Análisis de series de tiempo de contaminantes atmosféricos en la ciudad de Bogotá a partir del desarrollo de modelos estadísticos arima

Abstract

Nowadays, monitoring and forecasting the air quality status is a task of high relevance for environmental authorities, due to the negative effects that air pollution have on life’s population quality. Taking into account the need to move forward in the development of alternative tools for the adequate management of the atmospheric resource, the present investigation proposes a forecast simulation of the air quality status the development of ARIMA models having as input the reports of ten stations of Bogotá's air quality monitoring system. The results showed that ARIMA (Autoregressive Integrated Moving Average) models of particulate matter and ozone were better adjusted in the monthly time scale and at the same time, the temporal relationship between stations and pollutants was evidenced by revealing a similar behavior by zones of the city from the structure of the models. The persistence and variability of the pollutants were explained using the meteorological conditions, which finally played an important role in the forecast alerts for exceedances and atypical values of concentration in the air. It was found that the variable that had the greatest influence on the pollution episodes was the direction of the wind coming from Bogota’s west side.