Available via license: CC BY 4.0
Content may be subject to copyright.
1
Aerosol Composition Trends during 2000-2020: In depth insights from model
1
predictions and multiple worldwide observation datasets
2
3
Alexandra P. Tsimpidi1, Susanne M.C. Scholz1, Alexandros Milousis1, Nikolaos
4
Mihalopoulos2, and Vlassis A. Karydis1
5
1 Forschungszentrum Jülich, Inst. for Energy and Climate Research, IEK-8, Jülich, Germany
6
2 National Observatory of Athens, Inst. for Environm. Res. & Sustainable Dev., Athens, 15236,
7
Greece.
8
9
Correspondence to: Alexandra P. Tsimpidi (a.tsimpidi@fz-juelich.de)
10
11
Abstract
12
13
Atmospheric aerosols significantly impact Earth’s climate and air quality. In
14
addition to their number and mass concentrations, their chemical composition
15
influences their environmental and health effects. This study examines global trends in
16
aerosol composition from 2000 to 2020, using the EMAC atmospheric chemistry-
17
climate model and a variety of observational datasets. These include PM2.5 data from
18
regional networks and 744 PM1 datasets from AMS field campaigns conducted at 169
19
sites worldwide. Results show that organic aerosol (OA) is the dominant fine aerosol
20
component in all continental regions, particularly in areas with significant biomass
21
burning and biogenic VOC emissions. EMAC effectively reproduces the prevalence of
22
secondary OA but underestimates the aging of OA in some cases, revealing
23
uncertainties in distinguishing fresh and aged SOA. While sulfate is a major aerosol
24
component in filter-based observations, AMS and model results indicate nitrate
25
predominates in Europe and Eastern Asia. Mineral dust also plays a critical role in
26
specific regions, as highlighted by EMAC. The study identifies substantial declines in
27
sulfate, nitrate, and ammonium concentrations in Europe and North America, attributed
28
to emission controls, with varying accuracy in model predictions. In Eastern Asia,
29
sulfate reductions due to SO2 controls are partially captured by the model. OA trends
30
differ between methodologies, with filter data showing slight decreases, while AMS
31
data and model simulations suggest slight increases in PM1 OA across Europe, North
32
America, and Eastern Asia. This research underscores the need for integrating advanced
33
models and diverse datasets to better understand aerosol trends and guide
34
environmental policy.
35
36
37
38
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
2
1. Introduction
39
Atmospheric aerosols are tiny solid or liquid particulate matter (PM) suspended in
40
the air, ranging in size from a few nanometers to several micrometers. Atmospheric
41
aerosol, especially fine particles with diameters less than 2.5 micrometers (PM2.5),
42
poses health risks as it can penetrate deep into the respiratory system (Who, 2003).
43
Long-term exposure to high levels of PM has been associated with respiratory and
44
cardiovascular diseases (Brook et al., 2010; George et al., 2017). Dominici et al. (2006)
45
and Pope et al. (2009) highlight the impact of PM on mortality and morbidity, while
46
more recent studies have determined that the air pollution by PM2.5 is responsible for
47
more than 3 million premature deaths per year worldwide (Lelieveld et al., 2015; Who,
48
2022). As a result, air pollution is recognized as the largest environmental threat to
49
human health in the recent WHO report (Who, 2021). Furthermore, aerosols can
50
directly influence the Earth's climate by scattering and absorbing sunlight, leading to
51
changes in radiation balance (Haywood and Boucher, 2000; Ipcc, 2013). Aerosols can
52
also affect the Earth’s energy balance indirectly through interactions with clouds, i.e.,
53
by serving as cloud condensation (CCN) and ice (IN) nuclei, affecting cloud formation,
54
cloud properties, and precipitation patterns (Andreae and Rosenfeld, 2008). Beside the
55
number and mass concentrations of atmospheric aerosol, its chemical composition
56
determines its aerosol-related climatic (Klingmuller et al., 2019; Klingmüller et al.,
57
2020; Kok et al., 2023) and health impacts (Lelieveld et al., 2015; Fang et al., 2017;
58
Karydis et al., 2021).
59
Atmospheric aerosols have various precursors, and they can be categorized into
60
primary and secondary aerosols based on their origin. Primary sources include natural
61
processes such as volcanic eruptions, wildfires, and sea spray, as well as human
62
activities like industrial emissions and transportation. Secondary aerosols are formed
63
through the oxidation of gas phase pollutants in the atmosphere. Sulfate aerosols are
64
formed through the oxidation of sulfur dioxide (SO2) which is primarily released from
65
the burning of fossil fuels, particularly coal, and natural sources like volcanoes. Nitrate
66
aerosols result from the atmospheric oxidation of nitrogen oxides (NOx) emitted from
67
combustion processes, such as those in vehicles and power plants. Ammonium is
68
formed by the reaction of ammonia (NH₃), which is emitted from agricultural activities
69
and waste treatment, with an acid. Secondary organic aerosols (SOA) can form through
70
the oxidation of volatile organic compounds (VOCs), which are emitted from
71
vegetation, industrial processes, and vehicle exhaust.
72
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
3
Several measures have been discussed and implemented to mitigate pollutants
73
emitted from specific source sectors including transport, energy (power generation,
74
industries etc.), waste management, urban planning and agriculture. A few of the most
75
prominent global conferences that have taken place for the purpose of combating
76
climate change and air pollution are the Conferences of the Parties (COP) since the
77
early 90s, and the supreme decision-making body of the United Nations’ Framework
78
Convention on Climate Change (UNFCCC). Their passed agreements binding the
79
parties to individual emission targets are for instance the Agenda 21 of 1992, the Kyoto
80
Protocol of 1997 and its successor - the Paris Agreement of 2015. Besides these global
81
agreements, the single parties had to implement national or continental plans to meet
82
air quality requirements. The resulting emission trends have been so drastic that aerosol
83
composition has been unevenly altered in different parts of the world. Most European
84
countries are bound by the Gothenburg Protocol targets from 1999 and its amendment
85
from 2012 and have in majority successfully reduced pollutant levels (Emep, 2021).
86
SOx emissions have declined the most, by more than 80% in the last two decades. NOx
87
emissions have declined significantly as well (by 50%), but for NH3 only very small
88
reductions have been achieved (~10%) (Hoesly et al., 2018; Emep, 2021). NMVOCs
89
have also been significantly decreased due to emission controls to the transportation
90
and the solvents sector (Hoesly et al., 2018). In the US, pollutant levels are controlled
91
through regulations imposed by the National Ambient Air Quality Standards
92
(NAAQS), the Regional Haze Rule and the US Clean Air act of 1970. The US and
93
Canada are also part of the Gothenburg protocol. Over Asia, South Korea and China
94
belong to the Newly Industrialized and high-growth economies. Especially from 1980
95
to the mid-2000s, pollutants emissions grew in China (Hoesly et al., 2018; Zhai et al.,
96
2019). However, in the face of the Beijing Olympic Games in 2008, there have been
97
drastic endeavors of air pollution control in Beijing and neighboring administrative
98
regions (Huang et al., 2010). In 2013, the first consistent and aggressive emission
99
controls started under the Clean Air Action (Zhai et al., 2019). The Clean Air Action
100
has identified three target regions, the megacity clusters of Beijing-Tianjin-Hebei,
101
Yangtze River Delta and the Pearl River Delta, while in 2018, the latter was replaced
102
by the Fenwei Plain (Zhai et al., 2019).
103
Air pollution concentration levels can vary by time of day, season, across large spans
104
of time, based on meteorological factors, and in connection to climate change. Trends
105
analysis of air pollution concentrations (Guerreiro et al., 2014; Lang et al., 2019) can
106
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
4
allow the assessment of the impact of various factors on air quality including changes
107
in industrial activities, traffic patterns, or energy production. Analyzing trends in air
108
pollutants enables comparisons between different regions or countries (Anttila and
109
Tuovinen, 2010; Chow et al., 2022; Kyllönen et al., 2020) as well as between different
110
datasets that provide information for the same pollutant. This can highlight areas that
111
are successfully addressing air quality issues, provide benchmarks for others to follow
112
but also highlight any kind of inability of each method to reproduce the concentration
113
levels of the pollutants.
114
In this study, we use the comprehensive atmospheric chemistry-climate model
115
EMAC to present 20-year global composition trends of fine aerosols in different regions
116
of the planet. Here, for the first time, EMAC uses a computationally lite version of the
117
organic aerosol module ORACLE (Tsimpidi et al., 2014) and the new highly
118
computationally efficient module ISORROPIA-lite (Kakavas et al., 2022; Milousis et
119
al., 2024). The large emission trends in our model are considered by employing the
120
Copernicus Atmosphere Monitoring Service (CAMS) inventory for anthropogenic
121
emissions (Granier et al., 2019). Model results are combined with a global observational
122
aerosol composition dataset to provide insights into the large spatiotemporal changes
123
in aerosol composition over the past two decades, driven by changes in aerosol
124
precursor emissions. The dataset includes observations from regional filter-based
125
monitoring networks that routinely collect PM2.5 (e.g. EMEP, IMPROVE, EPA,
126
EANET, SPARTAN), and a unique comprehensive compilation of 744 individual
127
Aerosol Mass Spectometer (AMS) field campaigns worldwide that provide in-situ
128
measurements of PM1 composition.
129
130
2. Observational Dataset
131
132
2.1 PM1 Dataset
133
Since the year 2000, the quadrupole-based Aerodyne aerosol mass spectrometer (Q-
134
AMS) and its successors enjoy great popularity as a method for atmospheric aerosol
135
sampling. A great advantage of AMS is its ability to deliver high-resolved real-time
136
quantitative data on mass concentration of particles between ~ 0.05 - 1 µm
137
(Canagaratna et al., 2007) as a function of their non-refractory chemical composition
138
(i.e., OA and inorganic SO42-, NO3-, NH4+, and Cl-) (Jayne et al., 2000). Thus, over the
139
years and numerous field campaigns, a lot of valuable chemical and microphysical
140
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
5
information about ambient aerosols has been obtained (Ng et al., 2011). During 2000s,
141
these campaigns did not last more than a month, however, the development of the
142
Aerosol Chemical Speciation Monitor (ACSM), a small and cost-efficient version of
143
AMS, allowed the long-term monitoring of the PM1 composition over several locations
144
during the 2010s.
145
146
2.1.1 AMS factor analysis techniques
147
The AMS spectra of OA are often further analyzed via factor analysis techniques in
148
order to extract detailed information about the OA composition as well. Among factor
149
analysis techniques (e.g., ME-2 (Paatero, 1999); PCA (Zhang et al., 2013); MCA
150
(Zhang et al., 2007; Cottrell et al., 2008)), the PMF (Paatero and Tapper, 1994; Paatero,
151
1997) is the most popular technique, occasionally in combination with the ME-2.
152
Overall, a mass spectrum that peaks at m/z = 44 (or ƒ44) and m/z = 43 (or ƒ43) is mostly
153
dominated by the CO2+ and C2H3O+ ions, respectively. The first is mostly linked to
154
acidic groups (i.e, -COOH), typically associated with chemically aged and oxygenated
155
organic aerosols (OOA), while the latter is dominated by non-acid oxygenates. OOA
156
can be further categorized into different levels of aging and volatility stages. Most
157
commonly, a less oxidized (semi-volatile) OA (L-OOA (Bougiatioti et al., 2014)) and
158
a more oxidized (low-volatile) OA (M-OOA (Bozzetti et al., 2017)) are distinguished
159
(Jimenez et al., 2009; Ng et al., 2010; Crippa et al., 2014; Stavroulas et al., 2019). The
160
two OOA factors could be identified on the basis of the ƒ44 to ƒ43 ratio: M-OOA
161
component spectra have a higher ƒ44, while L-OOA component spectra have slightly
162
higher ƒ43. Besides these general factors, other oxygenated OA compounds have been
163
resolved in some campaigns. One of the most important is the IEPOX-OA with
164
abundant ions at m/z = 53, 75, or 82. This “isoprene” factor correlates strongly with
165
molecular tracers of SOA that are derived from isoprene epoxydiols (Xu et al., 2015;
166
Budisulistiorini et al., 2013; Budisulistiorini et al., 2016). Several campaigns in North
167
America have found IEPOX-OA, as have campaigns in South America and Australia.
168
Furthermore, methane-sulfonic acid (MSA) is often retrieved from datasets of marine
169
sites (Crippa et al., 2014; Mallet et al., 2019). Some studies could identify a nitrogen-
170
enriched OA-factor, NOA, mainly composed of amino compounds formed from
171
industrial or marine emissions. A more local-SOA factor that is related to humic-like
172
substances, termed as HULIS OA, found in the Netherlands (Schlag et al., 2016) and
173
in Crete (Crippa et al., 2014). In Greece (Bougiatioti et al., 2014; Stavroulas et al., 2019;
174
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
6
Vasilakopoulou et al., 2023), in the Amazonian (De Sá et al., 2019) and often in Asia
175
(Zhang et al., 2015b; Chakraborty et al., 2015; Du et al., 2015) OOA factors directly
176
associated with biomass burning were found, that are processed from fresh biomass
177
burning emissions. Furthermore, OOA compounds that are verifiable only biogenically
178
oxygenated were also derived (Kostenidou et al., 2015).
179
Apart from the mass spectrum, OA types can also be distinguished by their oxygen
180
to carbon ratio (O:C), which is an indicator of photochemical aging. Primary organic
181
aerosol (POA) is fresh and has a lower oxygen content than OOA, therefore lower O:C
182
ratios. Yet, it sometimes has the same dominant m/z peaks. Some of the most commonly
183
resolved POA factors are the Hydrocarbon-like (HOA) and Biomass Burning (BBOA)
184
OA. HOA has spectra that are distinguished by clear hydrocarbon signatures,
185
dominated by the ion series CnH2n+1+ and CnH2n-1+ (Ng et al., 2010). HOA correlates
186
with fossil fuel combustion tracers like NOx, CO and elemental carbon (Lanz et al.,
187
2008; Tsimpidi et al., 2016), therefore, is very often observed to be traffic-related and
188
a rather dominant POA factor in urban areas (Crippa et al., 2014; Xu et al., 2015;
189
Budisulistiorini et al., 2016). On the other hand, BBOA typically originates from forest
190
and savanna fires as well as from anthropogenically induced agricultural fires (Hoesly
191
et al., 2018) and residential wood burning for heating. This makes the contribution of
192
BBOA to total OA highly episodic (Zhang et al., 2007) and seasonal, and in several
193
cases underestimated due to the rapid physicochemical transformation of these
194
emissions to OOA (Stavroulas et al., 2019; Vasilakopoulou et al., 2023). Typical tracers
195
to identify BBOA in the spectra are gas-phase acetonitrile, particle-phase levoglucosan
196
and potassium (K+) (Lanz et al., 2010; Crippa et al., 2014). However, its mass spectra
197
are also highly variable since they can be affected by different types of wood and
198
burning conditions (Crippa et al., 2014).
199
Furthermore, a coal combustion factor (CCOA) is often identified, which presents a
200
dominant contribution to POA during the heating season, mostly in Eastern Asia (Sun
201
et al., 2013; Zhang et al., 2014). In many cases, HOA shows remarkably similar spectral
202
patterns as CCOA, so that these two factors could not be separated and, instead, are
203
combined in a fossil fuel related OA factor (FFOA) (Sun et al., 2018; Xu et al., 2019).
204
Another relatively frequent primary type resolved by the factor analysis is the cooking
205
related OA (COA) (Mohr et al., 2012). Its spectral pattern is governed by OA from
206
fresh cooking emissions and, fittingly, the spectral profiles have a distinct diurnal cycle
207
which corresponds to typical (local) meal hours (Mohr et al., 2012; Sun et al., 2013;
208
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
7
Stavroulas et al., 2019). Occasionally, special types of COA are also resolved, including
209
coffee roastery OA (Timonen et al., 2013) and OA related to charbroiling (Lanz et al.,
210
2007).
211
212
2.1.2 AMS Dataset
213
Here, a collection of AMS and ACSM field campaign datasets during the period of
214
2000-2020 has been compiled. The dataset covers a wide range of environments and
215
seasons from almost every continental region worldwide (Figure 1), characterized by a
216
variety of atmospheric and climatological conditions as well as sources of pollutants.
217
The selected field campaigns lasted from at least one full week to several months.
218
Individual campaigns lasting more than one month are divided into shorter periods of
219
preferably only one month. All of these individual periods of campaign data (thus
220
covering a maximum of one month) are hereafter referred to as individual datasets.
221
222
The number of both PM1 and OA composition datasets found for each year is
223
increasing significantly for all regions through the years (Figure 2) due to the growing
224
popularity of the AMS devices and the continuous improvement of the analysis
225
Figure 1: Seasonal distribution of datasets per subcontinent. The colored bars
indicate the relative proportions by season. The numbers in the colored boxes
indicate the absolute number of field campaigns that occurred in each season.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
8
techniques. Especially during the second decade,
226
the number of field campaigns increase
227
drastically, supported by the use of ACSM
228
devices since 2010. The long-term campaigns in
229
South Africa (2010-2011; (Tiitta et al., 2014))
230
and the Southern Great Planes (2010- 2012;
231
(Parworth et al., 2015)) belong to the very first
232
where the ACSM has been utilized. Furthermore,
233
campaigns in regions downwind of urban
234
environments have gotten a growing attention
235
mostly after 2014, primarily in Europe. However,
236
usually these datasets are not factor analyzed and
237
lack information for the OA composition. It is
238
worth mentioning that the small number of
239
downwind datasets available can partially
240
attributed to the ambiguous definition of
241
downwind sites, which might have led instead to
242
the more conventional classifications of rural or
243
urban locations in some cases.
244
Overall, the compiled dataset includes PM1
245
aerosol composition from 744 AMS field
246
campaigns datasets at 169 observational sites
247
around the world, while factor analysis has been
248
used to estimate the OA composition in 398 cases
249
at 140 different observational sites (Table S1). The dataset includes an intermediate
250
level regional breakdown following the sixth assessment report of IPCC working group
251
III (Ipcc, 2022) as shown in Figure 3. The most represented subcontinents are Europe,
252
Eastern Asia and North America. Datasets from these three northern-hemisphere
253
continents are more or less evenly distributed over the seasons with only a little
254
imbalance for North America which is over-represented during summer (Figure 1). The
255
rest of the regions include a significantly lower number of datasets; therefore, the
256
seasonal distribution is often very uneven. As an example, 50% of the data over the
257
Asia-Pacific Developed region has been collected during spring. On the contrary, the
258
Figure 2: Total AMS (dark
red) and factor analysis
(green) datasets per year in
(a) rural, (b) urban-
downwind, and (c) urban
regions
(a) Rural/Remote
(b) urban/industrial Downwind
(c) Urban
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
9
changes between the wet and dry seasons are well represented over Africa where the
259
ACSM has been employed for year-long campaigns (Tiitta et al., 2014).
260
261
2.1.3 Observed PM1 Aerosol Composition
262
The PM1 aerosol composition derived from AMS field campaigns at 8 regions
263
around the world is depicted in Figure 4. The analysis of the AMS dataset reveals that
264
OA is the dominant component of PM1 in all continental regions. Campaign data from
265
tropical or subtropical regimes (e.g., Latin America and Southern/Southeast Asia) is
266
strongly affected by biomass burning and biogenic VOC emissions, illustrating
267
remarkably high OA fractions with regional means around 65% and a maximum of
268
92% in the Amazonian. However, OA concentration shares up to 90% are also found
269
over the Northern Hemisphere regions where the regional average OA contribution to
270
PM1 concentrations is around 50%. Overall, OA contributes between 17 - 92% (50%
271
on average) of total PM1. This agrees well with the ranges reported by Kanakidou et al.
272
(2005) (20%-90%) and Zhang et al. (2007) (18%-70% or 45% on average). Sulfate has
273
been the dominant inorganic compound in the aerosol composition in most regions
274
Figure 3: Worldwide distribution of AMS and ACSM datasets for the of period
2000 - 2020. The world map is colored according to the intermediate level regional
breakdown of the sixth assessment report of IPCC working group III (IPCC, 2022).
The rural (green), downwind (red) and urban (blue) campaign locations and the total
number of PM1 composition (and OA factor analysis in parenthesis) datasets for
each region are also shown.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
10
275
Figure 4: Bar chart plots depicting the distribution (violin) and the 25th, 50th and 75th
percentiles (box) of the mass concentration (in μg m–3) for the major PM1 aerosol
components, i.e., organic aerosol (green), sulfate (red), nitrate (blue), ammonium
(yellow), chloride (purple), and the total non-refractive PM1 (dark red). The 10th and
90th percentiles (whiskers) for each aerosol component are also shown. The number
of total months (m.) with AMS data and the number of campaigns (cmp.) is written
in small boxes under the violins.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
11
(Figure 4). The highest regional average share of sulfate is found over Asia-Pacific
276
Developed (37%) while the lowest over Europe (17%) where SO2 has been drastically
277
reduced due to strict air pollution mitigation strategies. Nitrate dominates over sulfate
278
over Europe and Eastern Asia. However, it is surprising that the PM1 inorganic
279
composition of North America is dominated by sulfate, even though similar mitigation
280
strategies have been enforced as in Europe. This might be due to an over-representation
281
of summer data in North America (Figure 1) which resulted in lower nitrate
282
concentrations since higher temperatures hinder the condensation of nitric acid in the
283
aerosol phase. At the same time, sulfate concentrations are higher during summer due
284
to the increased photochemical production of H2SO4. Overall, nitrate concentrations are
285
highest in winter in Europe and North America, accounting for roughly a quarter of
286
total PM1 (Figures S1 and S2). A similar proportion is observed in spring, although the
287
absolute concentration is lower. The lowest average nitrate concentrations and shares
288
occur in summer, when sulfate peaks and dominates the inorganic composition.
