A1110
Title: Detecting anomalies in emission trends: A statistical approach for environmental policy in Portugal
Authors: Ana Borges - CIICESI, ESTG, Politécnico do Porto (Portugal) [presenting]
Clara Cordeiro - FCiencias.ID, Associacao para a Investigacao e Desenvolvimento de Ciencias (Portugal) (Portugal)
M Rosario Ramos - FCiencias.ID Associacao para a Investigacao e Desenvolvimento de Ciencias (Portugal)
Mariana Carvalho - CIICESI ESTG Politecnico do Porto (Portugal)
Abstract: To improve understanding of variations in emissions that serve as key indicators of environmental quality and the effectiveness of policies, a statistical methodology is proposed for detecting anomalies in the time series dataset of monthly emissions in Portugal. The focuse is on CO, NH3, NMVOC, NOx, OC, PM2.5, PM10, and SO2 emissions from 2000 to 2022. This information on changes is crucial so that policies may be designed for effective strategies toward emission reduction and improvement of air quality. The analytical method adapts a technique that has been developed for analysing hydrological-related time series data. First, the seasonal trend decomposition based on the Loess method is used to decompose the time series. Thereafter, a breakpoint analysis is undertaken to search for drastic changes in the emission trends for the seasonally adjusted data. Mann-Kendall test and Sen's slope estimator are used to discover the presence of significant increases or decreases in emissions. The application of this methodology to emissions data has been successful in the identification of breakpoints corresponding to significant changes in emission patterns. This is highly useful in environmental surveillance since it facilitates proactive measures to abate air pollution and its impact on health and the environment.
