymsfield, A. J., Seinfeld, J. H., and Prather, K. A.: Flight-based chemical characterization of biomass burning aerosols within two prescribed burn smoke plumes, Atmos. Chem. Phys., 11, 12549–12565, 10.5194/acp-11-12549-2011, 2011.
Pun, B. K., Seigneur, C., and Lohman, K.: Modeling secondary organic aerosol
formation via multiphase partitioning with molecular data, Environ. Sci.
Technol., 40, 4722–4731, 2006.Reddington, C. L., Spracklen, D. V., Artaxo, P., Ridley, D. A., Rizzo, L. V., and Arana, A.: Analysis of particulate emissions from tropical biomass burning using a global aerosol model and long-term surface observations, Atmos. Chem. Phys., 16, 11083–11106, 10.5194/acp-16-11083-2016, 2016.Reid, J. S., Koppmann, R., Eck, T. F., and Eleuterio, D. P.: A review of biomass burning emissions part II: intensive physical properties of biomass burning particles, Atmos. Chem. Phys., 5, 799–825, 10.5194/acp-5-799-2005, 2005.
Riipinen, I., Pierce, J. R., Donahue, N. M., and Pandis, S. N.:
Equilibration time scales of organic aerosol inside thermodenuders:
Evaporation kinetics versus thermodynamics, Atmos. Environ., 44, 597–607,
2010.Robinson, A. L., Donahue, N. M., Shrivastava, M. K., Weitkamp, E. A., Sage,
A. M., Grieshop, A. P., Lane, T. E., Pierce, J. R., and Pandis, S. N.:
Rethinking organic aerosols: Semivolatile emissions and photochemical aging,
Science, 315, 1259–1262, 10.1126/science.1133061, 2007.Sakamoto, K. M., Allan, J. D., Coe, H., Taylor, J. W., Duck, T. J., and Pierce, J. R.: Aged boreal biomass-burning aerosol size distributions from BORTAS 2011, Atmos. Chem. Phys., 15, 1633–1646, 10.5194/acp-15-1633-2015, 2015.Saleh, R., Hennigan, C. J., McMeeking, G. R., Chuang, W. K., Robinson, E. S., Coe, H., Donahue, N. M., and Robinson, A. L.: Absorptivity of brown carbon in fresh and photo-chemically aged biomass-burning emissions, Atmos. Chem. Phys., 13, 7683–7693, 10.5194/acp-13-7683-2013, 2013.Saleh, R., Marks, M., Heo, J., Adams, P. J., Donahue, N. M., and Robinson, A. L.:
Contribution of brown carbon and lensing to the direct radiative effect of
carbonaceous aerosols from biomass and biofuel burning emissions, J.
Geophys. Res.-Atmos., 120, 10285–10296, 10.1002/2015JD023697, 2015.Sand, M., Berntsen, T., von Salzen, K., Flanner, M., Langner, J., and
Victor, D.: Response of arctic temperature to changes in emissions of
short-lived climate forcers, Nat. Clim. Change, 6, 286–289,
10.1038/nclimate2880, 2015.
Seinfeld, J. H. and Pandis, S. N.: Atmospheric chemistry and physics: From
air pollution to climate change, 3rd Edn., J. Wiley & Sons, Inc.,
New Jersey, 2016.Shiraiwa, M., Zuend, A., Bertram, A. K., and Seinfeld, J. H.: Gas-particle
partitioning of atmospheric aerosols: interplay of physical state, non-ideal
mixing and morphology, Phys. Chem. Chem. Phys., 15, 11441–11453,
10.1039/c3cp51595h, 2013a.Shiraiwa, M., Yee, L. D., Schilling, K. A., Loza, C. L., Craven, J. S.,
Zuend, A., Ziemann, P. J., and Seinfeld, J. H.: Size distribution dynamics
reveal particle-phase chemistry in organic aerosol formation, P. Natl.