289
Although both sulfate and nitrate are generated through photochemical reactions, this
290
seasonal shift is due to nitric acid remaining in the gas phase at higher temperatures.
291
Additionally, the increased production of sulfuric acid reduces the amount of free
292
ammonia available for ammonium nitrate formation, further contributing to the summer
293
nitrate decline (Seinfeld and Pandis, 2006). Ammonium concentrations remain
294
relatively stable throughout the seasons, presenting similar shares of PM1 (Figures S1
295
and S2). However, in contrast to Europe and North America, sulfate concentrations in
296
East Asia are highest in winter, closely followed by summer (Figure S3). While
297
photochemical reactions still dominate during warmer, sunnier seasons, aqueous phase
298
reactions are more influential in East Asian winter, particularly under high relative
299
humidity (RH) and severe haze conditions. These factors are often present in Chinese
300
winters and likely explain this regional pattern (Zhang et al., 2015a; Zhou et al., 2020a).
301
Over the southern regions, ammonium follows sulfate in the inorganic aerosol
302
composition due to the high agricultural activities. Overall, the global average
303
contribution of the inorganic compounds to total PM1 concentration is 20%, 18%, and
304
11%, and 1% by sulfate, nitrate, ammonium, and chloride, respectively. However,
305
Zhang et al. (2007) reported much stronger contribution by sulfate (32%), less by nitrate
306
(10%), and similar values of ammonium (13%) and chloride (1%). Given that Zhang et
307
al. (2007) utilized AMS observations from the early 2000s, this is a first indication that
308
the inorganic aerosol composition has been altered during the last 20 years.
309
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
12
2.1.4 Observed PM1 Organic Aerosol Composition
310
HOA concentrations are observed to be higher over North America and Eastern Asia
311
in comparison to Europe (Figure 5). This could be explained by the significant influence
312
of traffic emissions on HOA in the vicinity of urban areas. While urban locations are
313
equally represented with rural sites in the dataset collection of North America and
314
Eastern Asia, in Europe, rural sites are immensely over-represented (3 times more than
315
urban sites), diminishing the importance of HOA. On the other side, the over-
316
representation of rural sites in the European dataset resulted in high concentrations of
317
BBOA which is found to be the dominant primary source of OA in the region (Lanz et
318
al., 2010). Here, BBOA originates mostly from domestic wood burning during the
319
colder seasons in central Europe, including the Alps, rather than from open biomass
320
burning. Even though a few campaigns took place in the European boreal forests, only
321
very few factor analyses have distinguished BBOA as an individual component. Thus,
322
the contribution of European boreal forests to total European BBOA is unfortunately
323
not clear yet. Similarly, biomass burning is an important source of OA in North
324
America and Eastern Asia (Rattanavaraha et al., 2017; Zhou et al., 2020b) but less
325
important than HOA (Figure 5). Biomass burning also presents an especially important
326
source in tropical and subtropical regions (i.e., South Asia and the Developing Pacific,
327
Africa, and Latin America and Caribbean) due to episodic wildfires and harvest related
328
burning (Budisulistiorini et al., 2018; Cash et al., 2021). Overall, the concentration
329
range of BBOA is very high since it varies a lot with season. However, it should be
330
emphasized that the availability of factor analysis datasets in equatorial and southern
331
hemisphere continents is very low and therefore, there is not enough data available for
332
statistically profound statements. The last primary type of OA, COA, is population
333
dependent and therefore is mainly found in urban areas and highly populated regions
334
(Zhou et al., 2020b). Cooking is a very constant and local source throughout the year
335
with low variability and high contributions over Eastern Asia, Europe, North America,
336
and South Asia and developing Pacific, especially in urban campaign sites.
337
OOA is unequivocally the dominant contributor to total OA with a mean share of
338
60% in urban and 75% in rural regions. Overall, the OOA contribution range from 19%
339
(urban minimum) to 99% (rural maximum). The extreme shares were both found during
340
European campaigns. The mean OOA share in Europe however lies roughly in the same
341
magnitude as the global mean (~70%). The dominant OOA subfactors resolved are L-
342
OOA and M-OOA, while the more aged M-OOA dominates in the OA composition of
343
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
13
all examined regions (~60% of total OOA). This agrees with the findings of Ng et al.
344
(2010), who stated that OOA component spectra become increasingly similar to each
345
other with atmospheric oxidation, indicating that ambient OA converges towards highly
346
aged M-OOA.
347
Figure 5: Bar chart plots depicting the distribution (violin) and the 25th, 50th and
75th percentiles (box) of the mass concentration (in μg m–3) for the major PM1 OA
components calculated from the collected factor analysis datasets, i.e., COA (olive
green), BBOA (orange), HOA (dark red), L-OOA (light turquoise), M-OOA (dark
turquoise), OOA (blue), and total OA (green). The 10th and 90th percentiles
(whiskers) for each aerosol component are also shown. The number of datasets (m.)
and the number of campaigns (cmp.) is written in small boxes under the violins.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
14
2.2 PM2.5 Dataset
348
Routine filter measurement PM2.5 data from large observational networks in East
349
Asia, Europe and North America is used. The filter samplers have three modules that
350
independently collect PM2.5 species on a Teflon, a nylon and a quartz filter. The aerosol
351
chemical composition is determined by further analysis of the filters in the laboratory
352
via ion chromatography (inorganic ions), thermal-optical analysis (OC and EC), and X-
353
ray fluorescence (XRF; trace elements) (Solomon et al., 2014). Potential difficulties
354
that could arise when comparing on-line AMS and ACSM PM1 composition to off-line
355
filter based PM2.5 composition, are discussed in section 5. The Environmental
356
Protection Agency (EPA) network includes 211 monitor sites primarily in urban areas
357
of North America. The data used here cover monthly averaged PM2.5 aerosol
358
component measurements during 2000-2018
359
(https://aqs.epa.gov/aqsweb/airdata/download_files.html). The Interagency
360
Monitoring of Protected Visual Environments (IMPROVE) network includes 198
361
monitoring sites that are representative of the regional haze conditions over North
362
America. IMPROVE samplers collect 24-hour samples, every three days. The data used
363
here cover monthly averaged PM2.5 aerosol component measurements during 2000-
364
2018 (http://views.cira.colostate.edu/fed/QueryWizard/Default.aspx). It is worth
365
mentioning that ammonium measurements by IMPROVE are only available until the
366
year 2006. The European Monitoring and Evaluation Programme (EMEP) network
367
monitors the long-range transmission of air pollutants in Europe and Eastern Eurasia
368
(Figure 6). This network includes 70 monitoring sites. The data used here cover
369
monthly averaged PM2.5 aerosol component measurements during 2000-2018
370
(https://www.emep.int/). Finally, the Acid Deposition Monitoring Network in East Asia
371
(EANET) network includes 39 (18 remote, 10 rural, 11 urban) air concentration monitor
372
sites in Eurasia, Eastern Asia, South-East Asia and Developing Pacific, and Asia-
373
Pacific Developed. The data used here cover monthly averaged PM2.5 aerosol
374
component measurements during 2001-2017 (https://www.eanet.asia/). The global
375
particulate matter network SPARTAN (Snider et al., 2015; Snider et al., 2016) includes
376
a global federation of ground-level PM2.5 monitors situated primarily in highly
377
populated regions around the word (i.e,, North America, Latin America and Caribbean,
378
Africa, Middle East, Southern Asia, Eastern Asia, South-Eastern Asia and Developing
379
Pacific) (Figure 6). The data used here covers monthly averaged PM2.5 aerosol
380
component measurements of sulfate, nitrate, ammonium and sodium during 2013-2019
381
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
15
(https://www.spartan-network.org/). Finally, PM2.5 aerosol component measurements
382
from individual observational field campaigns over Latin America and Caribbean,
383
Africa, Europe, Eastern Asia, and Asia-Pacific Developed reported as campaign
384
averages in the literature are used ( Wang et al., 2019; Radhi et al., 2010; Favez et al.,
385
2008; Mkoma, 2008; Mkoma et al., 2009; Weinstein et al., 2010; Celis et al., 2004;
386
Bourotte et al., 2007; Fuzzi et al., 2007; Mariani and De Mello, 2007; Martin et al.,
387
2010; Souza et al., 2010; Gioda et al., 2011; Molina et al., 2010; Molina et al., 2007;
388
Kuzu et al., 2020; Aggarwal and Kawamura, 2009; Batmunkh et al., 2011; Cho and
389
Park, 2013; Feng et al., 2006; Li et al., 2010; Pathak et al., 2011; Zhang et al., 2012;
390
Zhao et al., 2013).
391
392
2.2.1 PM2.5 Aerosol Composition
393
The PM2.5 aerosol composition derived from filter observations around the world is
394
depicted in Figure 7. OA is the dominant component of PM2.5 in most regions,
395
especially over regions affected by the tropical forests of the southern hemisphere (e.g.,
396
Latin America & Caribbean and Africa). Over the Northern Hemisphere, OA and EC
397
dominate the aerosol composition in Eastern Asia (54% and 22% of total PM2.5,
398
respectively) and contribute significantly to PM2.5 over Europe (30% and 5% of total
399
PM2.5, respectively). On the other hand, over North America, OA share is equally
400
important to sulfate over rural areas (28% of total PM2.5 each) and less important over
401
urban areas (24% versus 33% of sulfate). Indeed, sulfate is the most important inorganic
402
component of PM2.5 around the world (~50% of the inorganic PM2.5 mass on average)
403
followed by nitrate and ammonium (~20% each). This contradicts the results from AMS
404
Figure 6: Worldwide distribution of filter-based observations for the period of
2000-2020. The world map is colored following the intermediate level regional
breakdown of the sixth assessment report of IPCC working group III (IPCC, 2022).
The black dots correspond to the location of the monitor stations.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
16
405
Figure 7: Bar chart plots depicting the distribution (violin) and the 25th, 50th and 75th
percentiles (box) of the mass concentration (in μg m–3) for the major PM2.5 aerosol
components, i.e., sulfate (red), nitrate (blue), ammonium (yellow), sodium (pink), chloride
(purple), crustal ions (brown), organic aerosol (green), and elemental carbon (black). The 10th
and 90th percentiles (whiskers) for each aerosol component are also shown.
Africa
Asia-Pacific Developed
Eastern Asia
Eurasia
Europe
Latin America and Caribbean
Middle East
North America
Southeast Asia and Developing Pacific
Southern Asia
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
17
campaigns showing that ammonium nitrate surpasses ammonium sulfate in the aerosol
406
composition, especially over Europe and North America. However, filter measurements
407
are prone to negative sampling artifacts due to evaporation losses of the semivolatile
408
ammonium nitrate under warm and dry conditions (Ames and Malm, 2001), in contrast
409
to the nonvolatile sulfate aerosols (Docherty et al., 2011). The contribution of sulfate
410
to the measured inorganic PM2.5 aerosol composition is highest over Middle East, while
411
nitrate contributes significantly over Europe (Figure 7). The dominant inorganic ion
412
varies with the season (Figures S1-S3). Nitrate is most important in winter, accounting
413
for about a quarter of total PM2.5, while sulfate is the dominant PM2.5 component in
414
summer and spring. Over the 8 regions where all 7 components are measured, the
415
average contribution of each species to total PM2.5 concentration is 21%, 12%, 10%,
416
2%, 3%, and 40%, and 12% by sulfate, nitrate, ammonium, sodium, chloride, OA, and
417
EC respectively.
418
419
3 Model calculated Dataset
420
The ECHAM/MESSy Atmospheric Chemistry (EMAC) model is used, a numerical
421
chemistry and climate simulation system that includes sub-models describing
422
atmospheric processes from the troposphere to the mesosphere and their interaction
423
with oceans, land, and human influences (Jöckel et al., 2006). EMAC uses the Modular
424
Earth Submodel System (MESSy2) (Jöckel et al., 2010) to link the different sub-models
425
with an atmospheric dynamical core, being an updated version of the 5th generation
426
European Centre - Hamburg general circulation model (ECHAM5) (Roeckner et al.,
427
2006). The EMAC model has been extensively described and evaluated against
428
observations and satellite measurements and can be applied to a range of spatial
429
resolutions (Tsimpidi et al., 2016; Karydis et al., 2016; Janssen et al., 2017; Tsimpidi
430
et al., 2018; Pozzer et al., 2022; Milousis et al., 2024). The spectral resolution used in
431
this study is T63L31, corresponding to a horizontal grid resolution of 1.875ox1.875o
432
and 31 vertical layers extending to 10 hPa at about 25 km from the surface. The
433
presented model simulations cover the period 2000–2020.
434
435
3.1 Model configuration
436
In the model configuration used, EMAC calculates fields of gas phase species online
437
through the Module Efficiently Calculating the Chemistry of the Atmosphere
438
(MECCA) submodel (Sander et al., 2019). MECCA calculates the concentration of a
439
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
18
range of gases, including aerosol precursor species such as SO2, NH3, NOx, DMS,
440
H2SO4 and DMSO. The concentrations of the major oxidant species (OH, H2O2, NO3,
441
and O3) are also calculated online. The loss of gas phase species to the aerosol through
442
heterogeneous reactions (e.g., N2O5 to form HNO3) is treated using the
443
MECCA_KHET submodel (Jöckel et al., 2010). The aqueous phase oxidation of SO2
444
and the uptake of HNO3 and NH3 in cloud droplets are treated by the SCAV submodel
445
(Tost et al., 2006; Tost et al., 2007).
446
Aerosol microphysics and gas/aerosol partitioning are calculated by the Global
447
Modal-aerosol eXtension (GMXe) module (Pringle et al., 2010). The aerosol size
448
distribution is described by 7 interacting lognormal modes (4 hydrophilic and 3
449
hydrophobic modes). The modes cover the aerosol size spectrum (nucleation, Aitken,
450
accumulation and coarse). The aerosol composition within each mode is uniform with
451
size (internally mixed), though can vary between modes (externally mixed). The
452
removal of gas and aerosol species through dry deposition is calculated within the
453
DRYDEP submodel (Kerkweg et al., 2006) based on the big leaf approach. The
454
sedimentation of aerosols is calculated within the SEDI submodel (Kerkweg et al.,
455
2006) using a first order trapezoid scheme. Cloud properties and microphysics are
456
calculated by the CLOUD submodel utilizing the detailed two-moment microphysical
457
scheme of Lohmann and Ferrachat (2010) and considering a physically based treatment
458
of the processes of liquid (Karydis et al., 2017) and ice crystal (Bacer et al., 2018)
459
activation.
460
461
3.2 State of the art modules for the inorganic thermodynamics
462
The inorganic aerosol composition is computed with the ISORROPIA-lite
463
thermodynamic equilibrium model (Kakavas et al., 2022) as implemented in EMAC by
464
Milousis et al. (2024). ISORROPIA-lite is an accelerated and simplified version of the
465
widely used ISORROPIA-II aerosol thermodynamics model which calculates the
466
gas/liquid/solid equilibrium partitioning of the K+-Ca2+-Mg2+-NH4+-Na+-SO42--NO3--
467
Cl--H2O aerosol system. ISORROPIA-lite assumes that the aerosol is always in a
468
metastable state (i.e., it is composed only of a supersaturated aqueous phase) and uses
469
binary activity coefficients from precalculated look-up tables to minimize the
470
computational cost. ISORROPIA-lite provides almost identical results with
471
ISORROPIA-II in a metastable mode and reduces its computational cost by 35%
472
(Kakavas et al., 2022). The application of ISORROPIA-lite in EMAC improved the
473
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
19
computational speed of the model by 4% (Milousis et al., 2024). The assumption of
474
thermodynamic equilibrium is a good approximation for fine mode aerosols which can
475
reach equilibrium within the time frame of one model timestep. However, the
476
equilibrium timescale for large particles is typically larger than the timestep of the
477
model (Meng and Seinfeld, 1996). To account for kinetic limitations, the process of
478
gas/aerosol partitioning is calculated in two stages (Pringle et al., 2010). In the first
479
stage the amount of the gas phase species that are able to kinetically condense onto the
480
aerosol phase within the model timestep is calculated assuming diffusion limited
481
condensation (Vignati et al., 2004). In the second stage ISORROPIA-lite re-distributes
482
the mass between the gas and the aerosol phase assuming instant equilibrium between
483
the two phases.
484
485
3.3 State of the art module for organic aerosol
486
The organic aerosol composition and evolution in the atmosphere is calculated by
487
the ORACLE module (Tsimpidi et al., 2024). ORACLE is a computationally efficient
488
version of the ORACLE module (Tsimpidi et al., 2014) which simulates a wide variety
489
of semi-volatile organic products separating them into bins of logarithmically spaced
490
effective saturation concentrations. ORACLE minimizes the number of surrogate
491
species used to describe POA and SOA formation from different emission sources,
492
while at the same time it reproduces similar total organic aerosol mass concentrations
493
with the ORACLE module (Tsimpidi et al., 2024). In this application ORACLE uses
494
three surrogate species with effective saturation concentration at 298 K of C* = 10-2,
495
101, and 104 µg m-3 to cover the volatility range of LVOCs, SVOCs and IVOCs
496
emissions from biomass burning and other combustion sources (biofuel and fossil fuel
497
combustion, and other urban sources). These organic compounds are allowed to
498
partition between the gas and aerosol phases resulting in the formation of POA. The
499
least volatile fraction, at 10-2 µg m-3, describes the low volatility organics in the
500
atmosphere that are mostly in the particulate phase even in remote locations. The 10 µg
501
m-3 volatility bin describes the semivolatile organics in the atmosphere which partition
502
between the particle and gas phase at atmospheric conditions. Finally, even under
503
highly polluted conditions the majority of the material in the 104 µg m-3 volatility bin
504
will exist almost exclusively in the vapor phase. Photochemical reactions that modify
505
the volatility of the emitted organic compounds that remain in the gas phase are taken
506
into account and the oxidation products are simulated separately in the module to keep
507
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
20
track of the SOA formation from SVOC and IVOC emissions. LVOCs are not allowed
508
to participate in photochemical reactions since they are already in the lowest volatility
509
bin. A similar approach is followed for SOA formed from VOCs. In the this version of
510
ORACLE, it is assumed that the oxidation of the anthropogenic and biogenic VOC
511
species results in two products for each precursor distributed in two volatility bins with
512
effective saturation concentrations at 298 K equal to 1 and 103 µg m-3 at 298 K. Overall,
513
we have assumed that functionalization and fragmentation processes after any
514
subsequent photochemical aging as a result of the reaction with OH results in a net
515
average decrease of volatility by a factor of 103 for SOA produced by SVOC/IVOC and
516
anthropogenic VOC, without a net average change of volatility for SOA produced by
517
biogenic VOC (Tsimpidi et al., 2024). In total 18 organic compounds are simulated
518
explicitly, i.e., 9 in each of the gas and aerosol phases. Based on the saturation
519
Figure 8: Schematic of the VBS resolution and the formation procedure of POA and
SOA from LVOCs, SVOCs, IVOCs and VOCs emissions in ORACLE-lite. Red
indicates that the compound is in the vapor phase and blue in the particulate phase.
The circles correspond to primary organic material that can be emitted either in the
gas or in the aerosol phase. The triangles indicate the formation of SOA from
SVOCs by fuel combustion and biomass burning sources, while the squares show
SOA from IVOCs by fuel combustion and biomass burning sources, and the
diamonds the formation of SOA from anthropogenic and biogenic VOC sources.
The partitioning processes, the aging reactions and the names of the species used to
track all compounds are also shown.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
21
concentration of each organic compound, ORACLE calculates the partitioning between
520
the gas and particle phases by assuming bulk equilibrium and that all organic
521
compounds form a pseudo-ideal solution. A schematic overview of the ORACLE
522
module and the different aerosol types and chemical processes considered here is
523
provided in Figure 8. More details about ORACLE can be found in Tsimpidi et al.
524
(2024).
525
526
3.4 Emissions
527
Fuel combustion and agriculture related emissions are based on the high resolution
528
(0.1°×0.1°) Copernicus Atmosphere Monitoring Service global anthropogenic
529
emission inventory applied at monthly intervals, CAMSv4.2 (Granier et al., 2019). The
530
emission factors used for the distribution of traditional POA emissions from fuel
531
combustion and open biomass burning sources into the three volatility bins considered
532
by ORACLE are based on the work of Tsimpidi et al. (2024). These emission factors
533
account additionally for IVOC emissions that are not included in the original emission
534
inventories. We assume that the missing IVOC emissions from anthropogenic
535
combustion are 1.5 times the traditional OA emissions included in the inventory.
