Item

Abstract

Atmospheric pollution by ozone is a complex phenomenon which science and policy still struggle to quantitatively comprehend and control, respectively. Urban atmospheric ozone is a secondary pollutant which arises from a complex mechanism, initiated by sunlight and partially self propagating, whose yield in ozone depends on the relative abundance of nitrogen oxides and volatile organic compounds. The latter are a primary pollutant produced by human activities and the effects of the covid-19-related lockdowns on their levels have been extensively studied. On the other hand, the former are a broad class of chronically understudied primary and secondary molecules. In the present work, we aim to study the effect of the limited human activities dictated by the covid-19 lockdowns on the levels of a particular volatile organic compound, benzene. A gradient-boosted trees model was constructed to reproduce the benzene levels in Berlin for the years 2016-2019. The model was built using meteorological data from Berlin, together with proxies for seasonal cycles, mid-term trends and human activities, such as the year, the month and the day of the week. The application of the model to the period of the lockdown allows its effect to be quantified and linked to the reduction of human activities, giving policy makers valuable information for future air quality management policies.