Particle Number Concentration Data Research Articles

Harnessing AI to unmask Copenhagen's invisible air pollutants: A study on three ultrafine particle metrics

Ultrafine particles (UFPs) are airborne particles with a diameter of less than 100 nm. They are emitted from various sources, such as traffic, combustion, and industrial processes, and can have adverse effects on human health. Long-term mean ambient average particle size (APS) in the UFP range varies over space within cities, with locations near UFP sources having typically smaller APS. Spatial models for lung deposited surface area (LDSA) within urban areas are limited and currently there is no model for APS in any European city. We collected particle number concentration (PNC), LDSA, and APS data over one-year monitoring campaign from May 2021 to May 2022 across 27 locations and estimated annual mean in Copenhagen, Denmark, and obtained additionally annual mean PNC data from 6 state-owned continuous monitors. We developed 94 predictor variables, and machine learning models (random forest and bagged tree) were developed for PNC, LDSA, and APS. The annual mean PNC, LDSA, and APS were, respectively, 5523 pt/cm3, 12.0 μm2/cm3, and 46.1 nm. The final R2 values by random forest (RF) model were 0.93 for PNC, 0.88 for LDSA, and 0.85 for APS. The 10-fold, repeated 10-times cross-validation R2 values were 0.65, 0.67, and 0.60 for PNC, LDSA, and APS, respectively. The root mean square error for final RF models were 296 pt/cm3, 0.48 μm2/cm3, and 1.60 nm for PNC, LDSA, and APS, respectively. Traffic-related variables, such as length of major roads within buffers 100–150 m and distance to streets with various speed limits were amongst the highly-ranked predictors for our models. Overall, our ML models achieved high R2 values and low errors, providing insights into UFP exposure in a European city where average PNC is quite low. These hyperlocal predictions can be used to study health effects of UFPs in the Danish Capital.

Application of the Bayesian spline method to analyze real-time measurements of ultrafine particle concentration in the Parisian subway.

Air pollution in subway environments is a growing concern as it often exceeds WHO recommendations for indoor air quality. Ultrafine particles (UFP), for which there is still no regulation nor a standardized exposure monitoring method, are the strongest contributor to this pollution when the number concentration is used as exposure metric. We aimed to assess the real-time UFP number concentration in the personal breathing zone (PBZ) of three types of underground Parisian subway professionals and analyze it using a novel Bayesian spline approach. Consecutively, we investigated the effect of job, week day, subway station, worker location, and some further events on UFP number concentrations. The data collection procedure originated from a longitudinal study and lasted for a total duration of 6 weeks (from October 7 to November 15, 2019, i.e. two weeks per type of subway professionals). Time-series were built from the real-time particle number concentration (PNC) measured in the PBZ of professionals during their work-shifts. Complementarily, contextual information expressed as Station, Environment, and Event variables were extracted from activity logbooks completed for every work-shift. A Bayesian spline approach was applied to model the PNC within a Bayesian framework as a function of the mentioned contextual information. Overall, the Bayesian spline method suited a real-time personal PNC data modeling approach. The model enabled estimating the differences in UFP exposure between subway professionals, stations, and various locations. Our results suggest a higher PNC closer to the subway tracks, with the highest PNC on subway station platforms. Studied event and week day variables had a lesser influence. It was shown that the Bayesian spline method is suitable to investigate individual exposure to UFP in underground subway settings. This method is informative for better documenting the magnitude and variability of UFP exposure, and for understanding the determinants in view of further regulation and control of this exposure.

Emission characterization of size-resolved particles in a pre-school classroom in relation to children’s activities

The objective of this study was to investigate the effect of children’s daily activity patterns on the size-resolved indoor particles in a pre-school classroom, qualitatively and quantitatively. Real-time particle number size distribution data in conjunction with activity were collected experimentally in a pre-school classroom during two seasons: late spring and early winter. We divided daily activities into six groups: Arrival, Large Group, Small Group, Naptime, Dismissal and Unoccupied. The concentration variation forms of coarse particles were compared qualitatively among activity groups. The results indicated that noteworthy coarse particle concentration variations were observed during each activity group. These variations were caused by four factors: presence/absence of children, variation in the number of children, activity intensity and presence of sources. The emission rates were estimated to quantify the impact of activities. Time series emission rates were deduced from the piecewise-constant inputs and outputs model. The piecewise-constant emission rates were averaged for sequential parts to reduce noise and ensure daily repeated peaks were not missed. The sequential part-constant emission rates were consistent with previously described particle number concentration variation results and the predicted indoor particle number concentration data with sequential part-constant emission rates agreed well with measured data.