536
LVOCs and SVOCs are assumed to be emitted in the aerosol phase, while IVOCs are
537
emitted in the gas phase. Then, they are allowed to partition between the gas and particle
538
phase. Figure S4 shows the temporal evolution of anthropogenic emissions of inorganic
539
(SO2, NH3, NOx) and organic (LVOC, SVOC, IVOC, VOC) aerosol precursors over
540
the last 20 years, while Table S5 shows their decadal percentage change between the
541
2000s and 2010s. Open biomass burning emissions are calculated online based on the
542
dry matter burned from observations (Kaiser et al., 2012) and the fire type which affect
543
the emission factors for the different tracers (Akagi et al., 2011). Similar to POA
544
emissions from fuel combustion, POA from biomass burning is distributed to LVOC,
545
SVOC, and IVOC emissions, however, no additional IVOC emissions are assumed for
546
open biomass burning and therefore the sum for the biomass burning emission factors
547
is unity (Tsimpidi et al., 2016).
548
Biogenic emissions of isoprene and terpenes are calculated online using the Model
549
of Emissions of Gases and Aerosol from Nature (MEGANv2.04; Guenther et al., 2012)
550
with an average emission flux of 454 and 81.7 Tg yr-1, respectively. The natural
551
emissions of NH3 are based on the GEIA database (Bouwman et al., 1997) and include
552
excreta from domestic animals, wild animals, synthetic nitrogen fertilizers, oceans,
553
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
22
biomass burning, and emissions from soils under natural vegetation. NOx produced by
554
lightning is calculated online and distributed vertically based on the parameterization
555
of Price and Rind (1992). The emissions of NO from soils are calculated online based
556
on the algorithm of Yienger and Levy (1995). Eruptive and non-eruptive volcanic
557
degassing emissions of SO2 are based on the AEROCOM data set (Dentener et al.,
558
2006). The oceanic DMS emissions are calculated online by the AIRSEA submodel
559
(Pozzer et al., 2006). Emission fluxes of sea spray aerosols are calculated online (Guelle
560
et al., 2001) assuming a composition of 55% Cl-, 30.6% Na+, 7.7% SO42-, 3.7% Mg2+,
561
1.2% Ca2+, 1.1% K+ (Seinfeld and Pandis, 2006). The average global emission flux of
562
sea spray aerosols is 5910 Tg yr-1. Dust emission fluxes are calculated online by using
563
the meteorological fields calculated by the EMAC model (temperature, pressure,
564
relative humidity, soil moisture and the surface friction velocity) together with specific
565
input fields for soil properties (i.e., the geographical location of the dust sources, the
566
clay fraction of the soils, the rooting depth, and the monthly vegetation area index)
567
(Astitha et al., 2012). The average global emission flux of dust particles is 5684 Tg yr-
568
1. Emissions of individual crustal species (Ca2+, Mg2+, K+, Na+) are estimated as a
569
constant fraction of mineral dust emissions. This fraction is determined based on the
570
geological information that exists for the different dust source regions of the planet
571
(Karydis et al., 2016) and is applied online on the calculated mineral dust emission
572
fluxes based on the location of the grid cell (Klingmuller et al., 2018).
573
574
3.5 Model calculated aerosol composition
575
The EMAC simulation corroborates the findings based on filters and AMS
576
observations that OA is the dominant component of fine atmospheric aerosols in all
577
continental regions (Figure 9). The strongest OA contribution to total PM2.5 (more than
578
50%) is calculated over regions affected by biomass burning and biogenic VOC
579
emissions: the tropical forests and savannas of Africa, Latin America and Caribbean,
580
Southern Asia, and Southeast Asia and Developing Pacific, as well as the boreal forests
581
of Eurasia. Considerable OA shares (30-35%) are also calculated over the industrialized
582
regions of the Northern Hemisphere (i.e., North America, Europe, Eastern Asia) and
583
the Middle East, where strong fossil and biofuel combustion related sources are located.
584
OA shares peak in the summer over Europe and North America and in the winter over
585
East Asia (Figures S1-S3). EMAC is also able to reproduce the dominance of SOA
586
(resolved by the AMS as OOA) in all regions, even in regions with strong primary
587
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
23
emissions, e.g., close to tropical forests or industrial areas. However, EMAC cannot
588
reproduce the dominance of aged SOA in many cases (resolved as M-OOA by the
589
AMS), especially over Eastern Asia, revealing weaknesses in the oxidation scheme of
590
its organic module (e.g., including missing sources and formation pathways). POA has
591
the strongest contribution (more than 20%) over heavily forested areas (e.g., Africa and
592
Eurasia) and the lowest (less than 10%) over highly industrialized regions (e.g., Europe
593
Figure 9: Pie charts showing the simulated 20-year average chemical composition
of PM2.5 in the 10 regions considered according to WGIII AR6. The central world
map shows the simulated average near-surface concentration of PM2.5 (in μg m-3)
during the period 2000-2020.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
24
and Middle East). Regarding the inorganic aerosol composition, the EMAC model is
594
not always consistent with the filter-based observations since in many regions it reveals
595
that nitrate overpasses sulfate in the aerosol composition, which is also supported by
596
the AMS results. These regions are Europe, North America, and Eastern Asia, where
597
nitrate accounts for 25-30% of total PM2.5, with higher contributions in winter and
598
lower contributions in summer (Figures S1-S3). Sulfate becomes the dominant
599
inorganic aerosol component only during winter over North America (Figures S1-S3).
600
On the other side, sulfate contribution is stronger over the Middle East and Latin
601
America and Caribbean (~30%). Ammonium follows the spatial distribution of sulfate
602
and nitrate with high contributions to PM2.5 composition (~10-15%) over the highly
603
populated and agriculturally intensive regions of North America, Europe, Eastern Asia
604
and Southern Asia. Mineral dust is simulated to be a significant natural contributor to
605
aerosol composition in some regions. Here we only focus on the chemically active
606
components of mineral dust, which are the crustal cations of calcium, potassium,
607
sodium, and magnesium. Their total share to PM2.5 composition is around 15% in
608
regions affected by desert emissions (e.g., Africa, Middle East, Eastern Asia) while in
609
other areas their contribution is limited (~ 1%). Finally, sodium and chloride from sea
610
salt emissions are found to be high over regions with long coastlines per land area. Most
611
notably, chloride consists of 8% of the total PM2.5 over the Asia Pacific Developed
612
region, while sodium is the dominant inorganic component in the same region with a
613
share of 8.5%.
614
615
4 In depth model Evaluation
616
4.1 Sulfate
617
The EMAC performance for sulfate is best over North America, where the model
618
tends to underpredict its concentrations with a MB of -0.45 μg m-3 (Figure 10a). The
619
model performs better over rural regions with very low NMB (-8%) and worst over
620
urban locations (NMB=-40%). This performance can be attributed to the low spatial
621
resolution used and to possible errors in the assumed injection height of SO2 (Yang et
622
al., 2019) which can affect sulfate concentrations close to sources. Furthermore, EMAC
623
tends to overestimate sulfate over the Midwest, while underestimating its
624
concentrations over the Eastern states (Figure 10). The coarse resolution of the model
625
cannot reproduce the orography of the mountainous Midwest and therefore
626
overestimates the sulfate concentrations at high altitude sites. On the other hand, due to
627
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
25
its coarse resolution, it underestimates the sulfate concentrations over the urban areas
628
of the densely populated Eastern states. Therefore, the model underpredicts
629
observations over the Eastern US, where sulfate concentrations are high, and
630
overpredicts observations over the Midwest, where sulfate concentrations are low. As
631
a result, the model produces a quite narrow range of concentrations (i.e., 0.3 - 2.5 µg
632
m-3) over the North America in contrast to the AMS observations which cover almost
633
three orders of magnitude, ranging from 0.1 to 10 µg m-3. The seasonal pattern of both
634
measured and observed data shows clear differences between summer and winter. The
635
model calculates the highest sulfate concentrations in autumn, in contrast to the AMS
636
observations which show a peak in summer. The lowest sulfate concentrations are
637
observed in winter which are well captured by the model at most sites (Figure 10a).
638
Figure 10: Deviations (in %) between EMAC results and the AMS and ACSM
datasets over the period 2000 – 2020 (top). Negative values (blue colors) correspond
to underprediction of sulfate concentrations by the model. Scatter plots comparing
model results for PM1 sulfate concentrations (in μg m-3) with AMS and ASCM
observations (bottom) over (a) North America, (b) Europe, and (c) Eastern Asia.
Each point represents the data set mean and is colored based on the season of the
field campaign. Also shown are the 1:1, 2:1, and 1:2 lines.
10
2
10
2
10
2
(a) North America
(b) Europe
(c) Eastern Asia
Deviation (%)
Sulfate
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
26
In Europe, the model underpredicts sulfate in all types of environments and all
639
seasons by about 40% due to errors in emissions and an underestimation of the
640
oxidation capacity of the atmosphere (Emep, 2021). However, a few overpredictions
641
are calculated over Italy and Greece. Around 65% of the simulated sulfate
642
concentrations over Europe are within a factor of 2 compared to measurements (Figure
643
10b). The performance of the model does not exhibit any clear seasonal pattern except
644
a slight tendency towards higher underpredictions during summer when the observed
645
sulfate concentrations are the highest of the year. Over Asia, sulfate concentrations are
646
significantly higher than over Europe and North America, however, the performance of
647
the model is similar. Sulfate is underpredicted most of the time (Figure 10c, Table 1).
648
The model performs better over rural locations (NME=-38%) and worst over urban
649
areas (50%). Furthermore, while the model underpredicts sulfate concentrations during
650
all seasons, its performance is worst in winter when sulfate exhibits its annual peak
651
concentrations (Figure 10c) due to its multiphase formation during haze events, a
652
pathway not accurately resolved by the model. Furthermore, similar to North America,
653
the concentration range of the simulated sulfate over Eastern Asia is much narrower
654
than the observed, covering little more than one order of magnitude compared to two
655
orders of magnitude reported by the AMS. Over the tropical and subtropical regions,
656
sulfate is underestimated again, mostly over the Asian regions (NME ≈ -45%) and less
657
over Africa and Latin America and Caribbean (NME ≈ -30%) (Table S2, Figure S5).
658
Table 1: Statistical evaluation of EMAC PM1 sulphate concentrations against AMS
and ACSM datasets over Europe, North America, and Eastern Asia for the period of
2000–2020.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
27
4.2 Nitrate
659
The model is able to capture the observed average nitrate concentrations over the
660
different regions and seasons with very low NMB (below 10%). However, the NME is
661
high over all regions (40-80%) indicating that the discrepancy between model results
662
and observations is highly scattered and not systematically biased (Table 2). The
663
accurate prediction of nitrate concentrations is rather complex. Nitrate is typically
664
formed in areas characterized by high ammonia and nitric acid concentrations and low
665
sulfate concentrations. At the same time, the thermodynamic equilibrium of ammonium
666
nitrate varies several orders of magnitude under typical atmospheric conditions
667
(Seinfeld and Pandis, 2006). This variation causes significant challenges in the
668
calculation of nitrate concentrations since small errors in RH and T can shift the
669
equilibrium of nitric acid to the gas or the aerosol phase. Therefore, even though the
670
scatter is not negligible, it is encouraging that the EMAC model seems to perform
671
surprising well under diverse environments and atmospheric conditions (Figure 11).
672
The scatter is more intense over North America (NME=88%), especially during the
673
summer season where the occurrence of high temperatures and the semi-volatile nature
674
of NH4NO3 hinder the model’s ability to capture the observations accurately (Figure
675
11a). However, the model is still able to capture the seasonality of nitrate concentrations
676
well with the highest concentrations calculated during the periods with the lowest
677
temperatures (i.e., winter), when almost all the nitric acid that is available is transferred
678
to the particulate phase.
679
Over Europe, despite some widely dispersed points, the majority of datapoints (70%)
680
lie within a factor of two compared to observations (Figure 11b). Similar to North
681
America, the seasonality is very well captured, and the model predictions are mostly
682
scattered during the warmer seasons. However, the overall performance is better here
683
with NMB = -4% and NME = 53%. Over Eastern Asia, the overestimation appears to
684
be more systematic, especially during the summer and fall (Figure 11c). However, with
685
an overall NMB of 7.7%, the performance can still be considered very good (Table 2).
686
Nitrate levels are significantly overestimated by the model, especially over the west
687
coast of South Korea and the Chinese inlands (Figure 11). However, Eastern China and
688
especially the coastal regions are well described by the model. The contribution of sea
689
salt to nitrate formation is important in these coastal regions due to their proximity to
690
the Pacific Ocean (Bian et al., 2017). Therefore, the overestimation of nitrate levels on
691
the west coast of Korea, in contrast to the well captured east coast, could be caused by
692
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
28
the dominant west-east winds in the Yellow Sea simulated by the model, leading to an
693
overestimation of the sea salt content that can contribute to nitrate formation. Over the
694
tropical and subtropical regions, the discrepancies between the simulated and observed
695
nitrate concentrations are less dispersed with a tendency towards overprediction by the
696
model in most regions (Figure S5; Table S2). Over Latin America and the Caribbean,
697
the model underpredicts nitrate (NMB = -50%) except for a few strong overpredictions,
698
mostly during the wet season, suggesting possible errors in simulated wet deposition
699
(Figure S5). On the other hand, over Africa, the model overpredicts nitrate during the
700
dry season, especially over Welegund, an observation site downwind of Johannesburg.
701
Nitrate is strongly overpredicted over the Asia Pacific Developed region, especially
702
over the industrialized regions of Japan and Australia. On the contrary, the model
703
performance for nitrate is good over the Southeast Asia and the Developing Pacific
704
Figure 11: Deviations (in %) between EMAC results and the AMS and ACSM
datasets over the period 2000 – 2020 (top). Negative values (blue colors) correspond
to underprediction of nitrate concentrations by the model. Scatter plots comparing
model results for PM1 nitrate concentrations (in μg m-3) with AMS and ASCM
observations (bottom) over (a) North America, (b) Europe, and (c) Eastern Asia.
Each point represents the data set mean and is colored based on the season of the
field campaign. Also shown are the 1:1, 2:1, and 1:2 lines.
Figure 11
(a) North America
(b) Europe
(c) Eastern Asia
10
2
10
2
10
2
Deviation (%)
Nitrate
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
29
(NMB = -3%) with few random over- and underpredictions during the monsoon and
705
the transition periods towards that season.
706
707
4.3 Ammonium
708
EMAC tends to underpredict ammonium over the three main subcontinents of the
709
Northern Hemisphere, however, its performance is considered satisfactory with
710
relatively low bias and scatter (Table 3). The model evaluation exhibits a large scatter
711
only over North America (NME = 63%), where 50% of the comparison sites are beyond
712
the factor 2 intervals (Figure 12a). Ammonium tends to be overestimated during autumn
713
and underestimated during the rest of the seasons; especially during the summer (Figure
714
12a). Over Europe, the model exhibits its best performance with low NMB (-9%). The
715
average deviation from the observations is also relatively low (Figure 12) and 75% of
716
the model results diverge less than a factor of two from measurements. Surprisingly,
717
the model performance is best over the Benelux region (Figure 12) where NH3
718
emissions are the highest over Europe. While the good model performance for
719
ammonium over Europe indicates an accurate emission inventory for agricultural and
720
livestock NH3, the overprediction of nitrate and underprediction of sulfate suggest that
721
the model overpredicts the fraction of ammonium that exists as ammonium nitrate
722
(instead of ammonium sulfate). Over Asia, the model strongly underestimates
723
ammonium (NMB = -30%), especially over Eastern China (Figure 12). While this
724
Table 2: Statistical evaluation of EMAC PM1 nitrate concentrations against AMS
and ACSM datasets over Europe, North America, and Eastern Asia for the period of
2000–2020.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
30
underestimation can be partially attributed to sulfate underpredictions, the simultaneous
725
overestimation of nitrate over the same areas indicates errors in the NH3 emission
726
inventory. On the other hand, ammonium is overpredicted close to the deserts of Inland
727
China (e.g., over Tibet) and over South Korea (Figure 12). Over the Tropics and the
728
southern continents, ammonium is underestimated to a higher extent than in the
729
northern continents (with NMB from -40 to -60%). The main problem in model
730
performance is over Asia Pacific Developed and Africa, where the model predicts low
731
ammonium shares that are not supported by AMS observations (Figure S2). On the
732
other hand, EMAC has the largest underprediction and highest NMB over Latin
733
America. Nevertheless, here and over South Asia, EMAC and AMS agree that
734
ammonium has the smallest fraction of PM1. Overall, deviations in ammonium can be
735
traced back to global livestock emission inventory uncertainties as criticized by Hoesly
736
et al. (2018).
737
Figure 12: Deviations (in %) between EMAC results and the AMS and ACSM
datasets over the period 2000 – 2020 (top). Negative values (blue colors) correspond
to underprediction of ammonium concentrations by the model. Scatter plots
comparing model results for PM1 ammonium concentrations (in μg m-3) with AMS
and ASCM observations (bottom) over (a) North America, (b) Europe, and (c)
Eastern Asia. Each point represents the data set mean and is colored based on the
season of the field campaign. Also shown are the 1:1, 2:1, and 1:2 lines.
10
2
10
2
10
2
Deviation (%)
(a) North America
(b) Europe
(c) Eastern Asia
Ammonium
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
31
738
739
4.4 Organic aerosol
740
The model performance for total OA concentration varies significantly between the
741
three continents. Over North America, the simulated mean OA represents well the
742
observed OA by AMS (NMB = -4%). However, the comparison exhibits a significant
743
scatter (NME = 64%) since the model tends to overpredict OA over rural locations
744
(NMB = 37%) and underpredict it over and downwind of urban sites (NMB = -28%).
745
The model roughly captures the seasonality of OA concentrations over North America,
746
with high OA concentrations in summer and autumn and lower concentrations in spring
747
and winter. OA concentrations peak during summer due to enhanced biogenic VOC
748
emissions and photochemistry (Goldstein and Galbally, 2007; Tsimpidi et al., 2016),
749
however, EMAC tends to overpredict some low OA concentrations measured by AMS
750
over a few rural locations during summertime (Figure 13a). Over Europe, the model
751
tends to underestimate OA during all seasons, except summer (Figure 13b). The model
752
performance is worst during wintertime, where sources from biomass burning,
753
particularly by domestic wood burning, and their dark oxidation have been recently
754
identified as a major source of model bias over Europe during wintertime (Tsimpidi et
755
al., 2016; Kodros et al., 2020). This also affects the simulated OA seasonality over
756
Europe where the model estimates higher OA concentrations during summer over all
757
types of environments, while the AMS observations reveal that this is true only over
758
Table 3: Statistical evaluation of EMAC PM1 ammonium concentrations against
AMS and ACSM datasets over Europe, North America, and Eastern Asia for the
period of 2000–2020.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
32
rural locations. According to AMS, over and downwind of urban areas, OA
759
concentrations peak during wintertime. Over Eastern Asia, the model exhibits its best
760
performance with relatively low bias (NMB = -29%) and scatter (NME = 49%). In
761
contrast to Europe, the wintertime OA is well captured by the model even over urban
762
locations (Table 4). The model has excellent performance over rural and urban-
763
downwind locations with 75% of the datapoints lying within a factor of two compared
764
to observations. However, as it is typical for every global model (Tsigaridis et al.,
765
2014), the model fails to reproduce some of the high OA concentrations observed over
766
large urban centers due to its limited spatial resolution. Over the rest of the continental
767
regions, the overall performance of the model is satisfying for OA. EMAC tends to
768
underpredict OA over the tropical regions of South Asia and Developing Pacific and
769
over the more urbanized regions of the Asia Pacific Developed, without any clear
770
seasonal pattern (Figure S5). In contrast, simulated OA are overestimated over Africa,
771
Figure 13: Deviations (in %) between EMAC results and the AMS and ACSM datasets
over the period 2000 – 2020 (top). Negative values (blue colors) correspond to
underprediction of organic aerosol concentrations by the model. Scatter plots comparing
model results for PM1 organic aerosol concentrations (in μg m-3) with AMS and ASCM
observations (bottom) over (a) North America, (b) Europe, and (c) Eastern Asia. Each
point represents the data set mean and is colored based on the season of the field
campaign. Also shown are the 1:1, 2:1, and 1:2 lines.
10
2
10
2
(a) North America
(c) Eastern Asia
Deviation (%)
10
2
(b) Europe
Organic Aerosol
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
33
mostly during the dry season. Over Latin America and Caribbean, the evaluation
772
datapoints are more scattered with a few significant overestimations during the
773
Amazonian wet season and underestimations during the dry season.
774
775
4.4.1 POA
776
The simulated POA concentrations are compared with the sum of the AMS HOA
777
and BBOA concentrations. POA concentrations are mostly underestimated by the
778
model over North America and Europe (NMB ≈ -45%) and significantly overestimated
779
over Eastern Asia (NMB = 98%). In North American rural regions, POA simulated
780
concentrations are highest during spring and winter and lowest during fall, consistent
781
with the observed POA levels. However, during summer, most of the observed data is
782
underestimated by the model (Figure 14). Over urban locations, POA is more severely
783
underestimated (NMB = -68%) due to the coarse spatial resolution of the model and the
784
evaporation of organic compounds upon emission. POA concentrations are also
785
underestimated over European urban regions (NMB = -52%), however, to a lesser
786
extent than over North America. Over rural locations, the model performance is
787
scattered during all seasons with a few cases of strong over and under predictions (NME
788
= 62%). Over Eastern Asia, a pronounced overestimation during winter is striking,
789
especially over mega-city clusters (NMB = 106%; Table 5) such as around Hong Kong
790
and Shanghai. This discrepancy can be related to overestimations in the emission
791
inventory (e.g., not including the emission reductions in the frame of the Chinese
792
Table 4: Statistical evaluation of EMAC PM1 OA concentrations against AMS and
ACSM datasets over Europe, North America, and Eastern Asia during 2000–2020.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
34
control action plans) but also to the overestimated partition of the freshly emitted
793
SVOCs to the aerosol phase during the low winter temperatures. Tsimpidi et al. (2016)
794
has also reported POA overestimations over Eastern Asia due to too high simulated
795
bbPOA transported from the surrounding boreal forests. Since in ORACLE POA do
796
not participate in aqueous phase and other heterogeneous reactions, they do not convert
797
to SOA via these pathways, which can explain part of the positive model bias during
798
winter.
