Aplicação e performance da análise de intervenção em séries de emissão de gases poluentes na Cidade do México

Abstract

This research aims to evaluate the performance of time-series models together with the use of analysis, when the time for action is determined by visual inspection of the series or the spectral analysis, greenhouse gas emission data. In carrying out this research were considered data from the time series of the measurement of ozone gas (O3), nitrogen dioxide (NO2), carbon monoxide (CO) and sulfur dioxide (SO2), understood in January 1986 to December 2014, referring to the first day of each month, totaling 348 observations. The models were performed pre-analysis, using spectral analysis to determine the period in which the intervention must be done, show better performance considering the AIC and BIC criteria. Indicating that whenever possible to use this tool as an aid for modeling time series who need intervention.