Temporal-causal modeling of air pollution in the city of Plovdiv, Bulgaria: a case study

Abstract

High levels of air pollution are dangerous to human health, which is a current problem for densely populated cities worldwide. Studying this problem can help detect pollutants’ time dependencies on basic meteorological measurements and other factors for future prediction and elaborate corresponding alarms when official upper pollution limits are exceeded. In this work, time-causal models based on previous daily time observations and meteorological measurements in the city of Plovdiv, Bulgaria, are applied. Vector-type temporal-causal models are constructed and analyzed for carbon dioxide (CO2), nitrogen dioxide (NO2), sulfur dioxide (SO2), and fine dust particles below size 10, 2.5, and 1 micron (PM10, PM2.5, and PM1), respectively. Pollution levels are predicted seven days ahead.