Air quality data analysis obtained by means of new generation sensor network

Air pollution measurement networks are meant to monitor the concentration of the primary pollutants. This type of infrastructures can be installed in urban areas in order to have an overall view of the environmental conditions. This thesis work aims to increase the quality of environmental data provided by air quality monitoring networks. In particular, the object of this thesis work is the treatment of the large number of data produced by the air quality monitoring networks implemented by Sense Square along the Italian territory both in urban and extra-urban context. The first issue faced was the definition of the correct positions for the sampling points of the network that is relevant to the network design. A modelistic approach was applied to define the most interesting points of the city where to install the measuring stations. The Operational Street Pollugtion M odel was used to simulate the PM10 concentrations in five chosen points. For each point it was possible to define the local conditions such as traffic intensity, street geometry, hourly number and type of vehicles passing through the streets and the meteorological parameters. The model adopted takes into account the pollutant dispersion using a Gaussian plume, while a box model is used to calculate the contribution from the recirculation vortex along the urban canyon. The state of art uses the dispersion models to define the pollutant behaviour dispersed in the atmosphere but doesn’t compare the modelled results with the real measured ones. In this case, the comparison of the data obtained by the OSPM model and the ones measured after the installation of the sampling stations showed a good agreement, suggesting that the points chosen were the most interesting from the air pollution point of view. A similar approach was used, to forecast the urban quality air using the concentration levels of particulate matter and meteorological conditions. To integrate the fixed monitoring network to have a higher space resolution, the real-time on-road monitoring stations were implemented. The real-time on-road mobile monitoring network provides for the concentrations of pollutant in an innovative configuration. In fact, the data coming from vans equipped with the measuring devices moving along the streets of Milan makes it possible to aggregate the data or by districts or by cell. For the micro-analysis, the whole investigated area was divided into square cells of 1km side to form a grid. For each cell, the large amount of data available allows the analysis of the hourly, weekly and, monthly average. This new type of monitoring has to be compared with the traditional one represented by the Milan municipality air quality network. From the comparison between the two sets of data it was possible to notice the good agreement between them, suggesting that even if the measurement techniques are different, the results are reliable. Obviously, the availability of data from the dynamic network is strictly connected with the vans operating hours from 09:00 to 17:00. Moreover, the urban context is full of traffic restrictions like traffic limited zones, pedestrian pathways and parks that doesn’t allows the vans passage. So, it is possible that some of the cells in the grid which the city is divided into have no data available. This missing values can be estimated by using a geospatial interpolation model. Another possibility to estimate air quality in the missing cells, is to use measured data in the neighbouring cells with purposely designed algorithms. In the present study, the results of the applications of few simple methods is assessed by comparing, in each cell where available data are present, the values predicted by the algorithm using the neighbouring data end the experimental value. Different criteria will be adopted and compared. In some of them the influence of the neighbouring cells on the prediction of the local concentration is based exclusively on the distance between the centre of the influencing cell and the centre of the tested cell, in others the direction and the speed of the wind are taken into consideration in the calculation procedure. The approach proposed involves various innovative aspects, in fact this study use air quality data provided by an extremely innovative technology really and currently measured in an Italian city. Moreover, the model is applied to a very large area. Finally, the developed method for the spatial estimation of missing data was compared with other tecniques to evaluate the best solution. During the first part of the work a deterministic approach was used to to investigate the behaviour of unexpected events like fires. This methodology was studied in deep to evaluate the dispersion of the pollutant generated after a fire occurred in a factory. The fire was described as an equivalent stack having the height of the observed cloud of smoke generated by the fire. The pollutant propagation was simulated with a Gaussian plume dispersion model. From February 2020, the progressive adoption of measures to contain Coronavirus's contagion has resulted in Italy, especially in Lombardy, a sudden change in anthropic activities. From a scientific point of view, the new situation represents a unique laboratory for understanding and predicting the consequences of specific measures aimed at improving air quality. In this part of the work, the effect of the lockdown on air quality in the city of Milan (Italy) was analyzed. The PM10 values measured by the ARPA Lombardia air quality monitoring network indicate the seasonality of these pollutants, which typically record the highest values in the coldest months of the year. March 2020 data analysis shows an alternation of days with higher and lower particulate matter concentrations values. Some episodes highlighted the complexity of the phenomena related to the formation, transport, and accumulation of atmospheric particulates. Others highlighted the contribution of the second component and the meteorological situation most favorable to accumulation. The study showed that the trend of a general reduction of pollutant concentrations observed must be attributed to the decrease in emissions, in particular from the transport sector, from the variation of meteorological and environmental conditions. Among the various phenomena that participate in air pollution, there is the influence of Saharan dust. These powders can participate in the increase of PM10 concentrations even though they are of a different nature compared to the particulate produced in urban areas. For this reason, a procedure has been implemented to take into account the contribution of these dusts to the extent of pollution. [edited by Author]