1. Introduction
The Fog and Aerosol Interaction Research Italy (FAIRARI) campaign 2021/22 was initiated to study aerosols, fog, and aerosol–fog interactions in the Italian Po Valley. Aerosol particles play a key role in many environmental problems, including climate change, as they affect Earth’s radiation budget, both directly through scattering and absorbing solar radiation and indirectly through their influence on clouds (e.g., Kanakidou 2013; Boucher et al. 2013; Ghan and Schwartz 2007). Their sources can be of natural and anthropogenic origin, and aerosol particles are either directly emitted or formed from precursor gases (Hamed et al. 2007) through a process called new particle formation (NPF). In the atmosphere, particles can undergo growth and photochemical aging and swell by taking up atmospheric water. At least half of the cloud condensation nuclei (CCNs), the nuclei on which atmospheric water vapor condenses on to form cloud and fog droplets, originate from NPF (Merikanto et al. 2009; Gordon et al. 2017; Ren et al. 2021). Furthermore, the newly formed particles, usually smaller than 100 nm in diameter, can induce a health risk (Schraufnagel 2020). Therefore, understanding the physical and chemical properties of aerosol particles as well as their interaction with atmospheric water is of great importance.
a. Fog and its interactions with aerosols.
Radiation fog usually forms during the night when the air close to the surface cools and reaches saturation with respect to water, leading to condensation of atmospheric water on available aerosol particles. In polluted areas, radiation fog is typically characterized by low supersaturation and high aerosol concentrations (Bott 1991; Mazoyer et al. 2019). Under these conditions, the available CCNs are in excess with respect to the number of particles that are actually activated to fog droplets, meaning that the number of CCNs is not limiting droplet formation, in contrast to very pristine regions like the Arctic (Reutter et al. 2009; Mauritsen et al. 2011). The droplet number concentration present in fogs in highly polluted environments (as measured by Ghude et al. 2023) is therefore often much larger than in less polluted environments (as measured by Price 2011). Besides the meteorology governing the overall fog life cycle (through parameters like the boundary layer height and wind), aerosol properties, such as concentration or composition, can affect both fog microphysics and lifetime. These aerosol–fog interactions include feedbacks that involve, for example, changes in droplet size distributions, colloidal stability, or sedimentation rates (Jia et al. 2019; Poku et al. 2019; Boutle et al. 2022; Mazoyer et al. 2022). The geographic extent of events of reduced visibility can be characterized by satellite remote sensing (Cermak et al. 2009; Egli et al. 2019) with clear differences between fog and haze: fog exhibits a more whitish covered area (Fig. 1d) while haze exhibits a more brownish and more optically transparent region (Fig. 1c). These differences in optical properties also illustrate the different impacts of fog and haze on the radiative budget within the atmospheric column.
Schematic of targeted aerosol- and fog-related processes and their impact on climate and human health during FAIRARI. The inlays depict examples of (a) the sampled fog water and (b) the observed low visibility. Satellite images in (c) and (d) show an example haze and fog event, respectively (indicated also in Fig. 3). The SPC site is indicated in red. Images are taken from NASA Worldview. The FAIRARI campaign logo is shown in the lower-left corner. The first four highlights of FAIRARI are marked in red.
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
b. The polluted Po Valley: A concern for human health.
The Italian Po Valley is a densely populated area, surrounded by the Alps to the north, the Apennines to the southwest, and the Adriatic Sea to the east. The orography favors radiation fog formation under anticyclonic conditions with a weak southwesterly flow (Egli et al. 2019). The long-lasting events of reduced visibility and fog are associated with a stably stratified lower atmosphere and high surface relative humidity, which can be exacerbated by gas and particulate matter pollution (Hesse 1954; Fuzzi et al. 1992; Cermak et al. 2009; Giulianelli et al. 2014; Gilardoni et al. 2020), affecting both traffic (Leigh et al. 1998; Pagowski et al. 2004, Fig. 1b) and human health (Fig. 1a, Balmes et al. 1989; Decesari et al. 2017; Patel and Rastogi 2018; Lelieveld et al. 2020). Long-term exposure of humans to air pollution is known to cause adverse health effects (Peng et al. 2008; Moradi et al. 2023). As an example, the loss of life expectancy from air pollution is slightly higher than that from tobacco smoking (Lelieveld et al. 2020). However, the effect of fog on air pollution and, therefore, on human health has not yet been quantified. Scavenging of particulate matter by fog droplets and their subsequent removal can reduce the exposure to harmful pollutants during and for short periods after fog episodes. At the same time, studies indicate that fog processing might lead to increased toxicity of particles upon fog dissipation (Decesari et al. 2017; Patel and Rastogi 2018). The Po Valley is one of the most polluted regions in Europe [Daellenbach et al. 2020; European Environment Agency (EEA) 2020], with, for example, annual mean values for particulate matter (PM2.5) (ambient aerosol particles with cut-size diameters D < 2.5 μm) larger than 25 μg m−3 in 2022 at several sites (EEA 2022), exceeding the World Health Organization (WHO) standard by more than a factor of 5. It is heavily impacted by anthropogenic emissions from traffic, industrial and agricultural activities, as well as residential biomass burning (Scotto et al. 2021; Paglione et al. 2020). Similar to other European locations, pollution levels in the Po Valley have shown a decrease with time in recent decades (Bigi and Ghermandi 2016; Gilardoni et al. 2020). At the same time, a decline in fog and low visibility days has been observed in Europe, potentially contributing 10%–20% to Europe’s daytime warming (Vautard et al. 2009). The decline is probably due to both increasing temperatures and improving air quality, and thereby reduced aerosol particle concentrations (Klemm and Lin 2016; Glantz et al. 2022). However, the effect of improved air quality on fog properties and thus on climate is still not fully understood, although some studies have linked decreased aerosol loads to the historical improvement of atmospheric visibility and decreased hazy conditions (Manara et al. 2019).
c. A historical perspective on the long-term observations at SPC.
The San Pietro Capofiume (SPC) measurement site, located in the Po Valley, has been monitoring changes in air pollution and their consequences for fog and atmospheric chemistry since the 1980s (Fuzzi et al. 1983, 1992). In more than 30 years of research and monitoring, the SPC station hosted field experiments focusing, among others, on aerosol activation in fog (Fuzzi et al. 1992; Noone et al. 1992; Frank et al. 1998; Gilardoni et al. 2014), aerosol and fog chemistry (Facchini et al. 2008; Gilardoni et al. 2014), fog processing (Bhandari et al. 2019), gas-to-particle partitioning of semivolatile inorganic and organic compounds (Ricci et al. 1998; Facchini et al. 1999), organic aerosol composition and source apportionment (Gilardoni et al. 2016; Paglione et al. 2020), NPF (Hamed et al. 2007; Lampilahti et al. 2021), and the redox potential of aerosol and fog water (Costabile et al. 2022; Decesari et al. 2017). The first aerosol–fog microphysical observations together with chemical composition measurements were conducted at the SPC site in 1989 (Noone et al. 1992). Since the early 1990s, the liquid water content (LWC) and chemical composition of the fog have been routinely monitored, producing clear trends in changes in atmospheric composition in the area (Giulianelli et al. 2014), while long-term aerosol size distribution measurements have started in 2002 (Hamed et al. 2007). Therefore, SPC is a unique location for atmospheric observations of aerosol and fog anchored in long-term monitoring of atmospheric composition.
d. From molecules to droplets: The FAIRARI campaign.
