Arunik Baruah

During personal exposure monitoring, short-term peaks, or “spikes”, in pollutant concentration, were observed frequently in proximity to the local pollution sources, especially road traffic. This phenomenon may influence the overall exposure. The current study examined spikes in particle number concentration (PNC) to differentiate local from background contributions by eliminating spikes from the personal exposure dataset.Personal exposures were measured on 33 walking trips alongside major and minor roads between 20th May and 26th June 2024, in a heavily populated residential area close to the University of Birmingham. A portable miniature particle counter (Testo DiSCmini) was carried in a backpack, measuring particle number concentration with a high temporal resolution of 1 second. Particle Number Count data (10-100nm) were also collected with a reference-grade instrument (Scanning Mobility Particle Sizer (SMPS)) at a local urban background air quality monitoring station.The time series of PNC contained short-term excursions (spikes) to higher concentrations. There were several steps involved in removing the spikes from the dataset including baseline calculation, spike identification and removal, and background data interpolation. The first step was done using a moving median with 10 minutes average before and after all data points. Secondly, a threshold i.e., 10% of the baseline was chosen which can capture spikes optimally based on visual observation. Data above the threshold was subsequently identified as spikes and excluded from the data set. Finally, the edited background data was interpolated using a linear method.The results show that roughly 25% (by time) of the walking data was categorized as short-term peaks. Removal resulted in a reduction of the overall average PNC by nearly 19%. Temporal variation according to weekday/weekend and period of the day revealed a decline in average PNC ranges of 12-34%, with the most significant fall of 34% occurring during weekday mornings (MWD), due to a substantial number of PNC spikes observed during this period. PNC measured during walking was scaled to SMPS-equivalent values using data from an instrument intercomparison and was compared to urban background SMPS data during measurement. It was shown that the average of corrected de-peaked PNC from walking (11801±8357 #/cm³) was 9% higher than that recorded in the urban background (10816±6711 #/cm³). Local diffuse sources are probably responsible for this higher concentration, while the spikes appear to be due to road traffic emissions and locally operating sources such as off-road mobile machinery.By separating short-term peak concentrations from personal exposure monitoring data, the data indicate that limiting pollutant hotspots, especially in areas with high population density, may reduce exposure to pollutants, particularly those with significant geographical and temporal variability such as ultrafine particles (UFP).

PM2.5 (Particulate matter < 2.5 mm in diameter) pollution is a significant environmental and public health concern in Europe. According to the European Environmental Agency (EEA, 2023), 97% of the urban population are exposed to concentrations over the World Health Organization's (WHO) 2021 annual limit of 5 µg m-3. Various models predict PM levels, such as Chemical Transport Models (CTMs) and statistical approaches based on meteorological variables. Machine Learning (ML) tools, particularly tree based alogorithms, outperform linear models due to the non-linear response of atmospheric species to environmental conditions and emissions.Our research aims to introduce a novel methodology for predicting PM2.5 levels at fine spatial and temporal scales using ML tools. The primary objective is to showcase the methodology's capability in estimating missing PM2.5 measurements in urban areas where direct observations are unavailable. To achieve this, we compiled a hybrid dataset using inputs from an intensive aerosol campaign conducted in the Selly Oak neighbourhood of Birmingham, UK, spanning from April 15th to June 20th, 2023. This campaign focused on a 1×1 km² block, housing approximately 10,000 university students. Four low-cost Optical Particle Counters (OPC-N3, Alphasense, UK) were strategically placed at fixed locations within the study area, measuring particle number size distribution (PNSD) in 24 bins from 0.35 – 40 µm, as well as PM1, PM2.5, and PM10 mass concentrations. An additional four OPC-N3 devices were employed for aerosol mapping through a mobile backpack-based arrangement. Data collection adopted a citizen science approach, collaborating with local businesses and schools for static sensors and engaging university students for the deployment of mobile sensors. All sensors underwent calibration by collocating with research-grade instruments at the Birmingham Air Quality Supersite (BAQS), both before and after the campaign.For a detailed analysis of PM2.5 distribution along different road segments, the network was divided into 30-meter segments, and the centroid was computed for each segment. Spatially resolved proxy variables of atmospheric emissions were assigned to each centroid, including population data, average traffic count by vehicle, road rank, and the average frequency distribution of vehicle speed. The hybrid dataset also integrated meteorological parameters from BAQS (wind speed, wind direction, atmospheric pressure, relative humidity, atmospheric temperature) and aerosol properties from reference instruments at BAQS.Three distinct calibration approaches were employed: 1) Standard Random Forest Regression (RF) with an 80-20 train-test split to predict PM2.5 levels based on input features (R2 = 0.85, MBE = -0.01 µg m-3). 2) Sensor Transferability Evaluation: Calibrating the RF on a specific OPC unit and evaluating its performance on an independent OPC (best performance R2 = 0.65, MBE = 0.43 µg m-3). This approach assesses the model's generalization across different sensors. 3) Road Transferability Evaluation: Calibrating the model on one road and evaluating its performance on a different new road (R2 = 0.71, MBE = -1.14 µg m-3). This approach explores the model's ability to generalize across different road types.This methodology holds significant potential for improving spatial resolution beyond regulatory monitoring infrastructure, refining air quality predictions, and enhancing exposure assessments critical for investigating health impacts.

Particulate Matter (PM) air pollution poses significant threats to public health. We introduce a novel machine learning methodology to predict PM_2.5 levels at 30 m long segments along the roads and at a temporal scale of 10 seconds. A hybrid dataset was curated from an intensive PM campaign in Selly Oak, Birmingham, UK, utilizing citizen scientists and low-cost instruments strategically placed in static and mobile settings. Spatially resolved proxy variables, meteorological parameters, and PM properties were integrated, enabling a fine-grained analysis of PM_2.5. Calibration involved three approaches: Standard Random Forest Regression, Sensor Transferability and Road Transferability Evaluations. This methodology significantly increased spatial resolution beyond what is possible with regulatory monitoring, thereby improving exposure assessments. The findings underscore the importance of machine learning approaches and citizen science in advancing our understanding of PM pollution, with a small number of participants significantly enhancing local air quality assessment for thousands of residents.

Low-cost electrochemical air quality sensors can provide deep insights into the intra-urban variability of different air pollutants after proper calibration using field co-location with regulatory Air Quality Monitoring stations.