Acad. Sci. USA, 110, 11746–11750, 10.1073/pnas.1307501110,
2013b.Shiraiwa, M., Li, Y., Tsimpidi, A. P., Karydis, V. A., Berkemeier, T.,
Pandis, S. N., Lelieveld, J., Koop, T., and Pöoschl, U.: Global
distribution of particle phase state in atmospheric secondary organic
aerosols, Nat. Commun., 8, 15002, 10.1038/ncomms15002, 2017.Shrivastava, M., Zelenyuk, A., Imre, D., Easter, R. C., Beranek, J., Zaveri,
R. A., and Fast, J. D.: Implications of low volatility and gas-phase
fragmentation reactions on SOA loadings and their spatial and temporal
evolution in the atmosphere. J. Geophys. Res.-Atmos., 118, 3328–3342,
10.1002/jgrd.50160, 2013.Shrivastava, M., Easter, R., Liu, X., Zelenyuk, A., Singh, B., Zhang, K.,
Ma, P-L, Chand, D., Ghan, S., Jimenez, J. L., Zhang, Q., Fast, J., Rasch,
P., and Tiitta, P.: Global transformation and fate of SOA: Implications of
low volatility SOA and gas phase fragmentation reactions, J. Geophys.
Res.-Atmos., 120, 4169–4195, 10.1002/2014JD022563, 2015.Shrivastava, M., Cappa, C., Fan, J., Goldstein, A., Guenther, A., Jimenez,
J., Kuang, C., Laskin, A., Martin, S., Ng, N., Petaja, T., Pierce, J.,
Rasch, P., Roldin, P., Seinfeld, J., Shilling, J., Smith, J., Thornton, J.,
Volkamer, R., Wang, J., Worsnop, D., Zaveri, R., Zelenyuk, A., and Zhang,
Q.: Recent advances in understanding secondary organic aerosol: Implications
for global climate forcing, Rev. Geophys., 55, 509–559,
10.1002/2016RG000540, 2017.Sofiev, M., Ermakova, T., and Vankevich, R.: Evaluation of the smoke-injection height from wild-land fires using remote-sensing data, Atmos. Chem. Phys., 12, 1995–2006, 10.5194/acp-12-1995-2012, 2012.Theodoritsi, G. N. and Pandis, S. N.: Simulation of the chemical evolution of biomass burning organic aerosol, Atmos. Chem. Phys., 19, 5403–5415, 10.5194/acp-19-5403-2019, 2019.Tiitta, P., Leskinen, A., Hao, L., Yli-Pirilä, P., Kortelainen, M., Grigonyte, J., Tissari, J., Lamberg, H., Hartikainen, A., Kuuspalo, K., Kortelainen, A.-M., Virtanen, A., Lehtinen, K. E. J., Komppula, M., Pieber, S., Prévôt, A. S. H., Onasch, T. B., Worsnop, D. R., Czech, H., Zimmermann, R., Jokiniemi, J., and Sippula, O.: Transformation of logwood combustion emissions in a smog chamber: formation of secondary organic aerosol and changes in the primary organic aerosol upon daytime and nighttime aging, Atmos. Chem. Phys., 16, 13251–13269, 10.5194/acp-16-13251-2016, 2016.Tosca, M. G., Randerson, J. T., and Zender, C. S.: Global impact of smoke aerosols from landscape fires on climate and the Hadley circulation, Atmos. Chem. Phys., 13, 5227–5241, 10.5194/acp-13-5227-2013, 2013.Tsigaridis, K. and Kanakidou, M.: The Present and Future of Secondary
Organic Aerosol Direct Forcing on Climate, Current Climate Change Reports,
