Spatial analysis of PM_2.5 using a Concentration Similarity Index applied to air quality sensor networks

Abstract. Air quality sensor (AQS) networks are useful for mapping PM_2.5 in urban environments, but quantitative assessment of the observed spatial and temporal variation is currently under-developed. This study introduces a new metric – the Concentration Similarity Index (CSI) – to facilitate a quantitative and time-averaged comparison of the concentration‑time profiles of PM_2.5 measured by each sensor within an air quality sensor network. Following development on a dataset with minimal unexplained variation and robust tests, the CSI function is ensured to represent an unbiased and fair depiction of the air quality variation within an area covered by a monitoring network. The measurement data is used to derive a CSI value for every combination of sensor pairs in the network, which can then be compared with others in the network, yielding valuable information on spatial variation in PM_2.5. This new method is applied to two separate AQS networks in Dungarvan and Cork City, Ireland. Dungarvan yielded a lower mean CSI, indicating lower overall similarity between locations in the network, possibly due to the town’s coastal location giving rise to higher variation within the network. In both networks, the average diurnal plots for each sensor exhibit an evening peak in PM_2.5 concentration due to emissions from residential solid fuel burning, however, there is considerable variation in the size of this peak. Clustering techniques applied to the CSI matrices identify two different location types in each network; locations in central or residential areas which experience more pollution from sold fuel burning and locations on the edge of the urban areas which experience cleaner air. Furthermore, the examination of isolated data periods (January and May) indicates higher PM_2.5 levels during periods of increased residential solid fuel burning act as a major driver for greater differences (lower similarity indices) between locations in both networks. Additionally, the CSI method facilitates the assessment of the representativeness of the PM_2.5 measured at regulatory air quality monitoring locations with respect to population exposure, showing here that location type is more important than physical proximity in terms of similarity assessment. Applying the CSI in this manner can allow for the placement of monitoring infrastructure to be optimised. The findings of this work underscore the influence of solid fuel combustion as a local contributor to PM_2.5 and the variation it can cause between the measurements at different monitoring locations in a network while also highlighting the importance of including wintertime PM data for accurate comparisons. The CSI method developed here could be a valuable tool for quantitative comparisons of air quality within a monitoring network, offering insights for further regulatory monitoring and exposure assessments.