799
800
Table 5: Statistical evaluation of EMAC PM1 POA concentrations against AMS and
ACSM datasets over Europe, North America, and Eastern Asia during 2000–2020.
Table 6: Statistical evaluation of EMAC PM1 SOA concentrations against AMS and
ACSM datasets over Europe, North America, and Eastern Asia during 2000–2020.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
35
801
4.4.2 SOA
802
The model simulated OOA concentrations over North America are in very good
803
agreement with the OOA derived by the PMF analysis of the AMS observations (NMB
804
= 4.5%). The model performs well over both urban and rural areas and during all
805
seasons, except winter when it tends to underpredict the AMS-OOA estimations (Table
806
6; Figure 14c). L-OOA concentrations are reproduced by the model particularly well
807
(Figure S6a), however, M-OOA concentrations are slightly underestimated during
808
spring and fall and severely underpredicted during winter (Figure S6d). Similarly, the
809
model performance for all OOA types over Europe is best during summer and worst
810
during winter when it underpredicts the AMS estimations, especially for the M-OOA
811
(Figure S6e). During summer, the high temperatures enhance the biogenic VOC
812
emissions from vegetation and, more importantly, the more abundant solar radiation
813
increase the transformation of gas phase organic compounds through photochemical
814
processing into particulate OOA (Seco et al., 2011; Xu et al., 2017; Tsimpidi et al.,
815
Figure 14: Scatter plots comparing model results for PM1 primary organic aerosol (a-
c) and secondary organic aerosol (d-f) concentrations (in μg m-3) with AMS and ASCM
observations of HOA+BOA and OOA, respectively, over North America (a, d), Europe
(b, e), and Eastern Asia (c, f). Each point represents the data set mean and is colored
based on the season of the field campaign. Also shown are the 1:1, 2:1, and 1:2 lines.
EMAC SOA (μg m-3)
10
2
(a) North America
(d) North America
(e) Europe
(b) Europe
(f) Eastern Asia
(c) Eastern Asia
10
2
10
2
10
2
10
2
10
2
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
36
2016). The model performance during summer suggests that the model can accurately
816
represent this process. In winter, however, photochemical processing has lower impact
817
on OOA formation and evolution (Xu et al., 2017). Therefore, in seasons with
818
decreasing temperatures and/or photochemical activity, the model performance is
819
worsening, strongly suggesting that other processes become increasingly more
820
important. Missing SOA formation processes are related to heterogeneous reactions
821
like oligomerization or aqueous phase processing (Hallquist et al., 2009; Tsimpidi et
822
al., 2016). Under high RH, aqueous phase processing can rapidly result in highly
823
oxidized OOA (i.e., M-OOA with high oxygen to carbon ratio, O:C), while the impacts
824
on fresher, less oxygenated OOA (i.e., L-OOA) are minor. For the latter, photochemical
825
aging processes under low RH are more important (Xu et al., 2017). Such processes
826
occur during all seasons, however, the meteorological conditions during winter favor
827
the formation of M-OOA from aqueous phase chemistry against the L-OOA formation
828
from gas-phase photochemical oxidation processes (Xu et al., 2017; Mortier et al.,
829
2020; Pozzer et al., 2022). Therefore, this missing formation pathway becomes
830
gradually more important from spring and fall to winter. Additionally, recent studies
831
have identified high production of SOA during wintertime which can be attributed to
832
the rapid oxidation of biomass burning OA by the NO3 radical during nighttime (Kodros
833
et al., 2020; Paglione et al., 2020; Liu, 2024). Since residential heating from woodstoves
834
is not included in the model and ORACLE includes only the predominant
835
photochemical processing of BBOA by OH, a non-consideration of dark chemical
836
processing of BBOA can lead to substantial underprediction of OOA during the cold
837
seasons. Over Eastern Asia, OOA is underestimated even during summer (Figure 14f),
838
mainly due to the underestimation of M-OOA since L-OOA is relatively well
839
represented during all seasons (Figure S6). In fact, Eastern Asia is characterized by high
840
RH even during summer, corroborating our hypothesis that aqueous phase processes
841
may be an important missing piece in simulating the SOA formation. Recent studies
842
have provided strong evidence that the uptake of water-soluble gas-phase oxidation
843
products (even small carbonyls like formaldehyde and acetic acid) to the aqueous phase
844
and their subsequent oxidation and oligomerization can lead to significant increases of
845
SOA mass during pollution events (Gkatzelis et al., 2021). Overall, EMAC performs
846
best over the Eastern Asian rural areas during summer and spring and worst in the
847
vicinity of urban regions during fall and winter. Especially during wintertime, while the
848
model simulates well the total OA, it significantly overpredicts POA (Figure 14c) and
849
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
37
at the same time underpredicts SOA (especially M-OOA). This disagreement can be
850
due to an overestimation of the POA formation from the emitted SVOC species, but
851
also due to a missing mechanism that can significantly transform POA to SOA in the
852
aerosol phase during winter.
853
854
5 Aerosol Trends
855
Here, the simulated 20-year global aerosol composition trends of fine aerosols are
856
presented and discussed against trends calculated based on observational data. For this,
857
it is vital to have data well distributed spatially and measured consistently in a
858
comparable way at all observational sites within a region (Tørseth et al., 2012; Hand et
859
al., 2011). These conditions, unfortunately, cannot be satisfied by the available PM1
860
datasets (Figure 2). Instead, here we summarize the available observational data from
861
each region for the 1st versus the 2nd decade of the examined period. This allows a rough
862
statistical comparison between the two decades and can give insights on the overall
863
tendency of the observed aerosol composition trends for each region. These trends are
864
compared against the simulated PM1 trends based on the respective spatiotemporal
865
model data, as well as based on all the available model data for the entire model domain
866
over the complete 20-year period (Figure 15). As the spatial and temporal AMS
867
campaign distribution is much higher for regions in the northern than the southern
868
hemisphere, only PM1 data of the former is plotted here. PM2.5 data from the large
869
monitoring networks is also used to calculate the aerosol composition trends within the
870
regions of North America, Europe, and Eastern Asia. These networks present
871
cooperative measurement efforts that, among others, provide routinely filter based
872
measured data of aerosol composition. Even though not every element is always
873
measured at all sites and despite data gaps for some places, collectively, the networks’
874
datasets provide the consistency and duration requirements mentioned above. The
875
calculated trends are compared against PM2.5 simulated results based on the respective
876
spatiotemporal model data. It is worth noting that a comparison of filter PM2.5 to AMS
877
detected PM1 is not completely straightforward. First, as seen in Sections 2.1.3 and
878
2.2.1, there are expected compositional differences between the two size ranges,
879
especially in polluted regions (Sun et al., 2020; Petit et al., 2015). Second, instrumental
880
differences of the real-time on-line AMS (Decarlo et al., 2006) versus the non-real-time
881
off-line filter instruments (Docherty et al., 2011; Hand et al., 2011) can manipulate the
882
measurements in different ways, as discussed in the following sections.
883
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
38
884
5.1 Europe
885
Figure 16 depicts the interannual and seasonal concentration change of filter
886
measured PM2.5 components with a polynomial fitted trendline, in comparison to the
887
corresponding concentration trends as calculated by the EMAC model. Both
888
observations and the model reveal a concentration decrease for the three main inorganic
889
components of PM2.5, following the emission reductions during the last 20 years.
890
Sulfate concentrations have decreased drastically during the last decade (i.e., -46%
891
compared to 2000s). However, the simulated reduction is not so apparent mainly
892
because filter observations show much higher concentrations during the first half of the
893
2000s than model simulations. Until 2005, observed sulfate concentrations rose during
894
all seasons, however, they rapidly dropped under the 2000 levels in the following years.
895
The average decline rate is -0.15 µg m-3 yr-1, compared to the simulated rate of -0.02
896
µg m-3 yr-1. AMS measurements (Figure 17) corroborate the findings of filter
897
observations, revealing a drastic decrease in PM1 sulfate concentrations during the
898
decade of 2010s (i.e., -18% compared to 2000s). EMAC underestimates European PM1
899
(b) Nitrate
(d) Organic Aerosol
(c) Ammonium
(a) Sulfate
Figure 15: Simulated decadal change in (a) sulfate, (b) nitrate, (c) ammonium, and (d)
anthropogenic organic aerosol concentrations between the 2000s and 2010s.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
39
900
Figure 16: Temporal evolution of the observed (a, c, e, and g subplots on the left)
and simulated (b, d, f, h subplots on the right) concentrations of PM2.5 sulfate (a, b),
nitrate (c, d), ammonium, (e, f), and organic aerosol (g, h) during the period 2000–
2018 over Europe. Black lines show the annual trend while the dark blue, light blue,
orange, and red lines represent the seasonal trends during winter, spring, summer,
and autumn. Ranges represent the 1σ SD (standard deviation).
(b) Sulfate
(d) Nitrate
(f) Ammonium
(h) Organic aerosol
EMAC simulations over Europe
EMEP observations over Europe
(a) Sulfate
(c) Nitrate
(e) Ammonium
(g) Organic aerosol
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
40
sulfate (Figure 10b) resulting in a less pronounced negative trend in its concentrations
901
(i.e., -11%) since the model underestimation is more pronounced during the 2000s. The
902
average simulated decadal change in sulfate PM1 concentrations for the entire European
903
domain is -15% (Figure 15). Similar to sulfate, filter measured nitrate concentrations
904
rose until 2005 (except during summer where they remain in low levels) and then
905
quickly dropped again with an average rate of -0.09 µg m-3 yr-1 (Figure 16c). The high
906
observed nitrate concentrations during the first half of the 2000s results in an average
907
decrease of -35% between the two decades. On the other hand, the calculated change
908
of AMS-PM1 nitrate concentrations between the 2000s and the 2010s is -10 %, which
909
is similar to the simulated drop of -12%. However, it is worth mentioning that the model
910
significantly overestimates the nitrate concentrations both in comparison to AMS
911
measurements (Figure 11b) and to filter observations, especially during summer
912
(Milousis et al., 2024). The analysis of model simulation and observations (both by
913
AMS and filters) reveal that ammonium concentrations exhibit strong reductions
914
between the decades of 2000s and 2010s. The average concentration reduction between
915
the two decades is -21% based on the AMS observations, -13% based on the EMAC
916
results for PM1 (or -16% for the entire European domain), and -56% for the PM2.5 filter
917
observations. Therefore, the reduction of ammonium is much stronger based on the
918
filter observations (i.e., -0.1 µg m-3 yr-1) than based on AMS measurements or modeled
919
data (i.e., -0.02 µg m-3 yr-1). It is worth emphasizing that ammonium is clearly
920
declining, even though NH3 emissions have only been slightly reduced. This apparent
921
inconsistency can be attributed to the strong reductions of SO2 and NOx. This results in
922
reduced availability of acids (i.e., H2SO4 and HNO3) preventing the formation of
923
ammonium and allowing the NH3 to reside in the gas phase. This is also verified by
924
NH3 observations, where no significant trends, and even statistical increases, have been
925
observed despite reported reductions in NH3 emissions (Fagerli et al., 2016; Liu et al.,
926
2024).
927
The downward trend of organic aerosol calculated based on the filter observations
928
(-0.04 µg m-3 yr-1) is milder than that of inorganic components and differs between
929
seasons (Figure 16e). During summer, there is no clear trend observed, while in winter,
930
OC concentration soars after 2003 until 2005 when it starts to gradually drop until it
931
reaches the concentration levels of the other seasons during the second half of 2010s.
932
Irregularities in the early first decade could be owed to a lack of OC data (Fagerli et al.,
933
2016). OC data during spring and autumn shows a mild downward trend after 2005 as
934
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
41
well. Overall, the average difference of OC concentration between the two decades is -
935
22%. However, model data does not corroborate this reduction; on the opposite a slight
936
increase is calculated by the model during the last five years (Figure 16h). This agrees
937
with the AMS observations which predict a positive OA trend (Figure 17d) with an
938
average increase of +0.44 µg m-3 (or 10%) from the first to the second decade. Despite
939
the prominent underestimation of PM1 OA by the model, the simulated PM1 OA trend
940
is also positive with an average decadal increase of +0.55 µg m-3 (or 31%). Overall,
941
inconsistencies between AMS and filter observations can be attributed to instrumental
942
differences. First, is the size of particulate matter observed which is 2.5 μm for filters
943
and up to 1 μm for the AMS. The size distribution of OA can be affected by multiple
944
factors, including RH and chemical composition. Sun et al. (2020) have shown that the
945
PM1/PM2.5 SOA ratio increases when RH is below 60% and the contribution of
946
inorganic components in the aerosol decreases. This increase is related to differences
947
in aerosol water content due to changes in aerosol hygroscopicity and phase state.
948
Simulated data reveals that the frequency of RH dropping below 60% over European
949
locations has marginally increased (by 1%) during the decade of 2010s. However, the
950
(d) Organic
Aerosol
(b) Nitrate
(a) Sulfate
(c) Ammonium
Figure 17: Decadal PM1 concentration trends in Europe expressed by the bar plots
of the mass concentration (in μg m–3) for (a) sulfate, (b) nitrate, (c) ammonium, and
(d) OA during the periods 2000 - 2010 (left) and 2011 - 2020 (right) as calculated
from the AMS observational dataset (dark colors) and the corresponding simulation
values (light colors). The upper and lower whiskers range from 10-90%, the
quartiles from 25-75% of the dataset. The black line is the median.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
42
drastic reduction of sulfate and nitrate levels during the same period can explain the
951
increase in PM1 OA, as measured by the AMS, as opposed to the decrease in PM2.5 OA
952
observed by filters. Another important difference between the AMS and the filters is
953
that the latter, in contrast to AMS, only detects the carbonaceous fraction (OC) of OA.
954
Then, the ratio of the total organic mass (OM) to OC must be considered when
955
comparing the measured OC to AMS or simulated OA. However, the OM:OC is
956
broadly debated in literature. OM:OC is closely correlated to the oxygen to carbo ratio
957
(O:C) and therefore it is dependent on the chemical aging degree of OA. For the range
958
of SOA found in the atmosphere, Aiken et al. (2008) calculated the OM/OC ratios
959
between 1.9 to 2.5. Similarly, the ratio for POA varies depending on the source and
960
composition between 1.3 and 1.5 (Aiken et al., 2008). As the EMEP stations in Europe
961
are a mix of urban and rural locations, the measured OC concentrations are typically
962
multiplied by a median OM:OC value of 1.7. However, the oxidation capacity of the
963
atmosphere has increased as anthropogenic emissions such as SO2 have decreased
964
(Dalsøren et al., 2016), leading to an increased oxidation rate of organic compounds
965
and the formation of SOA. Consequently, a growing SOA fraction over the last 20 years
966
would have been accompanied by a rising OM:OC ratio. It can be assumed that while
967
the OC measured by the filters showed a slight downward trend (Figure 16g), a
968
conversion into OA via adapted gradually increasing OM:OC ratios could have
969
compensated the OC reduction and show a better matching trend compared to the AMS
970
and EMAC OA.
971
972
5.2 North America
973
Over North America, the filter measured inorganic aerosol compound
974
concentrations declined strongly during the last 20 years, following their precursor
975
emission reductions, with higher reductions over urban locations (Figure 18) and less
976
over rural regions (Figure 19). Nitrate reductions are more pronounced over urban
977
regions (-0.07 µg m-3 yr-1), especially during winter, while over rural locations, the
978
decline is imperceptible (-0.01 µg m-3 yr-1) since the abundance of NH3 have
979
decelerated the decrease of NH4NO3. On the other hand, the drastic decrease of SO2
980
emissions (Table S5, Figure S4) resulted in strong reductions of sulfate concentrations
981
primarily over urban areas (-0.16 µg m-3 yr-1) but also over remote regions (-0.07 µg m-
982
3 yr-1), especially during the summer seasons. Following the reductions of sulfate and
983
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
43
984
(b) Sulfate
(d) Nitrate
(f) Ammonium
(h) Organic aerosol
EMAC simulations over urban N. America
EPA observations over N. America
(a) Sulfate
(c) Nitrate
(e) Ammonium
(g) Organic aerosol
Figure 18: Temporal evolution of the observed (a, c, e, and g subplots on the left)
and simulated (b, d, f, h subplots on the right) concentrations of PM2.5 sulfate (a, b),
nitrate (c, d), ammonium, (e, f), and organic aerosol (g, h) during the period 2000–
2018 over urban locations in North America. Black lines show the annual trend
while the dark blue, light blue, orange, and red lines represent the seasonal trends
during winter, spring, summer, and autumn. Ranges represent the 1σ SD (standard
deviation).
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
44
nitrate, ammonium decrease strongly over urban locations by -0.08 µg m-3 yr-1,
985
especially during the 2010s (Figure 18), even though NH3 emissions remain practically
986
unchanged (Figure S4). Similarly, over Canada, strong reductions in sulfate and nitrate
987
concentrations were observed by the Canadian Air and Precipitation Monitoring
988
Network (CAPMoN), driven by significant decreases in SO₂ and NOₓ emissions (Cheng
989
et al., 2022). While PM2.5 concentrations decreased in eastern Canada, as observed by
990
the National Air Pollution Surveillance (NAPS), emission reductions were less
991
effective in the west, where large-scale wildfires overwhelmed these improvements and
992
even led to occasional increases in PM2.5 concentrations (Yao and Zhang, 2024). These
993
Figure 19: Temporal evolution of the observed (a, c, e, and g subplots on the left) and
simulated (b, d, f, h subplots on the right) concentrations of PM2.5 sulfate (a, b), nitrate
(c, d), and organic aerosol (e, f) during the period 2000–2018 over rural locations in
North America. Black lines show the annual trend while the dark blue, light blue,
orange, and red lines represent the seasonal trends during winter, spring, summer, and
autumn. Ranges represent the 1σ SD (standard deviation).
(b) Sulfate
(d) Nitrate
(f) Organic aerosol
EMAC simulations over rural N. America
IMPROVE observations over N. America
(a) Sulfate
(c) Nitrate
(e) Organic aerosol
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
45
regional differences over Canada are also captured by the EMAC model (Figure 15).
994
Furthermore, EMAC simulates a weaker decline of sulfate concentrations over both
995
rural and urban locations (Figures 18 and 19), mainly due to its tendency to
996
underestimate sulfate concentrations during the 2000s and especially during summer.
997
Reductions on the simulated nitrate and ammonium concentrations are also noticeable
998
but to a lesser extent than on the filter observations (Figures 18 and 19). The observed
999
OA concentrations over urban regions decrease until 2009, however, they gradually
1000
increase during 2010s by 0.11 µg m-3 yr-1. On the other hand, the model calculated OA
1001
concentration levels remain practically unchanged during the simulated period. Both
1002
the simulated and the observed OA concentration trends are also very weak over the
1003
rural and remote regions (Figure 19).
1004
Figure 20 depicts the decadal PM1 concentration trends in North America between
1005
2000s and 2010s. The AMS data for PM1 aerosol composition is composed of
1006
observational datasets from 30 field campaigns during the 2000s and 58 during the
1007
2010s (Figure 2). This uneven distribution can statistically manipulate the calculations
1008
(d) Organic
Aerosol
(b) Nitrate
(a) Sulfate
(c) Ammonium
Figure 20: Decadal PM1 concentration trends in North America expressed by the bar
plots of the mass concentration (in μg m–3) for (a) sulfate, (b) nitrate, (c) ammonium,
and (d) OA during the periods 2000 - 2010 (left) and 2011 - 2020 (right) as calculated
from the AMS observational dataset (dark colors) and the corresponding simulation
values (light colors). The upper and lower whiskers range from 10-90%, the quartiles
from 25-75% of the dataset. The black line is the median.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
46
and hinder the extraction of valid statements for trends over North America. Sulfate
1009
concentrations exhibit a tighter distribution during the 2nd decade (Figure 20); however,
1010
the mean concentration remains unchanged between the two decades. On the other
1011
hand, the simulated sulfate concentrations increase during the 2010s, mainly due to the
1012
larger proportion of urban field campaigns during the second decade. Indeed, the model
1013
simulates a reduction of the continental average sulfate concentrations by 20%, with
1014
maximum differences exceeding 1 μg m-3 over the Southeast US (Figure 15). This
1015
contradicted behavior is also mirrored on nitrate concentrations where both the AMS
1016
dataset and the corresponding simulated results produce a positive trend between the
1017
two decades, while the simulated continental average nitrate concentrations decrease
1018
(Figure 15). Furthermore, compared to AMS observations, the model tends to
1019
underpredict sulfate concentrations and overpredict nitrate. This results in a strong
1020
correlation of the simulated ammonium with nitrate exhibiting a significant positive
1021
trend, which is not observed in the AMS dataset (Figure 20). Finally, as for PM2.5 OA,
1022
the observed and, to a lesser extent, the simulated PM1 OA concentrations increase
1023
slightly during the 2010s.