Estimate of main local sources to ambient ultrafine particle number concentrations in an urban area

Quantifying and apportioning the contribution of a range of sources to ultrafine particles (UFPs, D<100nm) is a challenge due to the complex nature of the urban environments. Although vehicular emissions have long been considered one of the major sources of ultrafine particles in urban areas, the contribution of other major urban sources is not yet fully understood. This paper aims to determine and quantify the contribution of local ground traffic, nucleated particle (NP) formation and distant non-traffic (e.g. airport, oil refineries, and seaport) sources to the total ambient particle number concentration (PNC) in a busy, inner-city area in Brisbane, Australia using Bayesian statistical modelling and other exploratory tools. The Bayesian model was trained on the PNC data on days where NP formations were known to have not occurred, hourly traffic counts, solar radiation data, and smooth daily trend. The model was applied to apportion and quantify the contribution of NP formations and local traffic and non-traffic sources to UFPs. The data analysis incorporated long-term measured time-series of total PNC (D≥6nm), particle number size distributions (PSD, D=8 to 400nm), PM2.5, PM10, NOx, CO, meteorological parameters and traffic counts at a stationary monitoring site. The developed Bayesian model showed reliable predictive performances in quantifying the contribution of NP formation events to UFPs (up to 4×104particlescm−3), with a significant day to day variability. The model identified potential NP formation and no-formations days based on PNC data and quantified the sources contribution to UFPs. Exploratory statistical analyses show that total mean PNC during the middle of the day was up to 32% higher than during peak morning and evening traffic periods, which were associated with NP formation events. The majority of UFPs measured during the peak traffic and NP formation periods were between 30–100nm and smaller than 30nm, respectively. To date, this is the first application of Bayesian model to apportion different sources contribution to UFPs, and therefore the importance of this study is not only in its modelling outcomes but in demonstrating the applicability and advantages of this statistical approach to air pollution studies.

UFP and BC at a mid-sized city in Po valley, Italy: Size-resolved partitioning between primary and newly formed particles

In Po valley (Northern Italy) concentrations of particulate matter (PM) often exceed air quality standards, and road traffic is reported as one of the main sources of pollution. This study investigates the size resolved particle number concentration and size distribution at one rural station, one urban station and one traffic station. The measured size-resolved particle number concentration has been reduced by means of cluster analysis to four particle size fractions (cluster 1: 7–29 nm, cluster 2: 29–95 nm, cluster 3: 95–264 nm, and cluster 4: 264–10,000 nm) based on their behavior in atmosphere according to common time patterns. The primary emissions from traffic are evaluated based on black carbon (BC) and size-resolved particle number concentration data, considering separately single size intervals and providing cluster-resolved information on primary and newly formed particle concentration. Particles directly emitted by vehicle exhaust exhibit similar numbers for the clusters 1 to 3 while newly formed particles mainly occurs in cluster 1. Furthermore, diurnal variation of directly emitted particles is found to closely follow the BC levels, while the trend of newly formed particles varies according to air temperature, solar radiation and particle pollution levels. The results release that in Po valley the variations of particle number levels do not always reflect the variation of road traffic emissions in urban areas as the large availability of anthropogenic precursors can favor summertime nucleation events with region-wide extension.

