Abstract
Airborne nanoparticles have prompted a strong research interest in the scientific community due to their adverse effects on human health and the environment. However, there is a notable lack of studies focusing on extreme summertime conditions, where ambient temperatures can reach ~48 °C, relative humidity falls to its minimum values, and dust events are frequently encountered. The overall aims of this research are to understand the behaviour and sources of airborne nanoparticles in hot and arid environmental conditions, develop a statistical prediction model for nanoparticles that uses routinely-monitored air pollutants, and investigate the mitigation measures (i.e., vegetation barriers) used to limit the penetration of on-road nanoparticles to the surrounding vicinity. Size-resolved measurements of particle number distribution (PNDs) and concentrations (PNCs) were carried out continuously for one month at a roadside location in the State of Kuwait using a fast-response differential mobility spectrometer (DMS500) to assess the influence of summertime meteorological conditions on nanoparticles. Further data of trace pollutants (NOx, O3, CO, SO2 and PM10) and meteorological variables (wind speed, wind direction, temperature, relative humidity, and solar radiation), were obtained from the Kuwait Environment Public Authority (KEPA). The collected data was analysed to assess the behaviour of nanoparticles during summertime and to understand any unusual behaviour of PNDs and PNCs during (i) the afternoon, when temperature reaches it maximum and relative humidity to its minimum, and (ii) during the occurrence of Arabian dust events. The collected PNDs data were used to apportion the major sources and their contribution to total PNCs using a positive matrix factorisation (PMF) model. Further, a preliminary attempt to predict nanoparticles in three size ranges (nucleation mode: 5–30 nm, Aitken mode: 30–100 nm, and accumulation mode: 100–300 nm) using artificial neural network (ANN), was made. For the prediction purpose, seven scenarios were considered using different combinations of the routinely-measured meteorological and trace pollutant data as covariates. In addition, intermittent monitoring of PNDs and the associated PNCs were performed using DMS50 at a kerbside location in the United Kingdom (UK) to investigate the effect of vegetation barriers on traffic-generated nanoparticles, as well as pedestrian exposure. PND data was collected at four sampling locations pseudo-simultaneously using a multi-probe switching system. These locations encompassed the vegetation barrier and allowed us to make novel comparisons. Despite high traffic volumes during noon hours, there was a substantial decrease in PNCs with a corresponding increase in geometric mean diameters (GMDs) due to high ambient temperature (∼48 °C) and wind speed (∼15 m s–1). The high wind speed has a dispersive effect (i.e., dilution), and saltation causes the suspension of particles and enhances the coagulation process. Based on the PMF modelling, traffic emissions were found to be a major contributor (73%) to the total apportioned PNCs, whereas Arabian dust transport was found to be the lowest contributor (3%). ANN succeeded in capturing the general trend between observed and predicted PNCs with R2 up to 0.79. The deviations between the observed and predicted PNCs were not substantial, as evidenced by the fact that predicted PNCs were within a factor of two of the observed PNCs. Vegetation barriers were found to reduce not only PNCs by ~37%, but also the associated particle respiratory deposited doses in the human respiratory tract (RDD) by ~36%. The implication of vegetation barrier results are of high importance in the reduction of PNCs and the associated RDD. Besides policy makers and environmental authorities, the findings of this work are important for the modelling community to treat major nanoparticle sources in dispersion modelling and health impact assessments in the region.