1024
1025
5.3 Eastern Asia
1026
EANET observations of PM2.5 sulfate reveal a significant increase of its
1027
concentrations until 2007 (Figure 21). However, in view of the upcoming Beijing
1028
Olympic Games in 2008, the first SO2 emission controls have started to be
1029
implemented, and sulfate gradually reduced by -0.27 µg m-3 yr-1. By the end of 2017,
1030
SO2 emissions have been declined by 59% following the Clean Air Action (Zhai et al.,
1031
2019), however, observed sulfate concentrations have decreased by only 23% due to an
1032
increased dry deposition and oxidation rate of SO2 during the same period (Fagerli et
1033
al., 2016). EMAC fails to reproduce the reduction of sulfate concentrations after 2008
1034
since the CAMS emission inventory assumes only a stabilization of SO2 emissions after
1035
the year 2013, instead of a strong decline (Figure S4). At the same period, NOx was
1036
reduced by 21% and NH3 by just 3% (Zhai et al., 2019). This however is not mirrored
1037
in the observed nitrate trends (Figure 21), where nitrate reduces by only -0.05 µg m-3
1038
yr-1 after 2007. The strong SO2 reduction hinders the decline of nitrate since reductions
1039
in (NH4)2SO4 release NH3 to react with HNO3 and form NH4NO3. In contrast to
1040
observations, the simulated nitrate and ammonium continues to increase until the end
1041
of 2010s following the trends in NOx emissions used as input in the model (Figure S4).
1042
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
47
The frequency of AMS field-campaigns started to grow significantly in Eastern Asia
1043
only after 2008, while after 2013, the first consistent and aggressive emission controls
1044
started in China under the Clean Air Action (Zhai et al., 2019). Thus, since 2013 marks
1045
a significant year for Eastern Asia and due to the lack of AMS campaigns prior to 2006
1046
in the region, the decade comparison for Eastern Asia is done for the periods of 2006-
1047
2012 and 2013-2020. Between these two periods, AMS observations reveal a -17%
1048
decline for sulfate, while the corresponding simulated sulfate concentrations reduce by
1049
just -5% (Figure 22). Similar to PM2.5, the average PM1 nitrate concentrations remain
1050
Figure 21: Temporal evolution of the observed (a, c, e, and g subplots on the left) and
simulated (b, d, f, h subplots on the right) concentrations of PM2.5 sulfate (a, b), nitrate
(c, d), and ammonium (e, f) during the period 2000–2018 over Eastern Asia. Black lines
show the annual trend while the dark blue, light blue, orange, and red lines represent the
seasonal trends during winter, spring, summer, and autumn. Ranges represent the 1σ SD
(standard deviation).
(b) Sulfate
(d) Nitrate
(f) Ammonium
EMAC simulations over E. Asia
EANET observations over E. Asia
(a) Sulfate
(c) Nitrate
(e) Ammonium
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
48
the same between the two periods with a marginal decline observed by the AMS and a
1051
marginal increase simulated by EMAC, while the observed ammonium reduces by 18%
1052
following the reduction in sulfate concentrations (Figure 22). In contrast to inorganic
1053
aerosol precursors, the anthropogenic VOC emissions over Eastern Asia continue to
1054
increase even after 2013, mostly due to the use of solvents but also due to the energy
1055
transformation and industrial sector (Hoesly et al., 2018). Thus, both the observed and
1056
the simulated PM1 OA concentrations increase between the two examined periods by
1057
15% and 33%, respectively (Figure 22).
1058
1059
6 Conclusion
1060
This study investigates global trends in atmospheric aerosol composition over the
1061
past two decades, using the EMAC chemistry-climate model and the CAMS
1062
anthropogenic emissions inventory. Results integrate model outputs with global
1063
observational data from 2000-2020, covering PM2.5 composition from regional
1064
monitoring networks (e.g., EMEP in Europe) and PM1 composition from 744 AMS
1065
observational datasets at 169 sites worldwide. Findings reveal substantial regional
1066
(d) Organic
Aerosol
(b) Nitrate
(a) Sulfate
(c) Ammonium
Figure 22: Decadal PM1 concentration trends in Eastern Asia expressed by the bar plots
of the mass concentration (in μg m–3) for (a) sulfate, (b) nitrate, (c) ammonium, and (d)
OA during the periods 2006 - 2012 (left) and 2013 - 2020 (right) as calculated from the
AMS observational dataset (dark colors) and the corresponding simulation values (light
colors). The upper and lower whiskers range from 10-90%, the quartiles from 25-75%
of the dataset. The black line is the median.
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
49
variations in aerosol composition driven by industrial activities, energy production, and
1067
air quality regulations, highlighting the complexity of air pollution dynamics and its
1068
management.
1069
AMS field campaign data show that OA are the dominant PM1 component globally,
1070
especially in tropical and subtropical regions affected by biomass burning and biogenic
1071
VOC emissions. Sulfate is the primary inorganic compound across most areas, though
1072
nitrate predominates in Europe and Eastern Asia. Notably, North America shows
1073
unexpected sulfate dominance, likely due to seasonal sampling bias. HOA levels are
1074
higher in North America and Eastern Asia, while BBOA is prominent in rural Europe
1075
and tropical regions. OOA, particularly aged M-OOA, is the largest OA contributor in
1076
rural regions across all studied areas.
1077
For PM2.5 composition, global filter observations indicate OA as the primary
1078
component in most regions, notably in Southern Hemisphere tropical forests. In Eastern
1079
Asia, OA and elemental carbon (EC) are prominent, while OA and sulfate have similar
1080
importance in rural North America. Globally, sulfate constitutes roughly 50% of the
1081
inorganic PM2.5 mass, followed by nitrate and ammonium. However, sulfate dominance
1082
observed in filter samples contrasts with AMS findings, likely due to sampling artifacts.
1083
Regionally, sulfate is highest in the Middle East, while nitrate plays a significant role
1084
in Europe. Across eight regions, PM2.5 averages are: 21% sulfate, 12% nitrate, 10%
1085
ammonium, 2% sodium, 3% chloride, 40% OA, and 12% EC.
1086
The EMAC model confirms OA as the dominant component of fine aerosols
1087
globally, with the highest concentrations in regions influenced by biomass burning,
1088
such as tropical forests and savannas. Northern industrialized regions exhibit
1089
substantial OA levels (30-35%) from fossil and biofuel combustion. While EMAC
1090
successfully reproduces the prominence of SOA, it struggles to accurately simulate
1091
aged SOA in areas like Eastern Asia. The model further suggests that nitrate surpasses
1092
sulfate in PM2.5 composition in Europe, North America, and Eastern Asia, consistent
1093
with AMS findings but differing from some filter observations. Ammonium mirrors
1094
sulfate and nitrate distribution, with significant contributions in populated and
1095
agricultural regions. Mineral dust and sea salt emissions also play key roles regionally.
1096
Overall, EMAC provides valuable insights into global fine aerosol composition, while
1097
indicating areas for model refinement.
1098
This study presents a 20-year analysis of global trends in fine aerosol composition,
1099
comparing EMAC model simulations with observed trends. Given limited and
1100
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
50
inconsistent PM1 datasets, the analysis focuses on broad regional trends across the first
1101
and second decades, using primarily Northern Hemisphere AMS campaign data and
1102
PM2.5 data from major monitoring networks in North America, Europe, and East Asia.
1103
While these comparisons offer insights, they are complicated by compositional
1104
differences between PM1 and PM2.5 and by differences between real-time AMS and
1105
non-real-time filter-based methods.
1106
Both filter-based data and EMAC simulations show a major decline in key inorganic
1107
components over Europe, especially in sulfate, which dropped by 46% in the last
1108
decade. The EMAC model, however, underestimates the sulfate reduction due to initial
1109
discrepancies in early 2000s concentrations. Nitrate and ammonium also declined
1110
significantly, though the model overestimates nitrate levels. Organic aerosol (OA)
1111
trends vary by method: filter data indicate a slight decrease, while AMS data and
1112
simulations suggest a mild OA increase in PM1, likely due to differences in particle size
1113
(PM2.5 vs. PM1) and instrument detection capabilities (filter-based OC vs. AMS OA).
1114
In North America, filter-based measurements reveal sharp declines in inorganic
1115
aerosol compounds, particularly in urban areas. Nitrate and sulfate concentrations
1116
decreased significantly due to lower SO2 and NOx precursor emissions, with
1117
ammonium levels following this trend, although ammonia itself remained stable in the
1118
2010s. The EMAC model, however, simulates a weaker sulfate and nitrate decline,
1119
underestimating sulfate in the early 2000s while overestimating nitrate. Observed OA
1120
concentrations in urban North America decreased until 2009, then rose in the 2010s, a
1121
trend only partially captured by the model. PM1 sulfate and nitrate levels from AMS
1122
data show inconsistent trends, with the model generally underestimating sulfate and
1123
overestimating nitrate, leading to a positive ammonium trend in the model not observed
1124
in AMS data.
1125
In Eastern Asia, EANET PM2.5 data show rising sulfate concentrations until 2007,
1126
followed by a decline as SO2 emission controls implemented prior to the 2008 Beijing
1127
Olympics. Despite a 59% reduction in SO2 emissions by 2017, sulfate concentrations
1128
fell by only 23%, likely due to increased dry deposition and oxidation rates. The EMAC
1129
model does not fully capture this trend, as it assumes stable SO2 emissions post-2013
1130
rather than a steep decline. Similarly, while observed nitrate and ammonium levels
1131
show minimal reductions after 2007, the model inaccurately projects continued
1132
increases, reflecting discrepancies in NOx emissions trends. AMS data indicate a 17%
1133
reduction in PM1 sulfate from 2006–2012 to 2013–2020, compared to a 5% reduction
1134
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
51
in the model, with observed PM1 OA concentrations increasing by 15% and model
1135
predictions showing a 33% rise, driven by sustained VOC emissions from solvents and
1136
industrial sources.
1137
Overall, despite the complexities and inconsistencies in long-term aerosol trend
1138
analysis due to instrumental and methodological differences, this study highlights the
1139
importance of consistent, long-term global aerosol trend analysis. By integrating model
1140
results and observational data over 20 years, the study reveals significant
1141
spatiotemporal changes in atmospheric aerosol composition over different regions of
1142
the planet, largely driven by recent changes in aerosol precursor emissions.
1143
1144
Code and data availability. The usage of MESSy (Modular Earth Submodel System)
1145
and access to the source code is licensed to all affiliates of institutions which are
1146
members of the MESSy Consortium. Institutions can become a member of the MESSy
1147
Consortium by signing the “MESSy Memorandum of Understanding”. More
1148
information can be found on the MESSy Consortium website: http://www.messy-
1149
interface.org (last access: 8 November 2024). The data produced in the study are
1150
available from the author upon request
1151
1152
1153
Authors contribution: APT designed the research with contributions from VAK. APT
1154
and VAK developed ORACLE-lite. AM and VAK implemented ISOROPIA-lite in
1155
EMAC. SS selected all AMS observations and NM provided specific observations from
1156
sites over the Mediterranean. APT performed the simulations. APT and SS analyzed
1157
the results. APT, SS and VAK wrote the manuscript with contributions from NM and
1158
AM. All co-authors made revisions and corrections.
1159
1160
Competing interests: The authors declare that no competing interests are present
1161
1162
Acknowledgements: The work described in this paper has received funding from the
1163
Initiative and Networking Fund of the Helmholtz Association through the project
1164
“Advanced Earth System Modelling Capacity (ESM)”. The authors gratefully
1165
acknowledge the Earth System Modelling Project (ESM) for funding this work by
1166
providing computing time on the ESM partition of the supercomputer JUWELS
1167
(Alvarez, 2021) at the Jülich Supercomputing Centre (JSC).
1168
1169
Financial support: This research has been supported by the project FORCeS funded
1170
from the European Union’s Horizon 2020 research and innovation program under grant
1171
agreement no. 821205.
1172
1173
References
1174
Aggarwal, S. G. and Kawamura, K.: Carbonaceous and inorganic composition in long-
1175
range transported aerosols over northern Japan: Implication for aging of water-
1176
soluble organic fraction, Atmospheric Environment, 43, 2532-2540,
1177
10.1016/j.atmosenv.2009.02.032, 2009.
1178
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
52
Aiken, A. C., Decarlo, P. F., Kroll, J. H., Worsnop, D. R., Huffman, J. A., Docherty,
1179
K. S., Ulbrich, I. M., Mohr, C., Kimmel, J. R., Sueper, D., Sun, Y., Zhang, Q.,
1180
Trimborn, A., Northway, M., Ziemann, P. J., Canagaratna, M. R., Onasch, T. B.,
1181
Alfarra, M. R., Prevot, A. S. H., Dommen, J., Duplissy, J., Metzger, A.,
1182
Baltensperger, U., and Jimenez, J. L.: O/C and OM/OC ratios of primary, secondary,
1183
and ambient organic aerosols with high-resolution time-of-flight aerosol mass
1184
spectrometry, Environmen. Sci. & Technol., 42, 4478-4485, 2008.
1185
Akagi, S. K., Yokelson, R. J., Wiedinmyer, C., Alvarado, M. J., Reid, J. S., Karl, T.,
1186
Crounse, J. D., and Wennberg, P. O.: Emission factors for open and domestic
1187
biomass burning for use in atmospheric models, Atmos. Chem. Phys., 11, 4039-
1188
4072, 10.5194/acp-11-4039-2011, 2011.
1189
Ames, R. B. and Malm, W. C.: Comparison of sulfate and nitrate particle mass
1190
concentrations measured by IMPROVE and the CDN, Atmospheric Environment,
1191
35, 905-916, 10.1016/s1352-2310(00)00369-1, 2001.
1192
Andreae, M. O. and Rosenfeld, D.: Aerosol-cloud-precipitation interactions. Part 1. The
1193
nature and sources of cloud-active aerosols, Earth-Science Reviews, 89, 13-41,
1194
10.1016/j.earscirev.2008.03.001, 2008.
1195
Anttila, P. and Tuovinen, J. P.: Trends of primary and secondary pollutant
1196
concentrations in Finland in 1994-2007, Atmospheric Environment, 44, 30-41,
1197
10.1016/j.atmosenv.2009.09.041, 2010.
1198
Astitha, M., Lelieveld, J., Kader, M. A., Pozzer, A., and de Meij, A.: Parameterization
1199
of dust emissions in the global atmospheric chemistry-climate model EMAC: impact
1200
of nudging and soil properties, Atmospheric Chemistry and Physics, 12, 11057-
1201
11083, 10.5194/acp-12-11057-2012, 2012.
1202
Bacer, S., Sullivan, S. C., Karydis, V. A., Barahona, D., Kramer, M., Nenes, A., Tost,
1203
H., Tsimpidi, A. P., Lelieveld, J., and Pozzer, A.: Implementation of a
1204
comprehensive ice crystal formation parameterization for cirrus and mixed-phase
1205
clouds in the EMAC model (based on MESSy 2.53), Geoscientific Model
1206
Development, 11, 4021-4041, 10.5194/gmd-11-4021-2018, 2018.
1207
Batmunkh, T., Kim, Y. J., Lee, K. Y., Cayetano, M. G., Jung, J. S., Kim, S. Y., Kim,
1208
K. C., Lee, S. J., Kim, J. S., Chang, L. S., and An, J. Y.: Time-Resolved
1209
Measurements of PM2.5 Carbonaceous Aerosols at Gosan, Korea, J. Air Waste
1210
Manage. Assoc., 61, 1174-1182, 10.1080/10473289.2011.609761, 2011.
1211
Bian, H. S., Chin, M., Hauglustaine, D. A., Schulz, M., Myhre, G., Bauer, S. E., Lund,
1212
M. T., Karydis, V. A., Kucsera, T. L., Pan, X. H., Pozzer, A., Skeie, R. B., Steenrod,
1213
S. D., Sudo, K., Tsigaridis, K., Tsimpidi, A. P., and Tsyro, S. G.: Investigation of
1214
global particulate nitrate from the AeroCom phase III experiment, Atmospheric
1215
Chemistry and Physics, 17, 12911-12940, 10.5194/acp-17-12911-2017, 2017.
1216
Bougiatioti, A., Stavroulas, I., Kostenidou, E., Zarmpas, P., Theodosi, C., Kouvarakis,
1217
G., Canonaco, F., Prevot, A. S. H., Nenes, A., Pandis, S. N., and Mihalopoulos, N.:
1218
Processing of biomass-burning aerosol in the eastern Mediterranean during
1219
summertime, Atmospheric Chemistry and Physics, 14, 4793-4807, 10.5194/acp-14-
1220
4793-2014, 2014.
1221
Bourotte, C., Curl-Amarante, A. P., Forti, M. C., Pereira, L. A. A., Braga, A. L., and
1222
Lotufo, P. A.: Association between ionic composition of fine and coarse aerosol
1223
soluble fraction and peak expiratory flow of asthmatic patients in Sao Paulo city
1224
(Brazil), Atmospheric Environment, 41, 2036-2048,
1225
10.1016/j.atmosenv.2006.11.004, 2007.
1226
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
53
Bouwman, A. F., Lee, D. S., Asman, W. A. H., Dentener, F. J., VanderHoek, K. W.,
1227
and Olivier, J. G. J.: A global high-resolution emission inventory for ammonia,
1228
Global Biogeochemical Cycles, 11, 561-587, 10.1029/97gb02266, 1997.
1229
Bozzetti, C., El Haddad, I., Salameh, D., Daellenbach, K. R., Fermo, P., Gonzalez, R.,
1230
Minguillón, M. C., Iinuma, Y., Poulain, L., Elser, M., Müller, E., Slowik, J. G.,
1231
Jaffrezo, J. L., Baltensperger, U., Marchand, N., and Prévôt, A. S. H.: Organic
1232
aerosol source apportionment by offline-AMS over a full year in Marseille, Atmos.
1233
Chem. Phys., 17, 8247-8268, 10.5194/acp-17-8247-2017, 2017.
1234
Brook, R. D., Rajagopalan, S., Pope, C. A., 3rd, Brook, J. R., Bhatnagar, A., Diez-
1235
Roux, A. V., Holguin, F., Hong, Y., Luepker, R. V., Mittleman, M. A., Peters, A.,
1236
Siscovick, D., Smith, S. C., Jr., Whitsel, L., and Kaufman, J. D.: Particulate matter
1237
air pollution and cardiovascular disease: An update to the scientific statement from
1238
the American Heart Association, Circulation, 121, 2331-2378,
1239
10.1161/CIR.0b013e3181dbece1, 2010.
1240
Budisulistiorini, S. H., Riva, M., Williams, M., Miyakawa, T., Chen, J., Itoh, M.,
1241
Surratt, J. D., and Kuwata, M.: Dominant contribution of oxygenated organic aerosol
1242
to haze particles from real-time observation in Singapore during an Indonesian
1243
wildfire event in 2015, Atmos. Chem. Phys., 18, 16481-16498, 10.5194/acp-18-
1244
16481-2018, 2018.
1245
Budisulistiorini, S. H., Baumann, K., Edgerton, E. S., Bairai, S. T., Mueller, S., Shaw,
1246
S. L., Knipping, E. M., Gold, A., and Surratt, J. D.: Seasonal characterization of
1247
submicron aerosol chemical composition and organic aerosol sources in the
1248
southeastern United States: Atlanta, Georgia,and Look Rock, Tennessee, Atmos.
1249
Chem. Phys., 16, 5171-5189, 10.5194/acp-16-5171-2016, 2016.
1250
Budisulistiorini, S. H., Canagaratna, M. R., Croteau, P. L., Marth, W. J., Baumann, K.,
1251
Edgerton, E. S., Shaw, S. L., Knipping, E. M., Worsnop, D. R., Jayne, J. T., Gold,
1252
A., and Surratt, J. D.: Real-Time Continuous Characterization of Secondary Organic
1253
Aerosol Derived from Isoprene Epoxydiols in Downtown Atlanta, Georgia, Using
1254
the Aerodyne Aerosol Chemical Speciation Monitor, Environmental Science &
1255
Technology, 47, 5686-5694, 10.1021/es400023n, 2013.
1256
Canagaratna, M. R., Jayne, J. T., Jimenez, J. L., Allan, J. D., Alfarra, M. R., Zhang, Q.,
1257
Onasch, T. B., Drewnick, F., Coe, H., Middlebrook, A., Delia, A., Williams, L. R.,
1258
Trimborn, A. M., Northway, M. J., DeCarlo, P. F., Kolb, C. E., Davidovits, P., and
1259
Worsnop, D. R.: Chemical and microphysical characterization of ambient aerosols
1260
with the aerodyne aerosol mass spectrometer, Mass Spectrometry Reviews, 26, 185-
1261
222, https://doi.org/10.1002/mas.20115, 2007.