Knowledge gaps in understanding coupled aerosol–fog interactions result in poor constraints on the role that fog plays in the climate system, modulating air quality in different environments, or in the interactions between future air quality and climate change mitigation measures (Klemm and Lin 2016). Therefore, predicting fog still remains a challenge where not only parameters such as turbulence and vertical extension need better representation but also detailed and real-time cloud droplet spectra are especially needed (Price et al. 2018; Román-Cascón et al. 2019; Westerhuis et al. 2020; Boutle et al. 2022). Recent research efforts, however, have shown the potential of integrating detailed meteorological and microphysical observations with tailored modeling approaches in simulating the processes governing fog formation and evolution (Lee et al. 2016; Mazoyer et al. 2017; Boutle et al. 2018; Schwenkel and Maronga 2019; Liu et al. 2021; Bodaballa et al. 2022). Compared to previous experiments targeting fog formation and aerosol–fog interactions (e.g., Haeffelin et al. 2010; Price et al. 2018; Mazoyer et al. 2019), FAIRARI carried out a more in-depth investigation of the multiphase chemistry of the fog system and connected the fog period with the following period of frequent NPF (Fig. 3b and Fig. A2 in appendix A). FAIRARI was in line with the first field campaigns conducted in the Po Valley in 1989 and 1994 but took advantage of the large advances made in, e.g., the chemical analysis and microphysics measurement techniques that have taken place since these pioneering studies. A unique combination of instrumentation and methods was used to determine the physical and chemical characteristics of gas, aerosol, and fog as well as potential health effects, covering the size range from gas molecules to droplets. Here, we will present an overview of the setup and the observations recorded, as well as first highlights including fog processes, new particle formation, aerosol aging, and their link to remote sensing.
2. The FAIRARI campaign
The FAIRARI campaign took place between November 2021 and May 2022 at the SPC research station in the Italian Po Valley with an intensive observation period (IOP) between February and April 2022 (Fig. A2). The main objectives of the campaign were to
- (i)identify the properties of particle precursor gases, new particle formation, aerosol particles, hydrated aerosol particles, fog droplets (activated aerosol particles), and dried fog droplets (i.e., droplet residuals);
- (ii)identify the impact of fog processing on aerosol composition, as well as the oxidative potential as an emerging proxy of particulate matter toxicity (Bates et al. 2019);
- (iii)use the novel observations to constrain and develop new theoretical models with detailed chemistry and physics for aerosol and hydrometeor populations, coupled with atmospheric dynamics;
- (iv)identify the characteristics of NPF and particle growth during and after the transition from fog to clear periods.
FAIRARI provided detailed measurements on the formation of aerosol particles, their growth and photochemical aging, as well as their potential impact on human health. Further, the aerosol interaction with water vapor and fog, which in turn impacts climate and air quality, was studied. One unique aspect of the campaign was to shed new light on the intersect between gas, particle, and fog phase, by, e.g., studying in detail the composition of aerosol clusters forming from gas phase molecules or by analyzing particles which were directly involved in the formation of fog droplets (Fig. 1). Fog and aerosol properties, as well as particle formation mechanisms, have been investigated previously. However, multiple questions still remain. For example, what controls the two distinct atmospheric states necessary for fog and new particle formation as well as the transition from one state to the other? Or, is there a difference in molecular composition of particles being formed in the atmosphere and/or serving as nuclei to form fog droplets? And if so, what are the respective implications for air quality and human health? In addition, we pose the questions: how is the composition of the gas phase reflected in (i) the occurrence of new particle formation and (ii) the burden of secondary organic aerosol? On a larger scale, we can ask how representative are the in situ observations of fog compared to satellite observations?
At the SPC site, which is located in a rural area 30 km northeast of the city of Bologna, Italy (Figs. 1 and 2), standard atmospheric composition and meteorological parameters are continuously recorded. These include, among others, particle mass concentrations and number size distributions, as well as trace gas parameters. One unique aspect of the site is the long history of conducting regular analysis of the physical and chemical properties of fog water. During FAIRARI, the monitoring activities were complemented by a large suite of dedicated atmospheric measurements to cover the size range and transitions from molecules to droplets. As shown in Fig. 1, this included the characterization of particle clusters at the intersect between gas phase and aerosol particles. During fog, the interstitial and activated aerosol were analyzed separately. One unique aspect of FAIRARI was the analysis of fog residuals, the subpopulation of aerosol particles that were involved in fog droplet formation. The atmospheric processing or aging of the aerosol was studied in dedicated chamber experiments, bridging gas phase processes to particle formation, which potentially can impact fog properties. All compartments of gas, particle, and fog phase are inherently influenced by the meteorology, transport, and surface processes. The latter include the evaporation, radiation, and emissions of trace gases and primary aerosol particles, which are accounted for in the subsequent interpretation of our results. Together with satellite analysis of key fog properties, we present first scientific highlights of FAIRARI, indicated in the schematic of Fig. 1, in section 4.
SPC measurement site, including SU’s aerosol–fog measurement container and the CNR aerosol-gas measurement container as well as three of the filter samplers, the fog water sampler, and the particle volume monitor (PVM). The container dedicated to the NPF measurements is outside of the photo (to the right) but shown together with the smog chamber in Fig. A1.
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
For further details on the experimental setup, including information on instrument type, size resolution, retrieval approaches, and other technical details, the reader is referred to Tables 1 and 2, as well as to appendix A. A schematic of the setup during the FAIRARI campaign is shown in Fig. A1.
Overview of the main measured parameters and instruments using online techniques during FAIRARI. Acronyms are given in Table A3 in appendix A. See Fig. A1 for location and placement of the respective instrument.
Overview of parameters measured during FAIRARI using various offline techniques. Acronyms are given in Table A3 in appendix A.