Characterisation of particle mass and number concentration on the east coast of the Malaysian Peninsula during the northeast monsoon

Particle mass concentrations (PM10, PM2.5 and PM1) and particle number concentration ((PNC); 0.27 μm ≤ Dp ≤ 34.00 μm) were measured in the tropical coastal environment of Bachok, Kelantan on the Malaysian Peninsula bordering the southern edge of the South China Sea. Statistical methods were applied on a three-month hourly data set (9th January to 24th March 2014) to study the influence of north-easterly winds on the patterns of particle mass and PNC size distributions. The 24-h concentrations of particle mass obtained in this study were below the standard values detailed by the Recommended Malaysian Air Quality Guideline (RMAQG), United States Environmental Protection Agency (US EPA) and European Union (EU) except for PM2.5, which recorded a 24-h average of 30 ± 18 μg m−3 and exceeded the World Health Organisation (WHO) threshold value (25 μg m−3). Principal component analysis (PCA) revealed that PNC with smaller diameter sizes (0.27–4.50 μm) showed a stronger influence, accounting for 57.6% of the variability in PNC data set. Concentrations of both particle mass and PNC increased steadily in the morning with a distinct peak observed at around 8.00 h, related to a combination of dispersion of accumulated particles overnight and local traffic. In addition to local anthropogenic, agricultural burning and forest fire activities, long-range transport also affects the study area. Hotspot and backward wind trajectory observations illustrated that the biomass burning episode (around February–March) significantly influenced PNC. Meteorological parameters influenced smaller size particles (i.e. PM1 and Dp (0.27–0.43 μm)) the most.

The application of wavelet decomposition to quantify the local and regional sources of ultrafine particles in cities

This study explores the application of wavelet decomposition as a means to distinguish between local and regional sources of ultrafine particles (UFP). Particle number concentrations were measured at a central site, two downtown sites, and four residential sites located across Toronto, Canada. Using a wavelet decomposition algorithm, particle concentration time series were separated into two signals: high frequency local-to-neighbourhood scale sources and low frequency urban-to-regional scale sources and processes. At the field sites, local–neighbourhood sources contributed between 13 and 32% of the total particle concentration. The urban–regional signal at each field site exhibited stronger correlation and greater homogeneity with respect to the central site than the original concentration time series. In contrast, the high frequency local–neighbourhood source signals exhibited limited correlation and high heterogeneity with respect to the central site. Traffic volume within a 2.5 km buffer explained 87% of the variability in the local–neighbourhood level signal observed between field sites while no significant association with traffic was found for the original particle number concentration data. This study has demonstrated that wavelet decomposition can be a useful tool for estimating exposure to UFP from local–neighbourhood and urban–regional scale sources and processes.

Exploratory study on a statistical method to analyse time resolved data obtained during nanomaterial exposure measurements

Most of the measurement strategies that are suggested at the international level to assess workplace exposure to nanomaterials rely on devices measuring, in real time, airborne particles concentrations (according different metrics). Since none of the instruments to measure aerosols can distinguish a particle of interest to the background aerosol, the statistical analysis of time resolved data requires special attention. So far, very few approaches have been used for statistical analysis in the literature. This ranges from simple qualitative analysis of graphs to the implementation of more complex statistical models. To date, there is still no consensus on a particular approach and the current period is always looking for an appropriate and robust method. In this context, this exploratory study investigates a statistical method to analyse time resolved data based on a Bayesian probabilistic approach. To investigate and illustrate the use of the this statistical method, particle number concentration data from a workplace study that investigated the potential for exposure via inhalation from cleanout operations by sandpapering of a reactor producing nanocomposite thin films have been used. In this workplace study, the background issue has been addressed through the near-field and far-field approaches and several size integrated and time resolved devices have been used. The analysis of the results presented here focuses only on data obtained with two handheld condensation particle counters. While one was measuring at the source of the released particles, the other one was measuring in parallel far-field. The Bayesian probabilistic approach allows a probabilistic modelling of data series, and the observed task is modelled in the form of probability distributions. The probability distributions issuing from time resolved data obtained at the source can be compared with the probability distributions issuing from the time resolved data obtained far-field, leading in a quantitative estimation of the airborne particles released at the source when the task is performed. Beyond obtained results, this exploratory study indicates that the analysis of the results requires specific experience in statistics.

