Ambient or ground-level ozone (O3), a key player in atmospheric chemistry, has been studied extensively throughout the world in recent decades due to its harmful effects on human health and the environment. The aim of this study is to examine in detail the time patterns of O3 in the Czech Republic, a Central European country with a long-term history of ambient air pollution. We have examined the interplay between annual trends and within-a-year seasonality in daily O3 concentrations at twelve stations representing different environments, geographical regions and altitudes in the Czech Republic in 1993-2018. We base our analysis on a generalised additive model (GAM) framework as a modern regression approach suitable for addressing non-linear trend shapes in a formalised and unified way. In particular, we use penalised splines. The inconsistent behaviour of the stations prior to 1998, two local maxima around 2003 and 2006 and a steady increase in O3 at all sites from 2014 to the present are the main features of our study. A seasonality O3 pattern with a clear maximum in May-June is similar for all stations. In parallel with the O3 concentration changes over the years, the deformation of seasonality profiles over a long period of time occurs, resulting in a shift of the peak O3 concentrations towards later days in the year. Our statistical modelling offers a detailed view of both long-term trend and seasonality components separately. Additionally, in a model with trend*seasonality interaction, we are able to study the gradual deformation of the seasonality profile over the years. Capsule: Our analysis of daily mean O3 concentrations at twelve Czech sites in 1993-2018 revealed a steady increase in O3 from 2014 and deformation of the seasonality O3 profiles over the year.
Keywords: 1993–2018; Generalised additive model; Ground-level ozone; Seasonality; Time trends.
Copyright © 2020 Elsevier B.V. All rights reserved.