3. Shift in aerosol and meteorological conditions during FAIRARI
The Italian Po Valley is known for its notoriously high air pollution levels. In the first 2 months of the IOP of FAIRARI, the daily PM2.5 concentration exceeded the WHO air quality threshold of 15 μg m−3 on all except of 4 days until the end of March [mean ± standard deviation: (26 ± 11) μg m−3, Fig. 3a]. From the first of April until the end of FAIRARI at the beginning of May, the PM2.5 concentrations were significantly lower with (8 ± 3) μg m−3, only once exceeding the WHO air quality standard threshold. Interestingly, the particle number concentration stayed elevated (mean ± std: 5200 ± 2600 cm−3 for particles > 7 nm) throughout the campaign and showed no decrease in early spring, in contrast to PM2.5, showing that NPF can sustain the high particle concentrations at SPC and indicating a change in particle sources. The particle number concentration showed a diurnal pattern with concentrations approximately 2 times higher during nighttime compared to daytime periods. Of the measured submicron mass concentrations, organic aerosol (OA) had the highest contribution, followed by nitrate (
Transition from winter to spring as observed by key aerosol, air quality, and meteorological parameters during FAIRARI. (a) Particle number and PM2.5 concentration, including the 24-h threshold for PM2.5 concentration (dashed line) as recommended by WHO (World Health Organization 2021). (b) CS, temperature, radiation, as well as observed fog and NPF events. The retrieved CS threshold for favorable conditions of NPF during FAIRARI is shown as a dashed line. The times of the satellite images shown in Fig. 1 are indicated with dotted lines. The number concentration and CS are given as hourly means (dots) and 24-h running means (lines) and the other parameters are given as daily arithmetic means. It should be noted that the fog measurements were terminated at the beginning of April.
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
Fog was more frequently observed at the beginning of the intensified observation period (Fig. 3b). Most fog events started around midnight, but the duration of fog events varied from 36 min to more than 12 h (see also appendix A). Some fog events were also divided into several events as the measured visibility fluctuated and reached values larger than 1 km. Following the fog classification scheme of Tardif and Rasmussen (2007), none of the 21 fog events measured during the IOP was caused by advection due to the low wind speeds (WSs) measured. However, Lin et al. (2023) state that a grouping based only on wind speed may not be sufficient. Unfortunately, a further classification following Tardif and Rasmussen (2007) is not possible for our dataset due to the lack of vertically resolved meteorological data. Nevertheless, we can conclude that the dominant type of fog during winter in the Italian Po Valley is radiation fog.
The transition from a period with a dominance of fog to a period with increased new particle formation is a unique aspect of FAIRARI that was captured by a wide range of observational methods. It contrasts with the phenomenology of even more polluted environments like the North China Plain, where new particle formation was often found concomitant to, and apparently acted as a driver of, haze formation (Kulmala et al. 2022). However, the exact drivers and underlying processes of this change of chemical regime in the Po Valley are not yet fully understood. As indicated above (Fig. 1), surface processes and meteorology are potential pivotal candidates. Indeed, the change from a fog- to a NPF-dominated period coincided with a subsequent increase in ambient temperature and global surface radiation (Fig. 3b) which can be responsible for, or associated to, increased atmospheric instability, mixing, and ventilation (i.e., reduced stagnation), as well as increased deposition and photochemical sinks. In addition, these parameters influence the formation of aerosols and the partitioning between gas and particle phase. The condensation sink, often used to predict the occurrence of NPF (Kulmala et al. 2012), showed a clear decrease at the end of March, which can as well explain the increased number of NPF events in April. The reason why such a net decrease in PM2.5 and condensation sink has occurred in the face of only a gradual change in the meteorological conditions is still a matter of investigation. It is worth noticing, however, that chemical regimes can respond very nonlinearly to temperature and humidity and be mediated by pH and aerosol liquid water content, especially in the case of the gas-to-particle partitioning of secondary inorganic aerosol compounds (Nenes et al. 2020)—like ammonium nitrate—which account for a major fraction of PM2.5 in the Po Valley. The characterization of aerosol acidity during FAIRARI, therefore, will help to elucidate how the transition from winter to spring in the Po Valley can result in sharp changes in atmospheric aerosol chemical regimes.
4. Highlights from FAIRARI
A comprehensive and detailed observational dataset was successfully recorded during FAIRARI that allows to study the interplay between gas, particle, and droplet phase and will serve as a basis for various modeling activities in the future. Within this section, first highlights from each of the core parts of the campaign will be showcased: the fog, the new particle formation, the controlled chamber study, and a first comparison to new satellite retrievals.
a. Highlight 1—Fog processes.
Between 16 and 24 February 2022, fog developed almost every night (Fig. 4). However, the preconditions during this week were highly variable, with a difference in temperature between sunset and the beginning of the fog event ranging from −0.9 to −12.7 K. In the hours preceding a fog event, the temporal evolution of the measured relative humidity and the relative humidity calculated based on radiative cooling agree well, supporting the hypothesis that mostly radiation fog was observed.
Period with almost daily fog occurrence during FAIRARI. Periods with visibility below 5, 2, and 1 km are shown along all panels by red, gray, and blue shaded areas, respectively, and in (a), additionally as horizontal lines. (a) Visibility and relative humidity. (b) Dried particle number concentration and fog droplet number concentration. (c) Droplet size distribution. (d) Aerosol mass concentration showing nitrate (
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
Similar to previous studies in Paris (Elias et al. 2009; Hammer et al. 2014; Elias et al. 2015) or India (Ghude et al. 2023), visibility slightly below 1 km did not always indicate droplet formation, as the visibility reduction was also strongly influenced by hydrated and not activated aerosol particles. An example is the fog event on 19–20 February 2022, when fog droplet formation started hours after the visibility was already slightly below 1 km but then dropped to values around 0.1 km and droplet (Dd > 2.65 μm) concentrations increased to around 150 cm−3 (Figs. 4a–c). On average, less than 1% of the aerosol particles were activated into droplets during the fog events of this week (comparing the average droplet number concentration of 50 cm−3 to the average total aerosol number concentration of 6300 cm−3, assuming that all particles larger than 2.65 μm are droplets). Furthermore, the droplet number size distribution was generally bimodal with a peak at around 6 μm and one between 10 and 40 μm (Fig. 4c), which is in the range of previous fog measurements (e.g., Lyu et al. 2022; Mazoyer et al. 2022; Ghude et al. 2023). This is also in agreement with observations made at SPC in 1994 by Wendisch et al. (1998), suggesting that it is a characteristic microphysical feature of radiation fogs in this environment. Bimodal fog droplet number size distributions were also found in the Paris basin (Mazoyer et al. 2022), in rural areas of the United Kingdom (Price 2011), Canada (Boudala et al. 2022), and in the Gulf countries (Weston et al. 2022). During the first week of the IOP, a buildup and a subsequent decrease in the aerosol mass concentration was measured (Figs. 4d and B2d), which was also visible in the integrated number concentration of dried aerosol particles larger than 286 nm (Fig. B2b). This buildup, observed between the fog events on 18 and 19 February, could include the effects of fog processing, which typically leads to the formation of secondary inorganic and secondary OA (SOA) formation (Gilardoni et al. 2016; Jia et al. 2023). However, due to the stable atmospheric conditions characterized by low wind speeds, high relative humidity, and the absence of precipitation, a considerable part of the buildup is most likely a result of accumulated pollution. As the daytime wind speed intensifies, starting on 20 February, the ambient mass concentration starts to decrease.
