Comparative study and prediction of ambient air quality of Durgapur Industrial Belt, West Bengal using time series forecast model

Abstract

Time series predictive forecast models can be used to monitor air levels and study, trends in ambient air quality. In this study, we have collected the data of Oxides of Sulphur (SOx), Suspended Particulate Matter (SPM10), and Oxides of Nitrogen (NOx) in Bidhannagar, Durgapur Industrial Belt, West Bengal from the West Bengal Pollution Control Board (2007-2019). The data was analysed both season-wise and month-wise to nullify the irregularity of the data obtained.Time series model can be established only using regular data. The study mainly focused on the comparision between month-wise and season-wise models, in order to identifya prediction model that provides best predictions. The forecast was performed using the seasonal autoregressive integrated moving average (SARIMA) model. The results confirmed that among all forecast models of SARIMA (P, D, Q) (p, d, q) the season-wise model provide best predictions. Graphs indicated a higher concentration of SPM10 and NOx in the winter. The season-wise model for SPM10 (1,0,0)(0,1,1) showed significant higher trends but similar season-wise forecast model for NOx (1,0,0)(0,1,1) showed a decreasing trend. The levels of SPM10 is predicted to show an increasing trend in next four years whereas NOx level is predicted to remain low for the next four years. The seasonal prediction models can be used in understanding the trend in ambient air quality in Durgapur Industrial Belt and facilitate in taking necessary steps to combat the prevailing environmental conditions.

Keywords

• air pollution
• oxides of nitrogen
• oxides of sulphur
• suspended particulate matter
• sarima
• forecast