Physicochemical and toxicological characteristics of welding fume derived particles generated from real time welding processes

Welding fume particles have been well studied in the past; however, most studies have examined welding fumes generated from machine models rather than actual exposures. Furthermore, the link between physicochemical and toxicological properties of welding fume particles has not been well understood. This study aims to investigate the physicochemical properties of particles derived during real time welding processes generated during actual welding processes and to assess the particle size specific toxicological properties. A compact cascade impactor (Harvard CCI) was stationed within the welding booth to sample particles by size. Size fractionated particles were extracted and used for both off-line physicochemical analysis and in vitro cellular toxicological characterization. Each size fraction was analyzed for ions, elemental compositions, and mass concentration. Furthermore, real time optical particle monitors (DustTrak™, TSI Inc., Shoreview, Minn.) were used in the same welding booth to collect real time PM2.5 particle number concentration data. The sampled particles were extracted from the polyurethane foam (PUF) impaction substrates using a previously developed and validated protocol, and used in a cellular assay to assess oxidative stress. By mass, welding aerosols were found to be in coarse (PM 2.5–10), and fine (PM 0.1–2.5) size ranges. Most of the water soluble (WS) metals presented higher concentrations in the coarse size range with some exceptions such as sodium, which presented elevated concentration in the PM 0.1 size range. In vitro data showed size specific dependency, with the fine and ultrafine size ranges having the highest reactive oxygen species (ROS) activity. Additionally, this study suggests a possible correlation between welders' experience, the welding procedure and equipment used and particles generated from welding fumes. Mass concentrations and total metal and water soluble metal concentrations of welding fume particles may be greatly influenced by these factors. Furthermore, the results also confirmed the hypothesis that smaller particles generate more ROS activity and should be evaluated carefully for risk assessment.

An inexpensive particle monitor for smoker behaviour modification in homes

ObjectiveTo compare the response of a new particle counting instrument (Dylos DC1700) with that produced by a device (TSI Sidepak Personal Aerosol Monitor AM510) commonly used to measure PM2.5 in...

Pedestrian Exposure to Size-Resolved Particles in Milan

ABSTRACT Measurement campaigns for airborne particles along a pedestrian route in the city center of Milan were performed by means of a portable instrument consisting of an optical particle counter and a global positioning system (GPS) signal receiver. Based on the size-resolved particle number concentration data and on proper density factors experimentally determined for Milan urban area, the mass concentrations were calculated in terms of particulate matter with aerodynamic diameters ≤10 μm (PM10), ≤2.5 μm (PM2.5), and ≤1 μm (PM1). Besides directly measuring the personal exposure to PM throughout the route, the measurement campaigns pointed out small spatial and temporal variations of the concentration ranges in the different urban microenvironments visited along the route as well as very peculiar features of the particles levels in the underground subway. These findings suggested that the personal exposure of pedestrians in the city center could be estimated by simply taking into account the exposure at the open air and in the subway. The comparison between measured and calculated exposures according to the microenvironment-based estimation results in reasonable accordance, even though the estimations tend to slightly underestimate (12%) the actual measured exposure. IMPLICATIONS This work demonstrates that the actual exposure of pedestrians in Milan city center can be estimated by exposure modeling based on ambient data from an urban background site, provided that high-time-resolution data are available. This suggests that instruments such as the optical particle counters (OPCs) should be also included in the air quality monitoring network, together with conventional monitors used for air quality standard verification, in order to allow personal exposure assessment at open-air urban sites. However, because OPCs provide indirect information on the particle mass concentration, local field tests are required for establishing the proper algorithm to obtain mass concentrations from particle counting.

PMF Analysis of Wide-Range Particle Size Spectra Collected on a Major Highway

Particle number concentration data have been collected on a very busy road in central London (Marylebone Road). Continuous size distributions from 15 nm to 10 μm diameter, collected over 21 days, were analyzed using positive matrix factorization which identified 10 factors, five of which were observed to make major contributions (greater than 8%) to either the total number or volume of particulate matter. The sources associated with each factor were identified from the size distribution, directional association, diurnal variation and their relationship to meteorological pollution and traffic volume variables. The factors related to the emissions on Marylebone Road accounted for 40.5% of particle volume and 71.9% of particle number. These comprised nucleation mode exhaust particles (3.6% of total volume and 27.4% of total number), solid mode exhaust particles (18.8% of total volume and 38.0% of total number), brake dust (13.7% of total volume and 1.7% of total number and resuspension (4.4% of total volume and 4.8% of total number). Another six factors were associated with the urban background accounting for 59.5% of total volume and 28.2% of total particle number count. The method is extremely successful at separating the components of on-road emissions including brake wear and resuspension.