The unique setup during FAIRARI provided measurements of the properties of the gas phase, aerosol, fog, and precipitation that covered almost seven orders of magnitude in size (out of which five are shown in Fig. 5a). The particle number size distribution showed a monomodal distribution with a median mode diameter of around 90 nm for our specific event (Fig. 5a), while the mode diameter showed more variability throughout the campaign ranging between 30 and 150 nm (Fig. B2c). The aerosol composition in the sub-micrometer range was dominated by nitrate and organics (Fig. 4d). Within the fog residuals, particles larger than 300 nm dominated the volume and mass of the aerosol population (Fig. 5b). The OA in the fog residuals mainly consisted of large particles around 600–800 nm [vacuum aerodynamic diameter (Dva)]. In this range, also most of the fog residual nitrate and sulfate particles were observed, indicating that the fog residuals are internally mixed. A smaller component of ammonium nitrate was almost exclusively present at 400–600 nm, which suggests that at least part of the fog residuals was externally mixed. Fog droplets are naturally dominated by water. For our example fog event, the nucleation seed (the solute) of the droplet contributed only 31.5 ppm to the total mass of the droplet (Fig. 5b). Similar to the chemical composition of the ambient aerosol, the composition of the particles that served as fog seeds revealed a dominance of nitrate and organic compounds (Figs. 5b,c). However, compared to the ambient and interstitial aerosol, the fog residuals consisted of a larger fraction of inorganic nitrate and sulfate, while the fractions of less hygroscopic components such as black carbon (BC) and OA were significantly lower. Further investigation of the OA revealed that the fog residuals were moderately oxidized, only showing a slightly higher fraction of ion
Physical and chemical properties of aerosol and fog during the mature phase of fog event 15 (night 23–24 Feb 2022). (a) Size distributions of aerosol particles and fog droplets. The quartiles are given as shaded areas. (b) Median volume size distribution of fog residuals (gray points) and mean chemical composition of the residual mass size distribution. The size ranges of the characterized aerosol and fog are shown as gray and blue areas, respectively. Note that aerosol sizing instruments measure an electrical mobility or optical diameter, while the SP-AMS measures a vacuum aerodynamic diameter. Moreover, the SP-AMS as well as the DMPS 2 data are only corrected for the CVI enrichment factor, not the sampling efficiency. (c) Chemical composition of interstitial aerosol (PM1, by HR-TOF-AMS) and dried droplets (by SP-AMS) given as mass fractions. Interstitial BC refers to multiangle absorption photometer (MAAP) measurements behind the PM2.5 inlet. The SP-AMS determines the refractory BC (rBC) concentration only as a non-size-resolved bulk value. The asterisk marks that the total mass concentration of the dried droplets is only corrected for the CVI enrichment factor, not the sampling efficiency. For a list of instrumental acronyms, see Tables 1 and A3.
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
The mass concentrations of the main interstitial aerosol chemical species were clearly reduced during fog events due to their growth into hydrated aerosol particles and droplets, which are not sampled by the PM2.5 inlet during fog. For example, during the period reported in Fig. 4, the scavenging ratios [i.e., the fraction of mass removed from the PM1 size range at the onset of the fog, calculated comparing the chemical composition of PM1 aerosol during the fog event with that observed during the hour preceding the fog event, following Noone et al. (1992) and Gilardoni et al. (2014)] are (50 ± 11) % for OA, (55 ± 16) % for
b. Highlight 2—New particle formation.
Between 19 and 26 March 2022, frequent new particle formation around noontime was observed, which can be seen from the size distribution evolution (Fig. 6b) and the significantly increasing total number concentration of particles larger than 1.7 nm (Fig. 6a). In March and April, clustering of sub-3-nm particles was frequently observed (39 days). The clustering and subsequent growth of particles toward larger sizes, classified as typical NPF event, was observed on 16 days, occurring more and more frequently toward April (Fig. 3 and Table A1). The measured rates of particle formation (median 87 cm−3 s−1 for 1.7-nm particles) and growth (median 4.6 nm h−1 for 3–7-nm particles) were well in line with previous measurements at the same site (Manninen et al. 2010; Kontkanen et al. 2016). At many sites around the world, new particle formation often occurs around noon (Kerminen et al. 2018) due to enhanced photochemistry. This leads to the production of sulfuric acid (SA)—an important component in the formation of atmospheric clusters together with bases and oxygenated organic molecules (Kulmala et al. 2013; Ehn et al. 2014; Bianchi et al. 2019). During the FAIRARI NPF period (Fig. A2, 1 March–30 April), the concentration of sulfuric acid was 4.6 × 106 cm−3 [1000–1400 local time (LT)], close to that in polluted Chinese megacities (5 × 106 –7 × 106 cm−3; Deng et al. 2020) and significantly higher than in clean environments such as Hyytiälä (9 × 105 cm−3; Nieminen et al. 2014) and the Jungfraujoch (5 × 105 cm−3; Bianchi et al. 2016). High abundance of oxygenated organic molecules (OOMs) (108–109 cm−3, see also Fig. 6c) was also found in the Po Valley region (Fig. S1 in the online supplemental material). Nitrogen-containing OOMs (CHON) composed 60%–70% of the total OOMs, close to the observations in polluted cities such as Nanjing (Nie et al. 2022) and Beijing (Cai et al. 2022; Guo et al. 2022). Such species were also present (albeit to a smaller extent) in the larger particles (Fig. 5c). In addition to the organics and sulfuric acid clusters, halogen species (e.g., I and Cl), likely coming from marine sources, were found. Based on measurements of ions and neutral clusters during NPF events, one can draw the preliminary conclusion that the initial nucleation in the Po Valley is dominated by sulfuric acid and bases such as amines and ammonia, most likely originating from agricultural activities in the region (Cai et al. 2024). OOMs seemed to play a minor role in the formation of sub-3-nm particles due to the lack of extremely low volatility organics (e.g., OOM dimers), but contributed significantly to the following growth (above 3 nm), which has been discussed in Cai et al. (2024). Days with a noontime condensation sink (CS) below 10 × 10−3 s−1 (Fig. 6a) and higher photoactivity toward springtime (March–April) make NPF occurrence more likely in the Po Valley compared to the fog period. The particle growth rates (GR) are faster compared to, for example, Beijing, but range mostly between 1 and 10 nm h−1 as often observed around the globe (Yan et al. 2021; Stolzenburg et al. 2023). This should provide high particle survival probabilities as the condensation sink is moderate (8.9 × 10−3 s−1) and lower compared to heavily polluted environments. Overall, this should lead to a considerable role of NPF in the Po Valley particle number budget in spring. Detailed information on the NPF mechanism in the Po Valley during FAIRARI can be found in Cai et al. (2024).
Physical and chemical properties of NPF during the period with almost daily occurrence. (a) CS and total particle number concentration for particles larger than 1.7 nm, (b) particle number size distribution, and (c) SA monomer, SA dimer, and total OOM concentrations.
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
c. Highlight 3—Ambient air smog chamber study.
