Análisis descriptivo para datos funcionales de la contaminación del aire en el centro histórico de Quito. Período 2019 – 2020.

Abstract

This research study aims to descriptively analyze the air pollution functional data in the historic center of Quito, for the years 2019 – 2020. For which a statistical software R, with the package "fda.usc" was used. The pollutants analyzed were: carbon monoxide, sulfur dioxide, nitrogen dioxide, ozone, and fine particulate matter. The analysis started by downloading the pollutant database from the Secretaría de Ambiente de Quito. Then a selection, ordering, and filling of missing data was carried out, by linear interpolation. The reading of the database was carried out with the function read.table and the transformation to functional data by the fdata function. In the general analysis, exploratory graphs were made, obtaining atypical curves, monthly mean curves with bands of confidence. In the analysis by periods, the mean curves for the class school periods were compared with school holidays, the pandemic period and no pandemic, and periods from Monday to Friday, and Saturday to Sunday. To finally present linear models with a scalar response. The results obtained from the study indicated that the behavior of the pollutants is irregular throughout the day. From the analysis by periods, it was possible to conclude that the functional mean is higher in the school classes period, not pandemic period, and Monday to Friday. Linear models with a scalar response of the variable: the daily mean of fine particulate matter, resulted in the variability of 31.61%, 35.08% and for carbon monoxide, sulfur dioxide, and nitrogen dioxide a 57.45% respectively. While the percentage of variability of the daily mean for the fine particulate matter explained by the variables of the three pollutants showed a percentage of 60.83%.