1262
Cash, J. M., Langford, B., Di Marco, C., Mullinger, N. J., Allan, J., Reyes-Villegas, E.,
1263
Joshi, R., Heal, M. R., Acton, W. J. F., Hewitt, C. N., Misztal, P. K., Drysdale, W.,
1264
Mandal, T. K., Shivani, Gadi, R., Gurjar, B. R., and Nemitz, E.: Seasonal analysis
1265
of submicron aerosol in Old Delhi using high-resolution aerosol mass spectrometry:
1266
chemical characterisation, source apportionment and new marker identification,
1267
Atmos. Chem. Phys., 21, 10133-10158, 10.5194/acp-21-10133-2021, 2021.
1268
Celis, J. E., Morales, J. R., Zaror, C. A., and Inzunza, J. C.: A study of the particulate
1269
matter PM10 composition in the atmosphere of Chillan, Chile, Chemosphere, 54,
1270
541-550, 10.1016/s0045-6535(03)00711-2, 2004.
1271
Chakraborty, A., Bhattu, D., Gupta, T., Tripathi, S. N., and Canagaratna, M. R.: Real-
1272
time measurements of ambient aerosols in a polluted Indian city: Sources,
1273
characteristics, and processing of organic aerosols during foggy and nonfoggy
1274
periods, Journal of Geophysical Research: Atmospheres, 120, 9006-9019,
1275
https://doi.org/10.1002/2015JD023419, 2015.
1276
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
54
Cheng, I., Zhang, L., He, Z., Cathcart, H., Houle, D., Cole, A., Feng, J., O'Brien, J.,
1277
Macdonald, A. M., Aherne, J., and Brook, J.: Long-term declines in atmospheric
1278
nitrogen and sulfur deposition reduce critical loads exceedances at multiple
1279
Canadian rural sites, 2000–2018, Atmos. Chem. Phys., 22, 14631-14656,
1280
10.5194/acp-22-14631-2022, 2022.
1281
Cho, S. Y. and Park, S. S.: Resolving sources of water-soluble organic carbon in fine
1282
particulate matter measured at an urban site during winter, Environmental Science-
1283
Processes & Impacts, 15, 524-534, 10.1039/c2em30730h, 2013.
1284
Chow, W. S., Liao, K. Z., Huang, X. H. H., Leung, K. F., Lau, A. K. H., and Yu, J. Z.:
1285
Measurement report: The 10-year trend of PM<sub>2.5 </sub>major components
1286
and source tracers from 2008 to 2017 in an urban site of Hong Kong, China,
1287
Atmospheric Chemistry and Physics, 22, 11557-11577, 10.5194/acp-22-11557-
1288
2022, 2022.
1289
Cottrell, L. D., Griffin, R. J., Jimenez, J. L., Zhang, Q., Ulbrich, I., Ziemba, L. D.,
1290
Beckman, P. J., Sive, B. C., and Talbot, R. W.: Submicron particles at Thompson
1291
Farm during ICARTT measured using aerosol mass spectrometry, Journal of
1292
Geophysical Research-Atmospheres, 113, 10.1029/2007jd009192, 2008.
1293
Crippa, M., Canonaco, F., Lanz, V. A., Aijala, M., Allan, J. D., Carbone, S., Capes, G.,
1294
Ceburnis, D., Dall'Osto, M., Day, D. A., DeCarlo, P. F., Ehn, M., Eriksson, A.,
1295
Freney, E., Ruiz, L. H., Hillamo, R., Jimenez, J. L., Junninen, H., Kiendler-Scharr,
1296
A., Kortelainen, A. M., Kulmala, M., Laaksonen, A., Mensah, A., Mohr, C., Nemitz,
1297
E., O'Dowd, C., Ovadnevaite, J., Pandis, S. N., Petaja, T., Poulain, L., Saarikoski,
1298
S., Sellegri, K., Swietlicki, E., Tiitta, P., Worsnop, D. R., Baltensperger, U., and
1299
Prevot, A. S. H.: Organic aerosol components derived from 25 AMS data sets across
1300
Europe using a consistent ME-2 based source apportionment approach, Atmospheric
1301
Chemistry and Physics, 14, 6159-6176, 10.5194/acp-14-6159-2014, 2014.
1302
Dalsøren, S. B., Myhre, C. L., Myhre, G., Gomez-Pelaez, A. J., Søvde, O. A., Isaksen,
1303
I. S. A., Weiss, R. F., and Harth, C. M.: Atmospheric methane evolution the last 40
1304
years, Atmos. Chem. Phys., 16, 3099-3126, 10.5194/acp-16-3099-2016, 2016.
1305
de Sá, S. S., Rizzo, L. V., Palm, B. B., Campuzano-Jost, P., Day, D. A., Yee, L. D.,
1306
Wernis, R., Isaacman-VanWertz, G., Brito, J., Carbone, S., Liu, Y. J., Sedlacek, A.,
1307
Springston, S., Goldstein, A. H., Barbosa, H. M. J., Alexander, M. L., Artaxo, P.,
1308
Jimenez, J. L., and Martin, S. T.: Contributions of biomass-burning, urban, and
1309
biogenic emissions to the concentrations and light-absorbing properties of
1310
particulate matter in central Amazonia during the dry season, Atmos. Chem. Phys.,
1311
19, 7973-8001, 10.5194/acp-19-7973-2019, 2019.
1312
DeCarlo, P. F., Kimmel, J. R., Trimborn, A., Northway, M. J., Jayne, J. T., Aiken, A.
1313
C., Gonin, M., Fuhrer, K., Horvath, T., Docherty, K. S., Worsnop, D. R., and
1314
Jimenez, J. L.: Field-deployable, high-resolution, time-of-flight aerosol mass
1315
spectrometer, Analytical Chemistry, 78, 8281-8289, 10.1021/ac061249n, 2006.
1316
Dentener, F., Kinne, S., Bond, T., Boucher, O., Cofala, J., Generoso, S., Ginoux, P.,
1317
Gong, S., Hoelzemann, J. J., Ito, A., Marelli, L., Penner, J. E., Putaud, J. P., Textor,
1318
C., Schulz, M., van der Werf, G. R., and Wilson, J.: Emissions of primary aerosol
1319
and precursor gases in the years 2000 and 1750 prescribed data-sets for AeroCom,
1320
Atmos. Chem. Phys., 6, 4321-4344, 2006.
1321
Docherty, K. S., Aiken, A. C., Huffman, J. A., Ulbrich, I. M., DeCarlo, P. F., Sueper,
1322
D., Worsnop, D. R., Snyder, D. C., Peltier, R. E., Weber, R. J., Grover, B. D.,
1323
Eatough, D. J., Williams, B. J., Goldstein, A. H., Ziemann, P. J., and Jimenez, J. L.:
1324
The 2005 Study of Organic Aerosols at Riverside (SOAR-1): instrumental
1325
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
55
intercomparisons and fine particle composition, Atmospheric Chemistry and
1326
Physics, 11, 12387-12420, 10.5194/acp-11-12387-2011, 2011.
1327
Dominici, F., Peng, R. D., Bell, M. L., Pham, L., McDermott, A., Zeger, S. L., and
1328
Samet, J. M.: Fine particulate air pollution and hospital admission for cardiovascular
1329
and respiratory diseases, Jama, 295, 1127-1134, 10.1001/jama.295.10.1127, 2006.
1330
Du, W., Sun, Y. L., Xu, Y. S., Jiang, Q., Wang, Q. Q., Yang, W., Wang, F., Bai, Z. P.,
1331
Zhao, X. D., and Yang, Y. C.: Chemical characterization of submicron aerosol and
1332
particle growth events at a national background site (3295 m a.s.l.) on the Tibetan
1333
Plateau, Atmos. Chem. Phys., 15, 10811-10824, 10.5194/acp-15-10811-2015, 2015.
1334
EMEP: EMEP Status Report: Transboundary particulate matter, photo-oxidants,
1335
acidifying and eutrophying components., Norwegian Meterological Institute, 2021.
1336
Fagerli, H., Tsyro, S., Denby, B., Olivie, D., Nyiri, A., Gauss, M., Simpson, D., Wind,
1337
P., Benedictow, A., Mortier, A., Jonson, J., Schulz, M., Kirkevåg, A., Valdebenito,
1338
A., Iversen, T., Seland, Ø., Aas, W., Hjellbrekke, A.-G., Solberg, S., and Varma, V.:
1339
Transboundary particulate matter, photo-oxidants, acidifying and eutrophying
1340
components. EMEP Status Report 2016, 10.13140/RG.2.2.27632.46088, 2016.
1341
Fang, T., Guo, H. Y., Zeng, L. H., Verma, V., Nenes, A., and Weber, R. J.: Highly
1342
Acidic Ambient Particles, Soluble Metals, and Oxidative Potential: A Link between
1343
Sulfate and Aerosol Toxicity, Environmental Science & Technology, 51, 2611-
1344
2620, 10.1021/acs.est.6b06151, 2017.
1345
Favez, O., Cachler, H., Sciare, J., Alfaro, S. C., El-Araby, T. M., Harhash, M. A., and
1346
Abdelwahab, M. M.: Seasonality of major aerosol species and their transformations
1347
in Cairo megacity, Atmospheric Environment, 42, 1503-1516,
1348
10.1016/j.atmosenv.2007.10.081, 2008.
1349
Feng, J., Hu, M., Chan, C. K., Lau, P. S., Fang, M., He, L., and Tang, X.: A comparative
1350
study of the organic matter in PM2.5 from three Chinese megacities in three different
1351
climatic zones, Atmospheric Environment, 40, 3983-3994,
1352
10.1016/j.atmosenv.2006.02.017, 2006.
1353
Fuzzi, S., Decesari, S., Facchini, M. C., Cavalli, F., Emblico, L., Mircea, M., Andreae,
1354
M. O., Trebs, I., Hoffer, A., Guyon, P., Artaxo, P., Rizzo, L. V., Lara, L. L.,
1355
Pauliquevis, T., Maenhaut, W., Raes, N., Chi, X. G., Mayol-Bracero, O. L., Soto-
1356
Garcia, L. L., Claeys, M., Kourtchev, I., Rissler, J., Swietlicki, E., Tagliavini, E.,
1357
Schkolnik, G., Falkovich, A. H., Rudich, Y., Fisch, G., and Gatti, L. V.: Overview
1358
of the inorganic and organic composition of size-segregated aerosol in Rondonia,
1359
Brazil, from the biomass-burning period to the onset of the wet season, Journal of
1360
Geophysical Research-Atmospheres, 112, 10.1029/2005jd006741, 2007.
1361
George, D. T., Howard, K., Isabella, A.-M., John, B., Robert, D. B., Kevin, C., Sara
1362
De, M., Francesco, F., Bertil, F., Mark, W. F., Jonathan, G., Dick, H., Frank, J. K.,
1363
Nino, K., Robert, L., Annette, P., Sanjay, T. R., David, R., Beate, R., Jonathan, M.
1364
S., Thomas, S., Torben, S., Jordi, S., and Bert, B.: A joint ERS/ATS policy
1365
statement: what constitutes an adverse health effect of air pollution? An analytical
1366
framework, European Respiratory Journal, 49, 1600419, 10.1183/13993003.00419-
1367
2016, 2017.
1368
Gioda, A., Amaral, B. S., Monteiro, I. L. G., and Saint'Pierre, T. D.: Chemical
1369
composition, sources, solubility, and transport of aerosol trace elements in a tropical
1370
region, Journal of Environmental Monitoring, 13, 2134-2142, 10.1039/c1em10240k,
1371
2011.
1372
Gkatzelis, G. I., Papanastasiou, D. K., Karydis, V. A., Hohaus, T., Liu, Y., Schmitt, S.
1373
H., Schlag, P., Fuchs, H., Novelli, A., Chen, Q., Cheng, X., Broch, S., Dong, H.,
1374
Holland, F., Li, X., Liu, Y. H., Ma, X. F., Reimer, D., Rohrer, F., Shao, M., Tan, Z.,
1375
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
56
Taraborrelli, D., Tillmann, R., Wang, H. C., Wang, Y., Wu, Y. S., Wu, Z. J., Zeng,
1376
L. M., Zheng, J., Hu, M., Lu, K. D., Hofzumahaus, A., Zhang, Y. H., Wahner, A.,
1377
and Kiendler-Scharr, A.: Uptake of Water-soluble Gas-phase Oxidation Products
1378
Drives Organic Particulate Pollution in Beijing, Geophysical Research Letters, 48,
1379
10.1029/2020gl091351, 2021.
1380
Goldstein, A. H. and Galbally, I. E.: Known and unexplored organic constituents in the
1381
earth's atmosphere, Environmental Science & Technology, 41, 1514-1521,
1382
10.1021/es072476p, 2007.
1383
Granier, C., Darras, S., Gon, H. D. v. d., Doubalova, J., Elguindi, N., Galle, B., Gauss,
1384
M., Guevara, M., Jalkanen, J.-P., Kuenen, J., Liousse, C., Quack, B., Simpson, D.,
1385
and Sindelarova, K.: The Copernicus Atmosphere Monitoring Service global and
1386
regional emissions (April 2019 version) 10.24380/d0bn-kx16, 2019.
1387
Guelle, W., Schulz, M., Balkanski, Y., and Dentener, F.: Influence of the source
1388
formulation on modeling the atmospheric global distribution of sea salt aerosol,
1389
Journal of Geophysical Research: Atmospheres, 106, 27509-27524,
1390
https://doi.org/10.1029/2001JD900249, 2001.
1391
Guenther, A. B., Jiang, X., Heald, C. L., Sakulyanontvittaya, T., Duhl, T., Emmons, L.
1392
K., and Wang, X.: The Model of Emissions of Gases and Aerosols from Nature
1393
version 2.1 (MEGAN2.1): an extended and updated framework for modeling
1394
biogenic emissions, Geosci. Model Dev., 5, 1471-1492, 10.5194/gmd-5-1471-2012,
1395
2012.
1396
Guerreiro, C. B. B., Foltescu, V., and de Leeuw, F.: Air quality status and trends in
1397
Europe, Atmospheric Environment, 98, 376-384, 10.1016/j.atmosenv.2014.09.017,
1398
2014.
1399
Hallquist, M., Wenger, J. C., Baltensperger, U., Rudich, Y., Simpson, D., Claeys, M.,
1400
Dommen, J., Donahue, N. M., George, C., Goldstein, A. H., Hamilton, J. F.,
1401
Herrmann, H., Hoffmann, T., Iinuma, Y., Jang, M., Jenkin, M. E., Jimenez, J. L.,
1402
Kiendler-Scharr, A., Maenhaut, W., McFiggans, G., Mentel, T. F., Monod, A.,
1403
Prevot, A. S. H., Seinfeld, J. H., Surratt, J. D., Szmigielski, R., and Wildt, J.: The
1404
formation, properties and impact of secondary organic aerosol: current and emerging
1405
issues, Atmospheric Chemistry and Physics, 9, 5155-5236, 10.5194/acp-9-5155-
1406
2009, 2009.
1407
Hand, J., Copeland, S. A., McDade, C., Day, D., Moore, Jr., Dillner, A., Pitchford, M.,
1408
Indresand, H., Schichtel, B., Malm, W., and Watson, J.: Spatial and seasonal patterns
1409
and temporal variability of haze and its constituents in the United States, IMPROVE
1410
Report V, 2011.
1411
Haywood, J. and Boucher, O.: Estimates of the direct and indirect radiative forcing due
1412
to tropospheric aerosols: A review, Reviews of Geophysics, 38, 513-543,
1413
10.1029/1999rg000078, 2000.
1414
Hoesly, R. M., Smith, S. J., Feng, L., Klimont, Z., Janssens-Maenhout, G., Pitkanen,
1415
T., Seibert, J. J., Vu, L., Andres, R. J., Bolt, R. M., Bond, T. C., Dawidowski, L.,
1416
Kholod, N., Kurokawa, J. I., Li, M., Liu, L., Lu, Z., Moura, M. C. P., O'Rourke, P.
1417
R., and Zhang, Q.: Historical (1750–2014) anthropogenic emissions of reactive
1418
gases and aerosols from the Community Emissions Data System (CEDS), Geosci.
1419
Model Dev., 11, 369-408, 10.5194/gmd-11-369-2018, 2018.
1420
Huang, X. F., He, L. Y., Hu, M., Canagaratna, M. R., Sun, Y., Zhang, Q., Zhu, T., Xue,
1421
L., Zeng, L. W., Liu, X. G., Zhang, Y. H., Jayne, J. T., Ng, N. L., and Worsnop, D.
1422
R.: Highly time-resolved chemical characterization of atmospheric submicron
1423
particles during 2008 Beijing Olympic Games using an Aerodyne High-Resolution
1424
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
57
Aerosol Mass Spectrometer, Atmospheric Chemistry and Physics, 10, 8933-8945,
1425
10.5194/acp-10-8933-2010, 2010.
1426
IPCC: (Intergovernmental Panel on Climate Change): The physical science basis.
1427
Contribution of working group I to the fifth assessment report of the
1428
intergovernmental panel on climate change. T.F. Stocker, D. Qin, G.-K. Plattner, M.
1429
Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex, and P.M. Midgley
1430
(eds.). Cambridge University Press, Cambridge, United Kingdom and New York,
1431
NY, USA, 2013.
1432
IPCC, P.R. Shukla, J. S., R. Slade, A. Al Khourdajie, R. van Diemen, D. McCollum,
1433
M. Pathak, S. Some, P. Vyas, R. Fradera, M. Belkacemi, A. Hasija, G. Lisboa, S.
1434
Luz, J. Malley (Ed.): Climate Change 2022: Mitigation of Climate Change.
1435
Contribution of Working Group III to the Sixth Assessment Report of the
1436
Intergovernmental Panel on Climate Change Cambridge University Press,
1437
Cambridge, UK and New York, NY, USA, 10.1017/9781009157926, 2022.
1438
Janssen, R. H. H., Tsimpidi, A. P., Karydis, V. A., Pozzer, A., Lelieveld, J., Crippa, M.,
1439
Prévôt, A. S. H., Ait-Helal, W., Borbon, A., Sauvage, S., and Locoge, N.: Influence
1440
of local production and vertical transport on the organic aerosol budget over Paris,
1441
Journal of Geophysical Research: Atmospheres, 122, 8276-8296,
1442
https://doi.org/10.1002/2016JD026402, 2017.
1443
Jayne, J. T., Leard, D. C., Zhang, X., Davidovits, P., Smith, K. A., Kolb, C. E., and
1444
Worsnop, D. R.: Development of an Aerosol Mass Spectrometer for Size and
1445
Composition Analysis of Submicron Particles, Aerosol Science and Technology, 33,
1446
49-70, 10.1080/027868200410840, 2000.
1447
Jimenez, J. L., Canagaratna, M. R., Donahue, N. M., Prevot, A. S. H., Zhang, Q., Kroll,
1448
J. H., DeCarlo, P. F., Allan, J. D., Coe, H., Ng, N. L., Aiken, A. C., Docherty, K. S.,
1449
Ulbrich, I. M., Grieshop, A. P., Robinson, A. L., Duplissy, J., Smith, J. D., Wilson,
1450
K. R., Lanz, V. A., Hueglin, C., Sun, Y. L., Tian, J., Laaksonen, A., Raatikainen, T.,
1451
Rautiainen, J., Vaattovaara, P., Ehn, M., Kulmala, M., Tomlinson, J. M., Collins, D.
1452
R., Cubison, M. J., Dunlea, E. J., Huffman, J. A., Onasch, T. B., Alfarra, M. R.,
1453
Williams, P. I., Bower, K., Kondo, Y., Schneider, J., Drewnick, F., Borrmann, S.,
1454
Weimer, S., Demerjian, K., Salcedo, D., Cottrell, L., Griffin, R., Takami, A.,
1455
Miyoshi, T., Hatakeyama, S., Shimono, A., Sun, J. Y., Zhang, Y. M., Dzepina, K.,
1456
Kimmel, J. R., Sueper, D., Jayne, J. T., Herndon, S. C., Trimborn, A. M., Williams,
1457
L. R., Wood, E. C., Middlebrook, A. M., Kolb, C. E., Baltensperger, U., and
1458
Worsnop, D. R.: Evolution of organic aerosols in the atmosphere, Science, 326,
1459
1525-1529, 2009.
1460
Jöckel, P., Kerkweg, A., Pozzer, A., Sander, R., Tost, H., Riede, H., Baumgaertner, A.,
1461
Gromov, S., and Kern, B.: Development cycle 2 of the Modular Earth Submodel
1462
System (MESSy2), Geoscientific Model Development, 3, 717-752, 2010.
1463
Jöckel, P., Tost, H., Pozzer, A., Bruehl, C., Buchholz, J., Ganzeveld, L., Hoor, P.,
1464
Kerkweg, A., Lawrence, M. G., Sander, R., Steil, B., Stiller, G., Tanarhte, M.,
1465
Taraborrelli, D., Van Aardenne, J., and Lelieveld, J.: The atmospheric chemistry
1466
general circulation model ECHAM5/MESSy1: consistent simulation of ozone from
1467
the surface to the mesosphere, Atmos. Chem. Phys., 6, 5067-5104, 2006.
1468
Kaiser, J. W., Heil, A., Andreae, M. O., Benedetti, A., Chubarova, N., Jones, L.,
1469
Morcrette, J. J., Razinger, M., Schultz, M. G., Suttie, M., and van der Werf, G. R.:
1470
Biomass burning emissions estimated with a global fire assimilation system based
1471
on observed fire radiative power, Biogeosciences, 9, 527-554, 10.5194/bg-9-527-
1472
2012, 2012.