A total of 16 ambient air smog chamber experiments, including blanks, were performed between 2 and 17 March 2022. The experiments started at different hours of the day (morning, noon, afternoon, and night) to investigate different states of the Po Valley atmosphere. In all experiments except one, the formation of SOA was observed in the perturbed chamber almost immediately after UV illumination and formation of OH radicals. The SOA produced ranged from 0.1 to 10 μg m−3, while no SOA formation was observed in the reference chamber. The highest concentrations of produced SOA were observed during the nighttime experiments, when the atmosphere close to the surface was more polluted. In the single experiment without SOA formation, new particle formation occurred in the perturbed chamber. During that experiment, the atmosphere in the area was the cleanest (PM1 of around 3 μg m−3). A significant production of ammonium nitrate was observed (in the range of 2–160 μg m−3) in all experiments. Although the addition of nitrous acid (HONO) used for the production of OH radicals contributes significantly to the production of nitric acid, the high levels of ammonium nitrate produced highlight the high concentrations of ammonia in the ambient atmosphere (around 20 ppb). An example of an experiment (10 March 2022) with significant production of both SOA and ammonium nitrate is shown in Fig. 7. These results illustrate the high SOA formation potential of the ambient air at the SPC site, limited only by the availability of atmospheric oxidants. The precursor volatile organic compounds (VOCs) of the formed SOA in the perturbed chamber could not be quantified due to instrumentation issues. For the ambient measurements, the contributions of gaseous compounds like VOCs is currently under evaluation. However, it is expected that the Po Valley is dominated by anthropogenic VOCs like toluene, aromatic compounds, and cyclohexane (Steinbacher et al. 2005a; Decesari et al. 2014). Smaller contribution is observed for the biogenic VOCs, with isoprene concentrations peaking during morning hours at around 2 ppb (Steinbacher et al. 2005b). A detailed analysis of the air smog chamber experiments during FAIRARI can be found in Aktypis et al. (2024).
Ambient air chamber experiment performed during FAIRARI. Mass concentrations of (a) organics, (b) sulfate, (c) ammonium, and (d) nitrate in the perturbed and the reference chambers during the experiment on 10 Mar 2022. All data are wall-loss and collection efficiency corrected.
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
d. Highlight 4—A view from above.
The satellite data show that large regions within the Po Valley were typically covered by coherent fog and low stratus patches during the IOP of the FAIRARI campaign (Fig. 8a, and Figs. S2 and S3). However, many events did not directly cover the measurement site SPC (Figs. S3 and S4). During the IOP, a large variability in the spatial distribution of the fog layers is apparent at the time of Moderate resolution Imaging Spectroradiometer (MODIS) overpasses (around 1000–1100 LT, Fig. S4) which can be explained by dissipation occurring heterogeneously over the area.
Satellite retrieval of fog properties during the IOP of FAIRARI. (a) Number of days with FLS cover present for more than 30 min for the duration of the example fog week during the FAIRARI campaign (16–24 Feb 2022). The “X” marks the measurement site SPC, and the letters T, M, B, and V stand for Torino, Milan, Bologna, and Venice, respectively. The data are retrieved from the SEVIRI on board the MSG. (b) ED measured in situ (blue) and retrieved from Terra/MODIS satellite data (gray). Only two of the events detected by the satellite were classified as a fog event at the measurement site SPC (blue shaded). The boxes represent the 25th and 75th percentiles and the whiskers the 5th and 95th percentiles. The median is given in solid; the mean is given in dashed. The number of data points used is given at the bottom.
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
The key physical properties of fog, such as the effective diameter (ED) of the fog droplets, can not only be retrieved using in situ measurements but also by satellite-based remote sensing. However, for polar orbiting satellites, such as Terra or Aqua, this can only be done when the fog period coincides with a temporally and spatially close overpass of the satellite. During the IOP, four satellite overpasses detected fog over and/or near the site of SPC. Two events on 18 and 24 February (Fig. 8b) showed coinciding fog directly at the SPC site. Interestingly, both events were characterized by similar values for the differently derived ED, giving further confidence in the satellite retrieval algorithm (the second event shows higher variability in the in situ data due to more patchy fog on that particular day). Meanwhile, the satellite-retrieved cloud droplet number concentration (CDNC) is overestimated in both the cases compared to in situ measurements (Fig. S5). The higher value of CDNC could be due to the adiabatic assumption in deriving CDNC, which uses the cloud optical thickness and effective radius following Quaas et al. (2006).
5. Summary and outlook
FAIRARI was initiated to investigate the properties and drivers of fog and aerosol as well as their respective formation mechanisms in one of the most polluted and densely populated areas in Europe, the Italian Po Valley. The abundance of fog and aerosols has adverse impacts on human health, visibility, and climate. At the site of San Pietro Capofiume, long-term observations show decreasing trends in fog occurrence and aerosol concentrations over the last decades, which are in part due to changes in meteorology and reductions in anthropogenic aerosol emissions. However, not all of the physical and chemical mechanisms behind the formation of particles or fog are fully understood yet. Using recent developments in aerosol technology, we characterized numerous fog and new particle formation events, complemented by controlled outdoor chamber experiments and remote sensing analysis. Fog droplet measurements often showed a bimodal particle number size distribution. The reduction in visibility during fog was not only caused by an increase in the extinction of light by fog droplets but was also significantly affected by the contribution of hydrated but not activated aerosol particles. Nitrate and organics contributed the most to the mass of aerosol particles, with nitrate dominating during the fog period and organics during the NPF period. During fog events, the mass of the dried particles consisted of a larger fraction of inorganic species, such as nitrates, sulfates, or ammonia, and a significantly smaller fraction of black carbon compared to periods without fog. Within the organic aerosol mass fraction of the residuals, an increased contribution of nitrogen-containing organic compounds was observed. During the course of FAIRARI, we observed a clear transition in the atmosphere and its chemical regimes driven by changes in meteorology, aerosol emissions, and other key aerosol parameters (e.g., the observed decrease in condensation sink). These changes lead to more fog episodes in winter, while spring was dominated by new particle formation events. Out of 39 days with clustering of sub-3-nm particles, 16 clear NPF event days were identified showing subsequent particle growth. The formation of new particles was generally dominated by sulfuric acid and bases such as amines and ammonia, originating from anthropogenic activities. Dedicated chamber experiments showed that the concentration of produced secondary OA (SOA) was highest when the atmosphere was more polluted. The satellite analysis revealed the large-scale impact of fog on the Italian Po Valley. It agreed well with in situ measurements of droplet effective diameter and fog occurrence but not with the cloud droplet number concentration. Future work will include a more detailed analysis of the comprehensive gas-phase, aerosol, and fog observations performed during FAIRARI. These efforts will be used to inform modeling studies on different scales to explore, for example, the role of cloud microphysical processes such as droplet sedimentation in the regulation of fog properties, as well as the influence of thermodynamic and chemical parameters (e.g., temperature, soil moisture, and aerosol concentrations) on long-term fog trends.
Acknowledgments.