Influence of high-voltage transmission lines on pollutant concentration in nonurbanized areas

The variations of ozone and aerosol particles as possible pollutants in the vicinity of high-voltage transmission lines are presented and discussed. The ozone, aerosol particle (D > 0.4µm) number concentrations and meteorological parameters near two high-voltage 330 kV transmission lines and at the background site at the 220 m distance from the lines were investigated. During the investigation period the average ozone concentration close to high-voltage lines was higher than the background concentration and the instantaneous ozone concentration level was higher by 38% in separate cases, especially during the nighttime. The influence of meteorological parameters on the ozone concentration variation is discussed and it is determined that the relative humidity, wind speed and direction were the most important parameters causing the ozone concentration increase in the surroundings of transmission lines in nonurbanized areas. The analysis of aerosol particle number concentration data showed that the high-voltage lines were not the source of aerosol particles in the area of the experiment. Keywords: high-voltage transmission lines, ozone, aerosol particles, concentration, meteorological parameters

Multisite Study of Particle Number Concentrations in Urban Air

Particle number concentration data are reported from a total of eight urban site locations in the United Kingdom. Of these, six are central urban background sites, while one is an urban street canyon (Marylebone Road) and another is influenced by both a motorway and a steelworks (Port Talbot). The concentrations are generally of a similar order to those reported in the literature, although higher than those in some of the other studies. Highest concentrations are at the Marylebone Road site and lowest are at the Port Talbot site. The central urban background locations lie somewhere between with concentrations typically around 20 000 cm(-3). A seasonal pattern affects all sites, with highest concentrations in the winter months and lowest concentrations in the summer. Data from all sites show a diurnal variation with a morning rush hour peak typical of an anthropogenic pollutant. When the dilution effects of windspeed are accounted for, the data show little directionality at the central urban background sites indicating the influence of sources from all directions as might be expected if the major source were road traffic. At the London Marylebone Road site there is high directionality driven by the air circulation in the street canyon, and at the Port Talbot site different diurnal patterns are seen for particle number count and PM10 influenced by emissions from road traffic (particle number count) and the steelworks (PM10) and local meteorological factors. Hourly particle number concentrations are generally only weakly correlated to NO(x) and PM10, with the former showing a slightly closer relationship. Correlations between daily average particle number count and PM10 were also weak. Episodes of high PM10 concentration in summer typically show low particle number concentrations consistent with transport of accumulation mode secondary aerosol, while winter episodes are frequently associated with high PM10 and particle number count arising from poor dispersion of local primary emissions.

Estimating time series of aerosol particle number concentrations in the five HEAPSS cities on the basis of measured air pollution and meteorological variables

Background: Ultrafine particles in urban air represent a potentially important health risk, and are not well characterized by mass concentrations like PM10 or PM2.5. The aerosol particle number concentration (PNC) is dominated by ultrafine particles, but urban PNC measurement campaigns have only recently started in many cities and missing data impedes much research. Hence, reliable estimation techniques are needed. Past estimations of ambient concentrations of particulate matter have focused on mass concentrations. This project developed city-specific models for estimating PNC using available data on other air pollutants and meteorological variables during a period when PNC was measured, and applied them retrospectively to predict daily PNC levels during the Health Effects of Air Pollution on Susceptible Subpopulations (HEAPSS) study period in order to enable epidemiological analyses. Methods: Monitoring of PNC began in April 2001 using condensation particle counters (3022A, TSI) in Augsburg, Barcelona, Helsinki, Rome, and Stockholm. Concurrent measurements of air pollutants and weather were used, as well as selected interactions between the two, to fit a regularized linear model (also called ridge regression). This technique is robust with respect to inclusion of irrelevant explanatory variables and can be modified to be highly tolerant of missing data, two highly beneficial features when there are many explanatory variables. Results: The most important predictor variables were the nitrogen oxides. The models appear to fit PNC data relatively well, with R2 of 0.77, 0.80, 0.58, 0.84, 0.81 respectively for the five cities. Split-halves analysis (modelling on half of the data with validation on the other half) indicates that the modelling process was fairly reliable. Conclusion: A statistical model can be applied to existing data on traffic-related air pollutants and weather variables in order to predict PNC levels. The retrospective prediction of PNC levels appears to be sufficiently reliable for use in epidemiological research.

