Meteorological normalization of NO2 concentrations in the Province of Bolzano (Italian Alps). (1st February 2021)

• Record Type: Journal Article
• Authors: Falocchi, Marco; Zardi, Dino; Giovannini, Lorenzo
• Is Part Of: Atmospheric environment. Volume 246(2021)
• Publication Date: 2021-02-01
• Subjects: Random Forest -- Meteorological normalization -- Nitrogen dioxide -- Air quality -- Complex mountainous terrain -- Valley; Air -- Pollution -- Periodicals; Air -- Pollution -- Meteorological aspects -- Periodicals; 551.51
• DOI: 10.1016/j.atmosenv.2020.118048
• Languages: English
• ISSNs: 1352-2310

Abstract

Abstract: Pollutant concentrations in the atmosphere are controlled not only by emissions, but also by meteorological processes. As a consequence, adverse atmospheric conditions may hinder the effects of policies intended to improve air quality through reduction of emissions. In particular, low ventilation conditions and temperature inversions may significantly inhibit pollutant transport and mixing, determining high concentrations close to the ground. In order to disentangle the contribution of weather conditions on observed pollutant concentrations, meteorological normalization techniques can be applied. In this study, a normalization procedure based on a random forest predictive model is applied to 8-year–long series of nitrogen dioxide (NO2 ) concentrations measured at five air quality stations in the Province of Bolzano (Italian Alps). The normalization is performed on daily–averages of NO2 concentrations, related to a dataset composed of time variables and meteorological data from seven weather stations and one temperature profiler. The strong dependence of observed NO2 concentrations on atmospheric variables ( i.e. air temperature, atmospheric stability and wind speed) measured at the valley floor justifies the application of a normalization procedure. The resulting normalized time series of NO2 concentrations, instead, clearly display changes in correspondence of roadworks or other measures capable of modifying the emission regime, and allow to address the reliability of the applied procedure. Graphical abstract: Image 1 Highlights: Weather conditions may mask the actual emission regime. Meteorological normalization disentangles weather contributions on observed pollutants. Time series of NO2 concentrations are meteorologically normalized. The meteorological normalization bases on a random forest predictive model. Changes in the normalized series are explained by roadworks and interventions.