Financial support from the European Union’s Horizon 2020 research and innovation programme (Project FORCeS under Grant Agreement 821205 and “NPF-PANDA” under Grant Agreement 895875), European Research Council (Consolidator Grant INTEGRATE 865799), Knut and Alice Wallenberg Foundation (Grant 2021.0169 and 2021.0298), Finnish Research Council (Projects 356134, 346370, and 325656), and Viennese Vienna Science and Technology Fund (Project VRG22-003) is gratefully acknowledged. This article is part of a project supported by the European Commission under the Horizon 2020—Research and Innovation Framework Programme, H2020-INFRAIA-2020-1, Grant Agreement 101008004, which supported transnational access to the San Pietro Capofiume measurement site. The authors acknowledge the use of imagery from the NASA Worldview application (https://worldview.earthdata.nasa.gov), part of the NASA Earth Observing System Data Information System (EOSDIS). We thank Gabriel Freitas for designing our campaign logo. Acknowledgments are given to Tabea Hennig, Kai Rosman, Birgitta Noone, and Zhara Hamzavi for their technical support in the preparation of the campaign. We thank Luca Di Liberto and David Hadden for their excellent technical support in the field.
Data availability statement.
The data from the ambient air smog chamber experiment can be found at https://doi.org/10.5281/zenodo.10621120. The data used in the satellite analysis (MODIS) are from the MODIS level 2 Atmosphere Joint Products. They are available for download from the Level 1 and Atmosphere Archive and Distribution System Distributed Active Archive Center (LAADS DAAC). The daily MODIS data can be found at https://doi.org/10.5067/MODIS/MYDATML2.006 (Aqua). All other data and metadata from FAIRARI are and/or will be available at the database of the Bolin Centre for Climate Research (https://bolin.su.se/data/) using the keyword “FAIRARI” or the FAIRARI data summary page at https://doi.org/10.17043/fairari-2021-2022.
APPENDIX A FAIRARI Campaign Description
a. San Pietro Capofiume field station.
The campaign took place at the SPC research site (44.65°N, 11.62°E, 5 m elevation, Figs. 1 and 2), which is located about 30 km northeast of Bologna, Italy. The measurement site is situated in an agricultural area with some smaller villages nearby. The SPC site is representative of the regional background of the eastern Po Valley. A schematic of the setup of permanent and temporal installations during FAIRARI is given in Fig. A1 and is briefly described below. A timeline of the campaign with main instrumental deployments and performed experiments is shown in Fig. A2. The campaign averages of key parameters for all main fog and new particle formation events are given in Tables A1 and A2, respectively. The list of acronyms is given in Table A3.
Overview of the experimental setup during the IOP of FAIRARI. (a) Aerosol-gas phase measurements, (b) aerosol–fog measurements, (c) NPF measurements, and (d) aerosol aging experiments. Instruments behind a drier are indicated by a light gray background. Inlets/tubings of ambient air are colored, and sampling lines with aged gas/aerosol particles are shown in black. Ambient measurements are colored blue. The dashed lines in (b) indicate that the instruments were sampling from the CVI inlet during fog; otherwise, the instruments were sampling from the whole-air inlet. A list of acronyms is given in Table A3.
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
Timeline of the FAIRARI campaign. (a) Setups and (b) conducted SOA formation experiments, as well as collected fog water and filter samples. SU basic refers to the measurements by the GFAS, MPS, SP-AMS, Vocus, and Filter Inlet for Gases and Aerosols coupled to a chemical ionization time-of-flight mass spectrometer (FIGAERO-CIMS) instruments before the arrival of the aerosol–cloud container. The SP-AMS, GFAS, and MPS were later installed in the SU aerosol–cloud container, while Vocus and FIGAERO-CIMS remained at their original locations (see Fig. A1). The IOP refers to time between mid-February and end of April, when the majority of instrumentation was running in parallel.
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
b. Monitoring and continuous measurements at SPC.
Atmospheric composition and meteorological parameters are continuously recorded at the site, both within the routine monitoring program of the Regional Agency for Prevention, Environment and Energy of Emilia-Romagna (ARPAE) and by the National Research Council of Italy–Institute of Atmospheric Sciences and Climate (CNR-ISAC) network. Those include, among others, particle mass concentrations such as PM1 and PM2.5, trace gas parameters such as SO2, O3, NOx, and NH3, as well as standard meteorological parameters. An automated ceilometer delivers profiles of the boundary layer structure as well as cloud and fog properties. During the FAIRARI campaign, radiometric measurements of both shortwave and longwave (both downward and upward) radiations were performed.
Within the CNR measurement container (Figs. 2 and A1a), the concentrations and composition of trace gases and sub-micrometer aerosol particles were measured using a PM2.5 inlet. During fog events, the PM2.5 inlet sampled only interstitial (unactivated) particles. The mass concentrations of the nonrefractory sulfate, nitrate, ammonium, chloride, and organics of the sub-micrometer particles (NR-PM1) and of black carbon were measured, as well as the particle mass size distribution. Additionally, the molecular-level chemical composition of semivolatile species in the particulate and gaseous phase was measured, as well as the chemical composition of VOCs. Furthermore, the particle number and mass concentrations of the coarse-mode aerosol were measured with a direct, short, and not dried inlet on the roof of the CNR container. The LWC measurements and fog water analysis are performed routinely.
c. Aerosol–cloud laboratory.
As part of the FAIRARI campaign, a new mobile laboratory developed by Stockholm University (SU), dedicated to studying aerosol–cloud interactions, was installed at the SPC site. The setup (Figs. 2 and A1b) includes a two-inlet system, consisting of a whole-air inlet and a CVI inlet, allowing parallel characterization of total dried aerosol and dried fog droplets (i.e., fog residuals), respectively. The CVI inlet, situated in a wind tunnel, is operated with dried (RH around 20%) and particle-free air, which is heated to around 40°C. This process leads to the evaporation of water from the sampled fog droplets leaving the residuals (i.e., in an ideal activation case, the original CCN) which are successively analyzed by the various in situ instruments. In parallel, measurements of meteorology, fog, and precipitation properties were made on the roof of the laboratory container using various in situ probes. This included the ambient particle size distribution as well as the fall velocity and the size distribution of large hydrometeors (diameters > 50 μm). Additionally, the aerosol size distribution, equivalent black carbon concentrations, and the droplet activation of the total dried aerosol were continuously measured. At a visibility below 1 km [following the definition of fog given by the World Meteorological Organization (WMO)], certain instruments were switched from measuring behind the whole-air inlet to sampling behind the CVI inlet and thereby determining essential fog residual characteristics: the size distribution, the chemical composition, and the droplet reactivation. Moreover, during fog, the filter samples of fog residuals were taken to analyze the chemical composition of residuals offline at the molecular level. Detailed information on the parameters measured related to the mobile aerosol–cloud laboratory can be found in Table 1.
d. Offline sampling.
To complement the online measurements with detailed chemical analyses of aerosol and fog, a set of offline sampling systems was deployed at the SPC field site. PM1 and PM10 were sampled on quartz fiber filters twice a day to better capture fog and nonfog periods (Fig. A2b). Water-soluble inorganic ions and organic acids were analyzed according to the methods described in Sandrini et al. (2016), and low-molecular-weight alkyl amines were analyzed according to Facchini et al. (2008). Furthermore, water-soluble organic carbon (WSOC) content was quantified and PM1 was collected on separate filters for subsequent Fourier transform infrared spectroscopy (FTIR) characterization of OA.