1473
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
58
Kakavas, S., Pandis, S. N., and Nenes, A.: ISORROPIA-Lite: A Comprehensive
1474
Atmospheric Aerosol Thermodynamics Module for Earth System Models, Tellus
1475
Series B-Chemical and Physical Meteorology, 74, 1-23, 10.16993/tellusb.33, 2022.
1476
Kanakidou, M., Seinfeld, J. H., Pandis, S. N., Barnes, I., Dentener, F. J., Facchini, M.
1477
C., Van Dingenen, R., Ervens, B., Nenes, A., Nielsen, C. J., Swietlicki, E., Putaud,
1478
J. P., Balkanski, Y., Fuzzi, S., Horth, J., Moortgat, G. K., Winterhalter, R., Myhre,
1479
C. E. L., Tsigaridis, K., Vignati, E., Stephanou, E. G., and Wilson, J.: Organic
1480
aerosol and global climate modelling: a review, Atmos. Chem. Phys., 5, 1053-1123,
1481
2005.
1482
Karydis, V. A., Tsimpidi, A. P., Pozzer, A., and Lelieveld, J.: How alkaline compounds
1483
control atmospheric aerosol particle acidity, Atmospheric Chemistry and Physics,
1484
21, 14983-15001, 10.5194/acp-21-14983-2021, 2021.
1485
Karydis, V. A., Tsimpidi, A. P., Pozzer, A., Astitha, M., and Lelieveld, J.: Effects of
1486
mineral dust on global atmospheric nitrate concentrations, Atmos. Chem. Phys., 16,
1487
1491-1509, 10.5194/acp-16-1491-2016, 2016.
1488
Karydis, V. A., Tsimpidi, A. P., Bacer, S., Pozzer, A., Nenes, A., and Lelieveld, J.:
1489
Global impact of mineral dust on cloud droplet number concentration, Atmospheric
1490
Chemistry and Physics, 17, 5601-5621, 10.5194/acp-17-5601-2017, 2017.
1491
Kerkweg, A., Buchholz, J., Ganzeveld, L., Pozzer, A., Tost, H., and Jöckel, P.:
1492
Technical Note: An implementation of the dry removal processes DRY DEPosition
1493
and SEDImentation in the Modular Earth Submodel System (MESSy), Atmos.
1494
Chem. Phys., 6, 4617-4632, 2006.
1495
Klingmuller, K., Lelieveld, J., Karydis, V. A., and Stenchikov, G. L.: Direct radiative
1496
effect of dust-pollution interactions, Atmospheric Chemistry and Physics, 19, 7397-
1497
7408, 10.5194/acp-19-7397-2019, 2019.
1498
Klingmuller, K., Metzger, S., Abdelkader, M., Karydis, V. A., Stenchikov, G. L.,
1499
Pozzer, A., and Lelieveld, J.: Revised mineral dust emissions in the atmospheric
1500
chemistry-climate model EMAC (MESSy 2.52 DU_Astitha1 KKDU2017 patch),
1501
Geoscientific Model Development, 11, 989-1008, 10.5194/gmd-11-989-2018, 2018.
1502
Klingmüller, K., Karydis, V. A., Bacer, S., Stenchikov, G. L., and Lelieveld, J.: Weaker
1503
cooling by aerosols due to dust-pollution interactions, Atmospheric Chemistry and
1504
Physics, 20, 15285-15295, 10.5194/acp-20-15285-2020, 2020.
1505
Kodros, J. K., Papanastasiou, D. K., Paglione, M., Masiol, M., Squizzato, S., Florou,
1506
K., Skyllakou, K., Kaltsonoudis, C., Nenes, A., and Pandis, S. N.: Rapid dark aging
1507
of biomass burning as an overlooked source of oxidized organic aerosol,
1508
Proceedings of the National Academy of Sciences of the United States of America,
1509
117, 33028-33033, 10.1073/pnas.2010365117, 2020.
1510
Kok, J. F., Storelvmo, T., Karydis, V. A., Adebiyi, A. A., Mahowald, N. M., Evan, A.
1511
T., He, C. L., and Leung, D. M.: Mineral dust aerosol impacts on global climate and
1512
climate change, Nature Reviews Earth & Environment, 4, 71-86, 10.1038/s43017-
1513
022-00379-5, 2023.
1514
Kostenidou, E., Florou, K., Kaltsonoudis, C., Tsiflikiotou, M., Vratolis, S.,
1515
Eleftheriadis, K., and Pandis, S. N.: Sources and chemical characterization of
1516
organic aerosol during the summer in the eastern Mediterranean, Atmospheric
1517
Chemistry and Physics, 15, 11355-11371, 10.5194/acp-15-11355-2015, 2015.
1518
Kuzu, S. L., Yavuz, E., Akyüz, E., Saral, A., Akkoyunlu, B. O., Özdemir, H., Demir,
1519
G., and Ünal, A.: Black carbon and size-segregated elemental carbon, organic carbon
1520
compositions in a megacity: a case study for Istanbul, Air Quality, Atmosphere &
1521
Health, 13, 827-837, 10.1007/s11869-020-00839-1, 2020.
1522
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
59
Kyllönen, K., Vestenius, M., Anttila, P., Makkonen, U., Aurela, M., Wängberg, I.,
1523
Mastromonaco, M. N., and Hakola, H.: Trends and source apportionment of
1524
atmospheric heavy metals at a subarctic site during 1996-2018, Atmospheric
1525
Environment, 236, 10.1016/j.atmosenv.2020.117644, 2020.
1526
Lang, P. E., Carslaw, D. C., and Moller, S. J.: A trend analysis approach for air quality
1527
network data, Atmospheric Environment-X, 2, 10.1016/j.aeaoa.2019.100030, 2019.
1528
Lanz, V. A., Alfarra, M. R., Baltensperger, U., Buchmann, B., Hueglin, C., and Prevot,
1529
A. S. H.: Source apportionment of submicron organic aerosols at an urban site by
1530
factor analytical modelling of aerosol mass spectra, Atmospheric Chemistry and
1531
Physics, 7, 1503-1522, 2007.
1532
Lanz, V. A., Alfarra, M. R., Baltensperger, U., Buchmann, B., Hueglin, C., Szidat, S.,
1533
Wehrli, M. N., Wacker, L., Weimer, S., Caseiro, A., Puxbaum, H., and Prevot, A. S.
1534
H.: Source attribution of submicron organic aerosols during wintertime inversions
1535
by advanced factor analysis of aerosol mass spectra, Environmental Science &
1536
Technology, 42, 214-220, 10.1021/es0707207, 2008.
1537
Lanz, V. A., Prevot, A. S. H., Alfarra, M. R., Weimer, S., Mohr, C., DeCarlo, P. F.,
1538
Gianini, M. F. D., Hueglin, C., Schneider, J., Favez, O., D'Anna, B., George, C., and
1539
Baltensperger, U.: Characterization of aerosol chemical composition with aerosol
1540
mass spectrometry in Central Europe: an overview, Atmospheric Chemistry and
1541
Physics, 10, 10453-10471, 10.5194/acp-10-10453-2010, 2010.
1542
Lelieveld, J., Evans, J. S., Fnais, M., Giannadaki, D., and Pozzer, A.: The contribution
1543
of outdoor air pollution sources to premature mortality on a global scale, Nature,
1544
525, 367-371, 10.1038/nature15371, 2015.
1545
Li, L., Wang, W., Feng, J., Zhang, D., Li, H., Gu, Z., Wang, B., Sheng, G., and Fu, J.:
1546
Composition, source, mass closure of PM2.5 aerosols for four forests in eastern
1547
China, Journal of Environmental Sciences, 22, 405-412, 10.1016/s1001-
1548
0742(09)60122-4, 2010.
1549
Liu, L., ; Thorsten Hohaus; Philipp Franke; Anne C. Lange; Ralf Tillmann; Hendrik
1550
Fuchs; Zhaofeng Tan; Franz Rohrer; Vlassis Karydis; Quanfu He; Vaishali Vardhan;
1551
Stefanie Andres; Birger Bohn; Frank Holland; Benjamin Winter; Sergej Wedel;
1552
Anna Novelli; Andreas Hofzumahaus; Andreas Wahner; and Astrid Kiendler-
1553
Scharr: Observational evidence reveals the significance of nocturnal chemistry in
1554
secondary organic aerosol formation across all seasons, npj Climate and
1555
Atmospheric Science, in review, 2024.
1556
Liu, X., Lara, R., Dufresne, M., Wu, L., Zhang, X., Wang, T., Monge, M., Reche, C.,
1557
Di Leo, A., Lanzani, G., Colombi, C., Font, A., Sheehan, A., Green, D. C.,
1558
Makkonen, U., Sauvage, S., Salameh, T., Petit, J.-E., Chatain, M., Coe, H., Hou, S.,
1559
Harrison, R., Hopke, P. K., Petäjä, T., Alastuey, A., and Querol, X.: Variability of
1560
ambient air ammonia in urban Europe (Finland, France, Italy, Spain, and the UK),
1561
Environment International, 185, 108519,
1562
https://doi.org/10.1016/j.envint.2024.108519, 2024.
1563
Lohmann, U. and Ferrachat, S.: Impact of parametric uncertainties on the present-day
1564
climate and on the anthropogenic aerosol effect, Atmos. Chem. Phys., 10, 11373-
1565
11383, 10.5194/acp-10-11373-2010, 2010.
1566
Mallet, M. D., D'Anna, B., Même, A., Bove, M. C., Cassola, F., Pace, G., Desboeufs,
1567
K., Di Biagio, C., Doussin, J. F., Maille, M., Massabò, D., Sciare, J., Zapf, P., di
1568
Sarra, A. G., and Formenti, P.: Summertime surface PM1 aerosol composition and
1569
size by source region at the Lampedusa island in the central Mediterranean Sea,
1570
Atmos. Chem. Phys., 19, 11123-11142, 10.5194/acp-19-11123-2019, 2019.
1571
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
60
Mariani, R. L. and de Mello, W. Z.: PM2.5-10, PM2.5 and associated water-soluble
1572
inorganic species at a coastal urban site in the metropolitan region of Rio de Janeiro,
1573
Atmospheric Environment, 41, 2887-2892, 10.1016/j.atmosenv.2006.12.009, 2007.
1574
Martin, S. T., Andreae, M. O., Artaxo, P., Baumgardner, D., Chen, Q., Goldstein, A.
1575
H., Guenther, A., Heald, C. L., Mayol-Bracero, O. L., McMurry, P. H., Pauliquevis,
1576
T., Poschl, U., Prather, K. A., Roberts, G. C., Saleska, S. R., Dias, M. A. S.,
1577
Spracklen, D. V., Swietlicki, E., and Trebs, I.: SOURCES AND PROPERTIES OF
1578
AMAZONIAN AEROSOL PARTICLES, Reviews of Geophysics, 48,
1579
10.1029/2008rg000280, 2010.
1580
Meng, Z. Y. and Seinfeld, J. H.: Time scales to achieve atmospheric gas-aerosol
1581
equilibrium for volatile species, Atmospheric Environment, 30, 2889-2900,
1582
10.1016/1352-2310(95)00493-9, 1996.
1583
Milousis, A., Tsimpidi, A. P., Tost, H., Pandis, S. N., Nenes, A., Kiendler-Scharr, A.,
1584
and Karydis, V. A.: Implementation of the ISORROPIA-lite aerosol
1585
thermodynamics model into the EMAC chemistry climate model (based on MESSy
1586
v2.55): implications for aerosol composition and acidity, Geosci. Model Dev., 17,
1587
1111-1131, 10.5194/gmd-17-1111-2024, 2024.
1588
Mkoma, S. L.: Physico-Chemical Characterisation of Atmospheric Aerosols in
1589
Tanzania, with Emphasis on the Carbonaceous Aerosol Components and on
1590
Chemical Mass Closure, 2008.
1591
Mkoma, S. L., Maenhaut, W., Chi, X. G., Wang, W., and Raes, N.: Characterisation of
1592
PM10 atmospheric aerosols for the wet season 2005 at two sites in East Africa,
1593
Atmospheric Environment, 43, 631-639, 10.1016/j.atmosenv.2008.10.008, 2009.
1594
Mohr, C., DeCarlo, P. F., Heringa, M. F., Chirico, R., Slowik, J. G., Richter, R., Reche,
1595
C., Alastuey, A., Querol, X., Seco, R., Penuelas, J., Jimenez, J. L., Crippa, M.,
1596
Zimmermann, R., Baltensperger, U., and Prevot, A. S. H.: Identification and
1597
quantification of organic aerosol from cooking and other sources in Barcelona using
1598
aerosol mass spectrometer data, Atmospheric Chemistry and Physics, 12, 1649-
1599
1665, 10.5194/acp-12-1649-2012, 2012.
1600
Molina, L. T., Kolb, C. E., de Foy, B., Lamb, B. K., Brune, W. H., Jimenez, J. L.,
1601
Ramos-Villegas, R., Sarmiento, J., Paramo-Figueroa, V. H., Cardenas, B.,
1602
Gutierrez-Avedoy, V., and Molina, M. J.: Air quality in North America's most
1603
populous city – overview of the MCMA-2003 campaign, Atmos. Chem.
1604
Phys., 7, 2447-2473, 10.5194/acp-7-2447-2007, 2007.
1605
Molina, L. T., Madronich, S., Gaffney, J. S., Apel, E., de Foy, B., Fast, J., Ferrare, R.,
1606
Herndon, S., Jimenez, J. L., Lamb, B., Osornio-Vargas, A. R., Russell, P., Schauer,
1607
J. J., Stevens, P. S., Volkamer, R., and Zavala, M.: An overview of the MILAGRO
1608
2006 Campaign: Mexico City emissions and their transport and transformation,
1609
Atmospheric Chemistry and Physics, 10, 8697-8760, 10.5194/acp-10-8697-2010,
1610
2010.
1611
Mortier, A., Gliß, J., Schulz, M., Aas, W., Andrews, E., Bian, H., Chin, M., Ginoux, P.,
1612
Hand, J., Holben, B., Zhang, H., Kipling, Z., Kirkevåg, A., Laj, P., Lurton, T.,
1613
Myhre, G., Neubauer, D., Olivié, D., von Salzen, K., Skeie, R. B., Takemura, T.,
1614
and Tilmes, S.: Evaluation of climate model aerosol trends with ground-based
1615
observations over the last 2 decades – an AeroCom and CMIP6 analysis, Atmos.
1616
Chem. Phys., 20, 13355-13378, 10.5194/acp-20-13355-2020, 2020.
1617
Ng, N. L., Canagaratna, M. R., Jimenez, J. L., Zhang, Q., Ulbrich, I. M., and Worsnop,
1618
D. R.: Real-Time Methods for Estimating Organic Component Mass Concentrations
1619
from Aerosol Mass Spectrometer Data, Environmental Science & Technology, 45,
1620
910-916, 10.1021/es102951k, 2011.
1621
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
61
Ng, N. L., Canagaratna, M. R., Zhang, Q., Jimenez, J. L., Tian, J., Ulbrich, I. M., Kroll,
1622
J. H., Docherty, K. S., Chhabra, P. S., Bahreini, R., Murphy, S. M., Seinfeld, J. H.,
1623
Hildebrandt, L., Donahue, N. M., DeCarlo, P. F., Lanz, V. A., Prévôt, A. S. H.,
1624
Dinar, E., Rudich, Y., and Worsnop, D. R.: Organic aerosol components observed
1625
in Northern Hemispheric datasets from Aerosol Mass Spectrometry, Atmos. Chem.
1626
Phys., 10, 4625-4641, 10.5194/acp-10-4625-2010, 2010.
1627
Paatero, P.: Least squares formulation of robust non-negative factor analysis,
1628
Chemometrics and Intelligent Laboratory Systems, 37, 23-35, 10.1016/s0169-
1629
7439(96)00044-5, 1997.
1630
Paatero, P.: The Multilinear Engine—A Table-Driven, Least Squares Program for
1631
Solving Multilinear Problems, Including the n-Way Parallel Factor Analysis Model,
1632
Journal of Computational and Graphical Statistics, 8, 854-888,
1633
10.1080/10618600.1999.10474853, 1999.
1634
Paatero, P. and Tapper, U.: Positive matrix factorization-A nonnegative factor model
1635
with optimal utilization of error-estimates of data values, Environmetrics, 5, 111-
1636
126, 10.1002/env.3170050203, 1994.
1637
Paglione, M., Gilardoni, S., Rinaldi, M., Decesari, S., Zanca, N., Sandrini, S.,
1638
Giulianelli, L., Bacco, D., Ferrari, S., Poluzzi, V., Scotto, F., Trentini, A., Poulain,
1639
L., Herrmann, H., Wiedensohler, A., Canonaco, F., Prévôt, A. S. H., Massoli, P.,
1640
Carbone, C., Facchini, M. C., and Fuzzi, S.: The impact of biomass burning and
1641
aqueous-phase processing on air quality: a multi-year source apportionment study in
1642
the Po Valley, Italy, Atmos. Chem. Phys., 20, 1233-1254, 10.5194/acp-20-1233-
1643
2020, 2020.
1644
Parworth, C., Fast, J., Mei, F., Shippert, T., Sivaraman, C., Tilp, A., Watson, T., and
1645
Zhang, Q.: Long-term measurements of submicrometer aerosol chemistry at the
1646
Southern Great Plains (SGP) using an Aerosol Chemical Speciation Monitor
1647
(ACSM), Atmospheric Environment, 106, 43-55, 10.1016/j.atmosenv.2015.01.060,
1648
2015.
1649
Pathak, R. K., Wang, T., Ho, K. F., and Lee, S. C.: Characteristics of summertime
1650
PM2.5 organic and elemental carbon in four major Chinese cities: Implications of
1651
high acidity for water-soluble organic carbon (WSOC), Atmospheric Environment,
1652
45, 318-325, 10.1016/j.atmosenv.2010.10.021, 2011.
1653
Petit, J. E., Favez, O., Sciare, J., Crenn, V., Sarda-Estève, R., Bonnaire, N., Močnik,
1654
G., Dupont, J. C., Haeffelin, M., and Leoz-Garziandia, E.: Two years of near real-
1655
time chemical composition of submicron aerosols in the region of Paris using an
1656
Aerosol Chemical Speciation Monitor (ACSM) and a multi-wavelength
1657
Aethalometer, Atmos. Chem. Phys., 15, 2985-3005, 10.5194/acp-15-2985-2015,
1658
2015.
1659
Pope, C. A., Ezzati, M., and Dockery, D. W.: Fine-Particulate Air Pollution and Life
1660
Expectancy in the United States, New England Journal of Medicine, 360, 376-386,
1661
10.1056/NEJMsa0805646, 2009.
1662
Pozzer, A., Joeckel, P. J., Sander, R., Williams, J., Ganzeveld, L., and Lelieveld, J.:
1663
Technical note: the MESSy-submodel AIRSEA calculating the air-sea exchange of
1664
chemical species, Atmos. Chem. Phys., 6, 5435-5444, 2006.
1665
Pozzer, A., Reifenberg, S. F., Kumar, V., Franco, B., Kohl, M., Taraborrelli, D.,
1666
Gromov, S., Ehrhart, S., Jöckel, P., Sander, R., Fall, V., Rosanka, S., Karydis, V.,
1667
Akritidis, D., Emmerichs, T., Crippa, M., Guizzardi, D., Kaiser, J. W., Clarisse, L.,
1668
Kiendler-Scharr, A., Tost, H., and Tsimpidi, A.: Simulation of organics in the
1669
atmosphere: evaluation of EMACv2.54 with the Mainz Organic Mechanism (MOM)
1670
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
62
coupled to the ORACLE (v1.0) submodel, Geoscientific Model Development, 15,
1671
2673-2710, 10.5194/gmd-15-2673-2022, 2022.
1672
Price, C. and Rind, D.: A SIMPLE LIGHTNING PARAMETERIZATION FOR
1673
CALCULATING GLOBAL LIGHTNING DISTRIBUTIONS, Journal of
1674
Geophysical Research-Atmospheres, 97, 9919-9933, 1992.
1675
Pringle, K. J., Tost, H., Message, S., Steil, B., Giannadaki, D., Nenes, A., Fountoukis,
1676
C., Stier, P., Vignati, E., and Leieved, J.: Description and evaluation of GMXe: a
1677
new aerosol submodel for global simulations (v1), Geoscientific Model
1678
Development, 3, 391-412, 2010.
1679
Radhi, M., Box, M. A., Box, G. P., Mitchell, R. M., Cohen, D. D., Stelcer, E., and
1680
Keywood, M. D.: Optical, physical and chemical characteristics of Australian
1681
continental aerosols: results from a field experiment, Atmospheric Chemistry and
1682
Physics, 10, 5925-5942, 10.5194/acp-10-5925-2010, 2010.
1683
Rattanavaraha, W., Canagaratna, M. R., Budisulistiorini, S. H., Croteau, P. L.,
1684
Baumann, K., Canonaco, F., Prevot, A. S. H., Edgerton, E. S., Zhang, Z., Jayne, J.
