TIME SERIES ANALYSIS AND PREDICTION OF AIR POLLUTANT CONCENTRATIONS (PM10, SO2 AND NO2) USING LINEAR REGRESSION APPROACH IN KAZANLAK, BULGARIA


Authors: Svetla Stoykova, Miroslava Ivanova, Diyana Dermendzhieva, Lilko Dospatliev
Keywords: air pollution, linear regression model, statistical software R, Bulgaria

Abstract:
The aim of the present study is to conduct time series analysis and to predict of air pollutant concentrations applying linear regression approach using PM10, SO2 and NO2 data from January, 01, 2023 to April 30, 2024 in the second largest city in the region Stara Zagora, Bulgaria - Kazanlak. All statistical computing, test and graphics were performed with the statistical software R. We received three linear regression models, which showed that if SO2 increase by 1%, the effect of this increase would result in an increase in NO2 by 0.90%; if PM10 increase by 1%, the effect of this increase would result in an increase in SO2 and NO2 by 0.19% and 0.23%, respectively.Download article