Fog water samples were collected when the LWC exceeded 0.08 g m−3 (a threshold corresponding to a visibility of approximately 200 m; Tomasi and Tampieri 1976). Direct pH measurements of fog samples were carried out and the same analytical techniques were applied as for PM1 filters. In addition, the inorganic and organic ion contribution was characterized and the fog water samples were resuspended and analyzed with respect to their chemical composition (Table 2).
To assess the potential health effects, the PM1, PM10, and fog water samples were analyzed for the oxidative potential (OP) of the aerosol (Charrier and Anastasio 2012). The absorbance was measured and the volume-normalized OP was calculated for all samples, while the mass normalized OP is available for the fog and PM1 samples.
Ambient gaseous SO2, HNO3, HNO2, and NH3 and associated particulate nitrate, sulfate, and ammonium of PM2.5 were determined. To study aerosol partitioning in/out of fog, for some of the days (15 in total), two filters per day were collected, separating day and night periods, and generally corresponding to out-of-fog and in-fog conditions, respectively. Other samples (13 in total) were collected daily to cover a longer period and to catch late spring conditions.
e. New particle formation setup.
To characterize the NPF events, a suite of instruments was installed in an additional container (Fig. A1c and Table 1). Direct aerosol gas-phase precursors and initial clusters were characterized with respect to their composition. In addition, the NPF events were analyzed with respect to their frequency of occurrence, formation strength, and growth rate. Moreover, the chemical composition of the naturally charged cluster ions was measured, as well as the concentration of H2SO4 and OOMs. At the same location, the particle size distributions were also determined from the nucleation to the accumulation mode.
f. Ambient air smog chamber.
The Foundation for Research and Technology Hellas Mobile Dual Atmospheric Chamber System (FORTH-MSC) was deployed at the SPC site to quantify the potential of ambient air masses to form secondary organic and inorganic aerosol. The experimental setup (Fig. A1d) consisted of two identical polytetrafluoroethylene (PTFE) chambers (1.3 m3 each), five UV light panels, and a series of instruments measuring the concentration of gases and particles inside the chambers. The chambers’ relative UV light intensity (
g. Satellite remote sensing.
Satellite observations and, in particular, geostationary satellites can provide a coherent view of the spatial distribution and development of fog and help to contextualize detailed in situ measurements. A main challenge for the satellite retrieval of fog properties is the separation between fog and other low-level stratiform clouds, in particular when relying on passive-sensor imagery. Although attempts have been made to achieve this separation using physical information and machine learning approaches (Cermak and Bendix 2011; Egli et al. 2018), the errors in estimating the altitude of the cloud base are typically in the order of a few hundred meters, and thus not accurate enough to always confidently identify fog. However, the separation between fog and cloud is unnecessary for many research questions. Thus, fog and low stratus (FLS) clouds are typically treated as one category in remote sensing. For the purpose of illustrating the scale of the fog events during the IOP of the FAIRARI campaign, a well-established FLS detection technique based on daytime (here: 0800–1700 LT) geostationary satellite observations is used (Cermak and Bendix 2008). Data from the Spinning Enhanced Visible and Infrared Imager (SEVIRI) on board the Meteosat Second Generation (MSG) are used to detect FLS based on a series of spectral thresholds and object-oriented tests at the native resolution of the imager (3-km nadir, every 15 min).
Instruments on board polar-orbiting satellites, such as the MODIS (Platnick et al. 2017) on board the Terra and Aqua satellites, can provide regional and global coverage of relevant fog properties at 1-km spatial resolution (Bendix et al. 2005). However, the inherent limitation of passive remote sensing measurements on board of MODIS satellites results in reduced temporal resolution of retrievals. During the FAIRARI campaign, only a few fog events lasted until the MODIS overpass and were evaluated with in situ measurements.
h. Fog classification.
Depending on the specific research question to be addressed, different definitions of fog have been applied in the literature (Spänkuch et al. 2022). Visibility is one of the most common parameters used to define the presence of fog. According to Vautard et al. (2009), it can be distinguished between haze (visibility below 5 km), mist (visibility below 2 km), and fog (visibility below 1 km), which is the most widely used definition of fog (e.g., Tardif and Rasmussen 2007; Vautard et al. 2009; Haeffelin et al. 2010; Maier et al. 2013; Hammer et al. 2014; Mazoyer et al. 2019, 2022). However, in polluted regions, it is important to also include other parameters such as the LWC to distinguish between low visibility due to high levels of pollution or due to fog droplets (Figs. 4 and B1; Elias et al. 2009; Hammer et al. 2014; Elias et al. 2015; Ghude et al. 2023). The specific classification of fog periods also depends on the temporal resolution of the measured parameters of interest. In this work, we define fog with constant visibility below 1 km for at least 36 min (based on the temporal resolution of the particle size distribution measurements). Following Elias et al. (2009) and Vautard et al. (2009), the nonfog regime is defined by a visibility higher than 5 km. We assume that aerosol particles larger than 2.65 μm have been activated to fog droplets as the differentiation between hydrated and activated aerosol particles during FAIRARI is ongoing work. This choice [based on our instrument, ground-based fog and aerosol spectrometer (GFAS), and its low gain channel] might include nonactivated, hydrated aerosol particles and therefore result in an overestimation of CDNC and a minor underestimation of ED.
APPENDIX B Fog Statistics and Overview of Aerosol Measurements
The fog events during FAIRARI showed a great variability, both with respect to visibility and liquid water content (Fig. B1). During most fog events, visibility was below 200 m. SPC’s meteorological parameters and aerosol levels exhibit a clear diurnal pattern, with peak aerosol particle number and mass concentration occurring at night (Fig. B2) when the planetary boundary layer is shallow. During the campaign, only a few rain events (using a threshold of 0.4 mm h−1 as suggested by Elias et al. 2015) were recorded (Fig. B2a), and none of them coincided with fog. During the night, the median particle number concentration was about 6700 cm−3 [interquartile range (IQR): 5000–8000 cm−3, diameter range: 13–792 nm] and about 2 times higher than during the day, on average for the entire campaign. The mode diameter of the monomodal dry particle size distribution was largest (median of 140 nm) around 1400 LT. There was a clear dependence of the particle number concentration on the wind direction (WD), with higher concentrations during southwesterly winds (240°–260°, median: 6700 cm−3, IQR: 5000–8400 cm−3), indicating a contribution from the city of Bologna compared to the cleaner air during northeasterly winds (40°–60°, median: 4200 cm−3, IQR: 2600–6500 cm−3). Of the measured submicron mass concentrations, OA had the highest contribution (median of 4.3 μg m−3, IQR: 2.4–6.9 μg m−3), followed by
Fog statistics. (a) Visibility and (b) LWC during the fog events. The boxes represent the 25th and 75th percentiles; the whiskers represent the 5th and 95th percentiles. The median is given in orange; the mean is given in dashed green.