Quality control of semi-continuous mobility size-fractionated particle number concentration data

Fine and ultrafine particles have been postulated to play an important role in the association between ambient particulate matters and adverse health effects. As part of the EPA Supersite Program, the Southern California Particle Center & Supersite has conducted a series of monitoring campaigns that contribute to a better understanding of the sources, chemical composition and physical state of ambient aerosols. The Scanning Mobility Particle Sizer (SMPS) was deployed to semi-continuously measure mobility size-fractioned particle number concentrations. As part of the quality control efforts, we developed a two-stage graphic and statistical procedure to label and identify potentially discordant observations. The first stage considered the entire size-fractionated data by date-time as a whole to plot total concentration (TC) vs. coefficient of variation (CV), both in log scale. TC represents the magnitude of overall concentration for a size distribution; while CV represents the relative variability. This plot was used to partition all size distributions into four to five distinct regions. In each region, a generalized extreme studentized deviate (ESD) and a modified Z-score procedure were applied to identify potential discordant outliers. We have found that the majority of particle size distributions are concentrated within a ‘normal’ region, with TC ranging from 102 to 105 cm−3 and CV varying between 20% and 200%. Size distributions that are contaminated with discordant outliers are displayed distinctly from the ‘normal’ region and form four to five clusters in the Log TC–Log CV plot. The pattern of clusters in the plot is consistent among the four sampling sites in this study, suggesting the robustness of this technique. The generalized ESD and modified Z-score effectively identify discordant outliers and reveal that the pattern of clustering outliers are consistent within each distinct region. It has, thus, been concluded that the new approach is a useful quality control tool to identify potential discordant outliers in SMPS data.

Quality control of semi-continuous mobility size-fractionated particle number concentration data

Abstract Fine and ultrafine particles have been postulated to play an important role in the association between ambient particulate matters and adverse health effects. As part of the EPA Supersite Program, the Southern California Particle Center & Supersite has conducted a series of monitoring campaigns that contribute to a better understanding of the sources, chemical composition and physical state of ambient aerosols. The Scanning Mobility Particle Sizer (SMPS) was deployed to semi-continuously measure mobility size-fractioned particle number concentrations. As part of the quality control efforts, we developed a two-stage graphic and statistical procedure to label and identify potentially discordant observations. The first stage considered the entire size-fractionated data by date-time as a whole to plot total concentration (TC) vs. coefficient of variation (CV), both in log scale. TC represents the magnitude of overall concentration for a size distribution; while CV represents the relative variability. This plot was used to partition all size distributions into four to five distinct regions. In each region, a generalized extreme studentized deviate (ESD) and a modified Z -score procedure were applied to identify potential discordant outliers. We have found that the majority of particle size distributions are concentrated within a ‘normal’ region, with TC ranging from 10 2 to 10 5 cm −3 and CV varying between 20% and 200%. Size distributions that are contaminated with discordant outliers are displayed distinctly from the ‘normal’ region and form four to five clusters in the Log TC–Log CV plot. The pattern of clusters in the plot is consistent among the four sampling sites in this study, suggesting the robustness of this technique. The generalized ESD and modified Z -score effectively identify discordant outliers and reveal that the pattern of clustering outliers are consistent within each distinct region. It has, thus, been concluded that the new approach is a useful quality control tool to identify potential discordant outliers in SMPS data.