Seasonal Variations of Air Quality Measurements of Aba Metropolis and Suburbs Using MATLAB and ANN

Abstract

Air pollution is a major life-threatening problem in industrialized and commercially vibrant cities like Aba metropolis and its suburbs in Abia State Nigeria. The study of selected air pollutants in these areas were performed using Matrix Laboratory (MATLAB) and Artificial Neural Networks (ANN) pollution models. Primary data was collected by conducting sampling analysis on air samples during dry and rainy seasons from 2024 and 2025. MATLAB and ANN pollution models were generated by integrating measurements and spatial databases using polynomial expressions. The MATLAB 7th degree linear regression polynomial described the relationship between dependent and independent variables for the pollutants. The correlation methods verified that most MATLAB models could accurately predict or forecast concentration levels. Also the Artificial Neural Network demonstrated tracking of the actual plots on MATLAB. The analysis of variance (ANOVA) was also deployed which showed p < 0.05 for carbon monoxide (CO), nitrogen dioxide (NO2), sulphur dioxide (SO2), and total particulate matter (TPM), indicating that, there was a significant impact by the seasons on the concentrations of all gaseous pollutants under study(i.e. seasonal variations of concentration was highly affected by the two seasons). ANN was able to track all gaseous pollutants represented by MATLAB successfully above 50%.