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
Time series of main meteorological and aerosol parameters at the surface. (a) Visibility (black) and rain intensity (light blue), (b) number concentration of total (13–792 nm) dried aerosol particles (red) and of dried aerosol particles > 286 nm (black dotted), (c) size distribution of the dried aerosol particles (13–792 nm), and (d) chemical composition of the dried aerosol particles (during fog events: composition of the dried droplets). The rain intensity is based on the 1-min data given by the instrument. All other data are based on 1-h median data.
Citation: Bulletin of the American Meteorological Society 106, 1; 10.1175/BAMS-D-23-0166.1
Main fog events during FAIRARI with 25%, 50%, and 75%-percentiles, and arithmetic mean value of CDNC, with droplet diameter Dd > 2.65 μm, ED, with Dd > 2.65 μm, and LWC, as well as arithmetic mean WD and WS. The time is given in hour:minute (LT) day:month in 2022. Fog microphysical parameters are only calculated for fogs with a median LWC ≥ 0.02 g m−3.
Main NPF events with growth during FAIRARI. Formation (J) and GR rates are given for the specified nanometer ranges. The time is given in hour:minute (LT) day:month in 2022.
List of acronyms.
References
Aalto, P., and Coauthors, 2001: Physical characterization of aerosol particles during nucleation events. Tellus, 53B, 344–358, https://doi.org/10.1034/j.1600-0889.2001.530403.x.
Aktypis, A., and Coauthors, 2024: Formation and chemical evolution of secondary organic aerosol in two different environments: A dual-chamber study. Atmos. Chem. Phys., 24, 13769–13791, https://doi.org/10.5194/acp-24-13769-2024.
Balmes, J. R., J. M. Fine, T. Gordon, and D. Sheppard, 1989: Potential bronchoconstrictor stimuli in acid fog. Environ. Health Perspect., 79, 163–166, https://doi.org/10.1289/ehp.8979163.
Bates, J. T., and Coauthors, 2019: Review of acellular assays of ambient particulate matter oxidative potential: Methods and relationships with composition, sources, and health effects. Environ. Sci. Technol., 53, 4003–4019, https://doi.org/10.1021/acs.est.8b03430.
Bendix, J., B. Thies, J. Cermak, and T. Nauß, 2005: Ground fog detection from space based on MODIS daytime data—A feasibility study. Wea. Forecasting, 20, 989–1005, https://doi.org/10.1175/WAF886.1.
Bhandari, J., and Coauthors, 2019: Extensive soot compaction by cloud processing from laboratory and field observations. Sci. Rep., 9, 11824, https://doi.org/10.1038/s41598-019-48143-y.
Bianchi, F., and Coauthors, 2016: New particle formation in the free troposphere: A question of chemistry and timing. Science, 352, 1109–1112, https://doi.org/10.1126/science.aad5456.
Bianchi, F., and Coauthors, 2019: Highly oxygenated organic molecules (HOM) from gas-phase autoxidation involving peroxy radicals: A key contributor to atmospheric aerosol. Chem. Rev., 119, 3472–3509, https://doi.org/10.1021/acs.chemrev.8b00395.
Bigi, A., and G. Ghermandi, 2016: Trends and variability of atmospheric PM2.5 and PM10–2.5. Atmos. Chem. Phys., 16, 15 777–15 788, https://doi.org/10.5194/acp-16-15777-2016.
Bodaballa, J. K., I. Geresdi, S. D. Ghude, and I. Salma, 2022: Numerical simulation of the microphysics and liquid chemical processes occur in fog using size resolving bin scheme. Atmos. Res., 266, 105972, https://doi.org/10.1016/j.atmosres.2021.105972.
Bott, A., 1991: On the influence of the physico-chemical properties of aerosols on the life cycle of radiation fogs. Bound.-Layer Meteor., 56 (1), 1–31, https://doi.org/10.1007/BF00119960.
Boucher, O., and Coauthors, 2013: Clouds and aerosols. Climate Change 2013: The Physical Science Basis. T. F. Stocker et al., Eds., Cambridge University Press, 571–657.
Boudala, F. S., D. Wu, G. A. Isaac, and I. Gultepe, 2022: Seasonal and microphysical characteristics of fog at a northern airport in Alberta, Canada. Remote Sens., 14, 4865, https://doi.org/10.3390/rs14194865.
Boutle, I., J. Price, I. Kudzotsa, H. Kokkola, and S. Romakkaniemi, 2018: Aerosol–fog interaction and the transition to well-mixed radiation fog. Atmos. Chem. Phys., 18, 7827–7840, https://doi.org/10.5194/acp-18-7827-2018.
Boutle, I., and Coauthors, 2022: Demistify: A large-eddy simulation (LES) and single-column model (SCM) intercomparison of radiation fog. Atmos. Chem. Phys., 22, 319–333, https://doi.org/10.5194/acp-22-319-2022.
Cai, J., and Coauthors, 2022: Influence of organic aerosol molecular composition on particle absorptive properties in autumn Beijing. Atmos. Chem. Phys., 22, 1251–1269, https://doi.org/10.5194/acp-22-1251-2022.
Cai, J., and Coauthors, 2024: Elucidating the mechanisms of atmospheric new particle formation in the highly polluted Po Valley, Italy. Atmos. Chem. Phys., 24, 2423–2441, https://doi.org/10.5194/acp-24-2423-2024.
Cermak, J., and J. Bendix, 2008: A novel approach to fog/low stratus detection using Meteosat 8 data. Atmos. Res., 87, 279–292, https://doi.org/10.1016/j.atmosres.2007.11.009.
Cermak, J., and J. Bendix, 2011: Detecting ground fog from space—A microphysics-based approach. Int. J. Remote Sens., 32, 3345–3371, https://doi.org/10.1080/01431161003747505.
Cermak, J., R. M. Eastman, J. Bendix, and S. G. Warren, 2009: European climatology of fog and low stratus based on geostationary satellite observations. Quart. J. Roy. Meteor. Soc., 135, 2125–2130, https://doi.org/10.1002/qj.503.
Charrier, J. G., and C. Anastasio, 2012: On dithiothreitol (DTT) as a measure of oxidative potential for ambient particles: Evidence for the importance of soluble transition metals. Atmos. Chem. Phys., 12, 9321–9333, https://doi.org/10.5194/acp-12-9321-2012.
Costabile, F., and Coauthors, 2022: On the redox-activity and health-effects of atmospheric primary and secondary aerosol: Phenomenology. Atmosphere, 13, 704, https://doi.org/10.3390/atmos13050704.
Daellenbach, K. R., and Coauthors, 2020: Sources of particulate-matter air pollution and its oxidative potential in Europe. Nature, 587, 414–419, https://doi.org/10.1038/s41586-020-2902-8.
DeCarlo, P. F., and Coauthors, 2006: Field-deployable, high-resolution, time-of-flight aerosol mass spectrometer. Anal. Chem., 78, 8281–8289, https://doi.org/10.1021/ac061249n.