4, 1–15, 10.1007/s40641-018-0092-3, 2018.Tsigaridis, K., Daskalakis, N., Kanakidou, M., Adams, P. J., Artaxo, P., Bahadur, R., Balkanski, Y., Bauer, S. E., Bellouin, N., Benedetti, A., Bergman, T., Berntsen, T. K., Beukes, J. P., Bian, H., Carslaw, K. S., Chin, M., Curci, G., Diehl, T., Easter, R. C., Ghan, S. J., Gong, S. L., Hodzic, A., Hoyle, C. R., Iversen, T., Jathar, S., Jimenez, J. L., Kaiser, J. W., Kirkevåg, A., Koch, D., Kokkola, H., Lee, Y. H., Lin, G., Liu, X., Luo, G., Ma, X., Mann, G. W., Mihalopoulos, N., Morcrette, J.-J., Müller, J.-F., Myhre, G., Myriokefalitakis, S., Ng, N. L., O'Donnell, D., Penner, J. E., Pozzoli, L., Pringle, K. J., Russell, L. M., Schulz, M., Sciare, J., Seland, Ø., Shindell, D. T., Sillman, S., Skeie, R. B., Spracklen, D., Stavrakou, T., Steenrod, S. D., Takemura, T., Tiitta, P., Tilmes, S., Tost, H., van Noije, T., van Zyl, P. G., von Salzen, K., Yu, F., Wang, Z., Wang, Z., Zaveri, R. A., Zhang, H., Zhang, K., Zhang, Q., and Zhang, X.: The AeroCom evaluation and intercomparison of organic aerosol in global models, Atmos. Chem. Phys., 14, 10845–10895, 10.5194/acp-14-10845-2014, 2014.Tsimpidi, A. P., Karydis, V. A., Pozzer, A., Pandis, S. N., and Lelieveld, J.: ORACLE 2-D (v2.0): an efficient module to compute the volatility and oxygen content of organic aerosol with a global chemistry–climate model, Geosci. Model Dev., 11, 3369–3389, 10.5194/gmd-11-3369-2018, 2018.Vakkari, V., Beukes, J. P., Dal Maso, M., Aurela, M., Josipovic, M., and van
Zyl, P. G.: Major secondary aerosol formation in southern African open
biomass burning plumes, Nat. Geosci., 11, 580–583,
10.1038/s41561-018-0170-0, 2018.van der Werf, G. R., Randerson, J. T., Giglio, L., van Leeuwen, T. T., Chen, Y., Rogers, B. M., Mu, M., van Marle, M. J. E., Morton, D. C., Collatz, G. J., Yokelson, R. J., and Kasibhatla, P. S.: Global fire emissions estimates during 1997–2016, Earth Syst. Sci. Data, 9, 697–720, 10.5194/essd-9-697-2017, 2017.
Verwer, J.: Gauss-Seidel iteration for stiff ODEs from chemical kinetics, J.
Sci. Comput., 15, 1243–1250, 1994.Wang, Q., Saturno, J., Chi, X., Walter, D., Lavric, J. V., Moran-Zuloaga, D., Ditas, F., Pöhlker, C., Brito, J., Carbone, S., Artaxo, P., and Andreae, M. O.: Modeling investigation of light-absorbing aerosols in the Amazon Basin during the wet season, Atmos. Chem. Phys., 16, 14775–14794, 10.5194/acp-16-14775-2016, 2016.Yao, H., Song, Y., Liu, M., Archer-Nicholls, S., Lowe, D., McFiggans, G., Xu, T., Du, P., Li, J., Wu, Y., Hu, M., Zhao, C., and Zhu, T.: Direct radiative effect of carbonaceous aerosols from crop residue burning during the summer harvest season in East China, Atmos. Chem. Phys., 17, 5205–5219, 10.5194/acp-17-5205-2017, 2017.Yokelson, R. J., Crounse, J. D., DeCarlo, P. F., Karl, T., Urbanski, S., Atlas, E., Campos, T., Shinozuka, Y., Kapustin, V., Clarke, A. D., Weinheimer, A., Knapp, D. J., Montzka, D. D., Holloway, J., Weibring, P., Flocke, F., Zheng, W., Toohey, D., Wennberg, P. O., Wiedinmyer, C., Mauldin, L., Fried, A., Richter, D., Walega, J., Jimenez, J. L., Adachi, K., Buseck, P. R., Hall, S. R., and Shetter, R.: Emissions from biomass burning in the Yucatan, Atmos. Chem. Phys., 9, 5785–5812, 10.5194/acp-9-5785-2009, 2009.Zhou, S., Collier, S., Jaffe, D. A., Briggs, N. L., Hee, J., Sedlacek III, A. J., Kleinman, L., Onasch, T. B., and Zhang, Q.: Regional influence of wildfires on aerosol chemistry in the western US and insights into atmospheric aging of biomass burning organic aerosol, Atmos. Chem. Phys., 17, 2477–2493, 10.5194/acp-17-2477-2017, 2017.