1685
T., Worsnop, D. R., Gold, A., Shaw, S. L., and Surratt, J. D.: Source apportionment
1686
of submicron organic aerosol collected from Atlanta, Georgia, during 2014–2015
1687
using the aerosol chemical speciation monitor (ACSM), Atmospheric Environment,
1688
167, 389-402, https://doi.org/10.1016/j.atmosenv.2017.07.055, 2017.
1689
Roeckner, E., Brokopf, R., Esch, M., Giorgetta, M., Hagemann, S., Kornblueh, L.,
1690
Manzini, E., Schlese, U., and Schulzweida, U.: Sensitivity of simulated climate to
1691
horizontal and vertical resolution in the ECHAM5 atmosphere model, Journal of
1692
Climate, 19, 3771-3791, 10.1175/jcli3824.1, 2006.
1693
Sander, R., Baumgaertner, A., Cabrera-Perez, D., Frank, F., Gromov, S., Grooss, J. U.,
1694
Harder, H., Huijnen, V., Jockel, P., Karydis, V. A., Niemeyer, K. E., Pozzer, A.,
1695
Hella, R. B., Schultz, M. G., Taraborrelli, D., and Tauer, S.: The community
1696
atmospheric chemistry box model CAABA/MECCA-4.0, Geoscientific Model
1697
Development, 12, 1365-1385, 10.5194/gmd-12-1365-2019, 2019.
1698
Schlag, P., Kiendler-Scharr, A., Blom, M. J., Canonaco, F., Henzing, J. S., Moerman,
1699
M., Prévôt, A. S. H., and Holzinger, R.: Aerosol source apportionment from 1-year
1700
measurements at the CESAR tower in Cabauw, the Netherlands, Atmos. Chem.
1701
Phys., 16, 8831-8847, 10.5194/acp-16-8831-2016, 2016.
1702
Seco, R., Peñuelas, J., Filella, I., Llusià, J., Molowny-Horas, R., Schallhart, S., Metzger,
1703
A., Müller, M., and Hansel, A.: Contrasting winter and summer VOC mixing ratios
1704
at a forest site in the Western Mediterranean Basin: the effect of local biogenic
1705
emissions, Atmos. Chem. Phys., 11, 13161-13179, 10.5194/acp-11-13161-2011,
1706
2011.
1707
Seinfeld, J. H. and Pandis, S. N.: Atmospheric Chemistry and Physics: From Air
1708
Pollution to Climate Change, Second, John Wiley & Sons, Inc., Hoboken, New
1709
Jersey2006.
1710
Snider, G., Weagle, C. L., Martin, R. V., van Donkelaar, A., Conrad, K., Cunningham,
1711
D., Gordon, C., Zwicker, M., Akoshile, C., Artaxo, P., Anh, N. X., Brook, J., Dong,
1712
J., Garland, R. M., Greenwald, R., Griffith, D., He, K., Holben, B. N., Kahn, R.,
1713
Koren, I., Lagrosas, N., Lestari, P., Ma, Z., Vanderlei Martins, J., Quel, E. J., Rudich,
1714
Y., Salam, A., Tripathi, S. N., Yu, C., Zhang, Q., Zhang, Y., Brauer, M., Cohen, A.,
1715
Gibson, M. D., and Liu, Y.: SPARTAN: a global network to evaluate and enhance
1716
satellite-based estimates of ground-level particulate matter for global health
1717
applications, Atmos. Meas. Tech., 8, 505-521, 10.5194/amt-8-505-2015, 2015.
1718
Snider, G., Weagle, C. L., Murdymootoo, K. K., Ring, A., Ritchie, Y., Stone, E., Walsh,
1719
A., Akoshile, C., Anh, N. X., Balasubramanian, R., Brook, J., Qonitan, F. D., Dong,
1720
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
63
J., Griffith, D., He, K., Holben, B. N., Kahn, R., Lagrosas, N., Lestari, P., Ma, Z.,
1721
Misra, A., Norford, L. K., Quel, E. J., Salam, A., Schichtel, B., Segev, L., Tripathi,
1722
S., Wang, C., Yu, C., Zhang, Q., Zhang, Y., Brauer, M., Cohen, A., Gibson, M. D.,
1723
Liu, Y., Martins, J. V., Rudich, Y., and Martin, R. V.: Variation in global chemical
1724
composition of PM2.5: emerging results from SPARTAN, Atmos. Chem. Phys., 16,
1725
9629-9653, 10.5194/acp-16-9629-2016, 2016.
1726
Solomon, P. A., Crumpler, D., Flanagan, J. B., Jayanty, R. K. M., Rickman, E. E., and
1727
McDade, C. E.: US National PM2.5 Chemical Speciation Monitoring Networks-
1728
CSN and IMPROVE: Description of networks, J. Air Waste Manage. Assoc., 64,
1729
1410-1438, 10.1080/10962247.2014.956904, 2014.
1730
Souza, P. A. d., Mello, W. Z. d., Mariani, R. L., and Sella, S. M.: Caracterização do
1731
material particulado fino e grosso e composição da fração inorgânica solúvel em
1732
água em São José dos Campos (SP), Química Nova, 33, 1247-1253, 2010.
1733
Stavroulas, I., Bougiatioti, A., Grivas, G., Paraskevopoulou, D., Tsagkaraki, M.,
1734
Zarmpas, P., Liakakou, E., Gerasopoulos, E., and Mihalopoulos, N.: Sources and
1735
processes that control the submicron organic aerosol composition in an urban
1736
Mediterranean environment (Athens): a high temporal-resolution chemical
1737
composition measurement study, Atmos. Chem. Phys., 19, 901-919, 10.5194/acp-
1738
19-901-2019, 2019.
1739
Sun, Y., Xu, W., Zhang, Q., Jiang, Q., Canonaco, F., Prévôt, A. S. H., Fu, P., Li, J.,
1740
Jayne, J., Worsnop, D. R., and Wang, Z.: Source apportionment of organic aerosol
1741
from 2-year highly time-resolved measurements by an aerosol chemical speciation
1742
monitor in Beijing, China, Atmos. Chem. Phys., 18, 8469-8489, 10.5194/acp-18-
1743
8469-2018, 2018.
1744
Sun, Y., He, Y., Kuang, Y., Xu, W., Song, S., Ma, N., Tao, J., Cheng, P., Wu, C., Su,
1745
H., Cheng, Y., Xie, C., Chen, C., Lei, L., Qiu, Y., Fu, P., Croteau, P., and Worsnop,
1746
D. R.: Chemical Differences Between PM1 and PM2.5 in Highly Polluted
1747
Environment and Implications in Air Pollution Studies, Geophysical Research
1748
Letters, 47, e2019GL086288, https://doi.org/10.1029/2019GL086288, 2020.
1749
Sun, Y. L., Wang, Z. F., Fu, P. Q., Yang, T., Jiang, Q., Dong, H. B., Li, J., and Jia, J.
1750
J.: Aerosol composition, sources and processes during wintertime in Beijing, China,
1751
Atmos. Chem. Phys., 13, 4577-4592, 10.5194/acp-13-4577-2013, 2013.
1752
Tiitta, P., Vakkari, V., Croteau, P., Beukes, J. P., van Zyl, P. G., Josipovic, M., Venter,
1753
A. D., Jaars, K., Pienaar, J. J., Ng, N. L., Canagaratna, M. R., Jayne, J. T., Kerminen,
1754
V. M., Kokkola, H., Kulmala, M., Laaksonen, A., Worsnop, D. R., and Laakso, L.:
1755
Chemical composition, main sources and temporal variability of PM<sub>1</sub>
1756
aerosols in southern African grassland, Atmos. Chem. Phys., 14, 1909-1927,
1757
10.5194/acp-14-1909-2014, 2014.
1758
Timonen, H., Carbone, S., Aurela, M., Saarnio, K., Saarikoski, S., Ng, N. L.,
1759
Canagaratna, M. R., Kulmala, M., Kerminen, V.-M., Worsnop, D. R., and Hillamo,
1760
R.: Characteristics, sources and water-solubility of ambient submicron organic
1761
aerosol in springtime in Helsinki, Finland, Journal of Aerosol Science, 56, 61-77,
1762
10.1016/j.jaerosci.2012.06.005, 2013.
1763
Tørseth, K., Aas, W., Breivik, K., Fjæraa, A. M., Fiebig, M., Hjellbrekke, A. G., Lund
1764
Myhre, C., Solberg, S., and Yttri, K. E.: Introduction to the European Monitoring
1765
and Evaluation Programme (EMEP) and observed atmospheric composition change
1766
during 1972–2009, Atmos. Chem. Phys., 12, 5447-5481, 10.5194/acp-12-
1767
5447-2012, 2012.
1768
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
64
Tost, H., Jockel, P. J., Kerkweg, A., Sander, R., and Lelieveld, J.: Technical note: A
1769
new comprehensive SCAVenging submodel for global atmospheric chemistry
1770
modelling, Atmos. Chem. Phys., 6, 565-574, 2006.
1771
Tost, H., Joeckel, P., Kerkweg, A., Pozzer, A., Sander, R., and Lelieveld, J.: Global
1772
cloud and precipitation chemistry and wet deposition: tropospheric model
1773
simulations with ECHAM5/MESSy1, Atmos. Chem. Phys., 7, 2733-2757, 2007.
1774
Tsigaridis, K., Daskalakis, N., Kanakidou, M., Adams, P. J., Artaxo, P., Bahadur, R.,
1775
Balkanski, Y., Bauer, S. E., Bellouin, N., Benedetti, A., Bergman, T., Berntsen, T.
1776
K., Beukes, J. P., Bian, H., Carslaw, K. S., Chin, M., Curci, G., Diehl, T., Easter, R.
1777
C., Ghan, S. J., Gong, S. L., Hodzic, A., Hoyle, C. R., Iversen, T., Jathar, S., Jimenez,
1778
J. L., Kaiser, J. W., Kirkevag, A., Koch, D., Kokkola, H., Lee, Y. H., Lin, G., Liu,
1779
X., Luo, G., Ma, X., Mann, G. W., Mihalopoulos, N., Morcrette, J. J., Mueller, J. F.,
1780
Myhre, G., Myriokefalitakis, S., Ng, N. L., O'Donnell, D., Penner, J. E., Pozzoli, L.,
1781
Pringle, K. J., Russell, L. M., Schulz, M., Sciare, J., Seland, O., Shindell, D. T.,
1782
Sillman, S., Skeie, R. B., Spracklen, D., Stavrakou, T., Steenrod, S. D., Takemura,
1783
T., Tiitta, P., Tilmes, S., Tost, H., van Noije, T., van Zyl, P. G., von Salzen, K., Yu,
1784
F., Wang, Z., Wang, Z., Zaveri, R. A., Zhang, H., Zhang, K., Zhang, Q., and Zhang,
1785
X.: The AeroCom evaluation and intercomparison of organic aerosol in global
1786
models, Atmospheric Chemistry and Physics, 14, 10845-10895, 10.5194/acp-14-
1787
10845-2014, 2014.
1788
Tsimpidi, A. P., Karydis, V. A., Pandis, S. N., and Lelieveld, J.: Global combustion
1789
sources of organic aerosols: model comparison with 84 AMS factor-analysis data
1790
sets, Atmos. Chem. Phys., 16, 8939-8962, 10.5194/acp-16-8939-2016, 2016.
1791
Tsimpidi, A. P., Karydis, V. A., Pozzer, A., Pandis, S. N., and Lelieveld, J.: ORACLE
1792
(v1.0): module to simulate the organic aerosol composition and evolution in the
1793
atmosphere, Geoscientific Model Development, 7, 3153-3172, 10.5194/gmd-7-
1794
3153-2014, 2014.
1795
Tsimpidi, A. P., Karydis, V. A., Pozzer, A., Pandis, S. N., and Lelieveld, J.: ORACLE
1796
2-D (v2.0): an efficient module to compute the volatility and oxygen content of
1797
organic aerosol with a global chemistry-climate model, Geoscientific Model
1798
Development, 11, 3369-3389, 10.5194/gmd-11-3369-2018, 2018.
1799
Vasilakopoulou, C. N., Matrali, A., Skyllakou, K., Georgopoulou, M., Aktypis, A.,
1800
Florou, K., Kaltsonoudis, C., Siouti, E., Kostenidou, E., Błaziak, A., Nenes, A.,
1801
Papagiannis, S., Eleftheriadis, K., Patoulias, D., Kioutsioukis, I., and Pandis, S. N.:
1802
Rapid transformation of wildfire emissions to harmful background aerosol, npj
1803
Climate and Atmospheric Science, 6, 218, 10.1038/s41612-023-00544-7, 2023.
1804
Vignati, E., Wilson, J., and Stier, P.: M7: An efficient size-resolved aerosol
1805
microphysics module for large-scale aerosol transport models, J. Geophys. Res.-
1806
Atmos., 109, doi: 10.1029/2003jd004485, 2004.
1807
Wang, Y., Li, W., Gao, W., Liu, Z., Tian, S., Shen, R., Ji, D., Wang, S., Wang, L.,
1808
Tang, G., Song, T., Cheng, M., Wang, G., Gong, Z., Hao, J., and Zhang, Y.: Trends
1809
in particulate matter and its chemical compositions in China from 2013–2017,
1810
Science China Earth Sciences, 62, 1857-1871, 10.1007/s11430-018-9373-1, 2019.
1811
Weinstein, J. P., Hedges, S. R., and Kimbrough, S.: Characterization and aerosol mass
1812
balance of PM2.5 and PM10 collected in Conakry, Guinea during the 2004
1813
Harmattan period, Chemosphere, 78, 980-988, 10.1016/j.chemosphere.2009.12.022,
1814
2010.
1815
WHO: Health aspects of air pollution with particulate matter, ozone and nitrogen
1816
dioxide : report on a WHO working group, Bonn, Germany 13-15 January 2003,
1817
2003.
1818
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
65
WHO: WHO global air quality guidelines: Particulate matter (PM2.5 and PM10), ozone,
1819
nitrogen dioxide, sulfur dioxide and carbon monoxide, 2021.
1820
WHO: Ambient (outdoor) air pollution: https://www.who.int/news-room/fact-
1821
sheets/detail/ambient-(outdoor)-air-quality-and-health, last access: 15/09/2024.
1822
Xu, L., Suresh, S., Guo, H., Weber, R. J., and Ng, N. L.: Aerosol characterization over
1823
the southeastern United States using high-resolution aerosol mass spectrometry:
1824
spatial and seasonal variation of aerosol composition and sources with a focus on
1825
organic nitrates, Atmos. Chem. Phys., 15, 7307-7336, 10.5194/acp-15-7307-2015,
1826
2015.
1827
Xu, W., Han, T., Du, W., Wang, Q., Chen, C., Zhao, J., Zhang, Y., Li, J., Fu, P., Wang,
1828
Z., Worsnop, D. R., and Sun, Y.: Effects of Aqueous-Phase and Photochemical
1829
Processing on Secondary Organic Aerosol Formation and Evolution in Beijing,
1830
China, Environ Sci Technol, 51, 762-770, 10.1021/acs.est.6b04498, 2017.
1831
Xu, W., Sun, Y., Wang, Q., Zhao, J., Wang, J., Ge, X., Xie, C., Zhou, W., Du, W., Li,
1832
J., Fu, P., Wang, Z., Worsnop, D. R., and Coe, H.: Changes in Aerosol Chemistry
1833
From 2014 to 2016 in Winter in Beijing: Insights From High-Resolution Aerosol
1834
Mass Spectrometry, Journal of Geophysical Research: Atmospheres, 124, 1132-
1835
1147, https://doi.org/10.1029/2018JD029245, 2019.
1836
Yang, Y., Smith, S. J., Wang, H., Lou, S., and Rasch, P. J.: Impact of Anthropogenic
1837
Emission Injection Height Uncertainty on Global Sulfur Dioxide and Aerosol
1838
Distribution, Journal of Geophysical Research: Atmospheres, 124, 4812-4826,
1839
https://doi.org/10.1029/2018JD030001, 2019.
1840
Yao, X. and Zhang, L.: Identifying decadal trends in deweathered concentrations of
1841
criteria air pollutants in Canadian urban atmospheres with machine learning
1842
approaches, Atmos. Chem. Phys., 24, 7773-7791, 10.5194/acp-24-7773-2024, 2024.
1843
Yienger, J. J. and Levy, H.: Empirical-model of global soil-biogenic NOx emissions,
1844
Journal of Geophysical Research-Atmospheres, 100, 11447-11464,
1845
10.1029/95jd00370, 1995.
1846
Zhai, S., Jacob, D. J., Wang, X., Shen, L., Li, K., Zhang, Y., Gui, K., Zhao, T., and
1847
Liao, H.: Fine particulate matter (PM2.5) trends in China, 2013–2018: separating
1848
contributions from anthropogenic emissions and meteorology, Atmos. Chem. Phys.,
1849
19, 11031-11041, 10.5194/acp-19-11031-2019, 2019.
1850
Zhang, F., Xu, L., Chen, J., Yu, Y., Niu, Z., and Yin, L.: Chemical compositions and
1851
extinction coefficients of PM2.5 in peri-urban of Xiamen, China, during June 2009-
1852
May 2010, Atmospheric Research, 106, 150-158, 10.1016/j.atmosres.2011.12.005,
1853
2012.
1854
Zhang, Q., Jimenez, J. L., Canagaratna, M. R., Allan, J. D., Coe, H., Ulbrich, I., Alfarra,
1855
M. R., Takami, A., Middlebrook, A. M., Sun, Y. L., Dzepina, K., Dunlea, E.,
1856
Docherty, K., DeCarlo, P. F., Salcedo, D., Onasch, T., Jayne, J. T., Miyoshi, T.,
1857
Shimono, A., Hatakeyama, S., Takegawa, N., Kondo, Y., Schneider, J., Drewnick,
1858
F., Borrmann, S., Weimer, S., Demerjian, K., Williams, P., Bower, K., Bahreini, R.,
1859
Cottrell, L., Griffin, R. J., Rautiainen, J., Sun, J. Y., Zhang, Y. M., and Worsnop, D.
1860
R.: Ubiquity and dominance of oxygenated species in organic aerosols in
1861
anthropogenically-influenced Northern Hemisphere midlatitudes, Geophys. Res.
1862
Lett., 34, doi: L13801 10.1029/2007gl029979, 2007.
1863
Zhang, Y., Sun, J., Zhang, X., Shen, X., Wang, T., and Qin, M.: Seasonal
1864
characterization of components and size distributions for submicron aerosols in
1865
Beijing, Science China Earth Sciences, 56, 890-900, 10.1007/s11430-012-4515-z,
1866
2013.
1867
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
66
Zhang, Y., Tang, L., Yu, H., Wang, Z., Sun, Y., Qin, W., Chen, W., Chen, C., Ding,
1868
A., Wu, J., Ge, S., Chen, C., and Zhou, H.-c.: Chemical composition, sources and
1869
evolution processes of aerosol at an urban site in Yangtze River Delta, China during
1870
wintertime, Atmospheric Environment, 123, 339-349,
1871
https://doi.org/10.1016/j.atmosenv.2015.08.017, 2015a.
1872
Zhang, Y. J., Tang, L. L., Wang, Z., Yu, H. X., Sun, Y. L., Liu, D., Qin, W., Canonaco,
1873
F., Prévôt, A. S. H., Zhang, H. L., and Zhou, H. C.: Insights into characteristics,
1874
sources, and evolution of submicron aerosols during harvest seasons in the Yangtze
1875
River delta region, China, Atmos. Chem. Phys., 15, 1331-1349, 10.5194/acp-15-
1876
1331-2015, 2015b.
1877
Zhang, Y. M., Zhang, X. Y., Sun, J. Y., Hu, G. Y., Shen, X. J., Wang, Y. Q., Wang, T.
1878
T., Wang, D. Z., and Zhao, Y.: Chemical composition and mass size distribution of
1879
PM<sub>1</sub> at an elevated site in central east China, Atmos. Chem. Phys., 14,
1880
12237-12249, 10.5194/acp-14-12237-2014, 2014.
1881
Zhao, P. S., Dong, F., He, D., Zhao, X. J., Zhang, X. L., Zhang, W. Z., Yao, Q., and
1882
Liu, H. Y.: Characteristics of concentrations and chemical compositions for PM2.5
1883
in the region of Beijing, Tianjin, and Hebei, China, Atmospheric Chemistry and
1884
Physics, 13, 4631-4644, 10.5194/acp-13-4631-2013, 2013.
1885
Zhou, W., Xu, W., Kim, H., Zhang, Q., Fu, P., Worsnop, D. R., and Sun, Y.: A review
1886
of aerosol chemistry in Asia: insights from aerosol mass spectrometer
1887
measurements, Environ Sci Process Impacts, 22, 1616-1653, 10.1039/d0em00212g,
1888
2020a.
1889
Zhou, W., Xu, W., Kim, H., Zhang, Q., Fu, P., Worsnop, D. R., and Sun, Y.: A review
1890
of aerosol chemistry in Asia: insights from aerosol mass spectrometer
1891
measurements, Environmental Science: Processes & Impacts, 22, 1616-1653,
1892
10.1039/D0EM00212G, 2020b.
1893
1894
1895
https://doi.org/10.5194/egusphere-2024-3590
Preprint. Discussion started: 4 December 2024
c
Author(s) 2024. CC BY 4.0 License.
