Andreas Gavros

The effect of COVID-19 lockdown (LD) on many ambient air pollutants (NO, NO₂, PM_2.5, PM₁₀, O₃ and SO₂) was assessed for the first time in the Western Balkans with an innovative approach that evaluates a variety of factors including the stringency of the LD measures, the type of location, the pollution sources, the correlation with traffic fluxes and the meteorology. To that end, observations from 10 urban sites were compared with historical time series. The time window 1 February–30 May 2020 was classified in sub-periods on the basis of the stringency of the circulation restrictions. NO₂ and O₃ are the pollutants most affected by restrictions to population circulation due to lockdown during the first phase of the COVID-19 pandemic, and are well correlated with traffic fluxes. A reduction in fine particulate matter (PM_2.5 and PM₁₀) concentrations is observed in all sites only during the full LD periods, while the relation between SO₂ average and maximum hourly concentrations and LD periods in industrial and traffic sites vary from site to site. The reduction in NO₂ concentrations during the LD resulted in a reduction in mortality associated with air pollution in the largest cities, while the interpretation of the changes in O₃ and particulate matter is less clear.

By deploying machine learning (ML) and deep learning (DL) algorithms, we address the problem of smell event modelling in the Pittsburgh metropolitan area. We use the Smell Pittsburgh dataset to develop a model that can reflect the relation between bad smell events and industrial pollutants in a specific urban territory. The initial dataset resulted from crowd-sourcing citizen reports using a mobile phone application, which we categorised in a binary matter (existence or absence of smell events). We investigate the mapping of smell data with air pollution levels that were recorded by a reference station located in the southeastern area of the city. The initial dataset is processed and evaluated to produce an updated dataset, which is used as an input to assess various ML and DL models for modelling smell events. The models utilise a set of air quality and climate data to associate them with a smell event to investigate to what extent these data correlate with unpleasant odours in the Pittsburgh metropolitan area. The model results are satisfactory, reaching an accuracy of 69.6, with ML models mostly outperforming DL models. This work also demonstrates the feasibility of combining environmental modelling with crowd-sourced information, which may be adopted in other cities when relevant data are available.