Development of Mathematical Model for Prediction of PM2.5 Concentrations in Ambient air of Metal Recycling Industry in Ogijo, Ogun State, South Western Nigeria
Taiwo T. Oshin *
Department of Chemical Sciences, Lagos State University of Science and Technology, Ikorodu, Lagos State, Nigeria.
James M. Okuo
Environmental Analytical Research Laboratory, Department of Chemistry, University of Benin, Benin City, Edo State, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
Aims: This study aimed to develop a mathematical model for predicting PM2.5 pollutant concentrations in the ambient air of the metal recycling industry.
Study Design: This research is a quantitative design and utilized a regression and correlational analysis. Three models were developed for predicting PM2.5 concentrations: Linear Regression (LRM), Nonlinear Polynomial Regression (NPRM), and Nonlinear Gamma Regression (NGRM) models. Error evaluation functions were employed to analyze how these models deviated from the experimental data. The applicability of the models was assessed using statistical tools, such as correlation coefficient (r), coefficient of determination (R²), coefficient of non-determination (K²), student’s t (t-test), equality of variance (F-test), and chi-square (X2 ) tests.
Place and Duration of Study: The study was conducted in the metal recycling industry in Ogijo, Southwestern Nigeria, from November 2021 to October 2022.
Methodology: Daily mean meteorological data including ambient temperature, rainfall, relative humidity (RH), wind speed (WS), wind direction (WD), solar radiation, and ultra-violet radiation were recorded using an automatic weather monitoring system positioned 2.0m above ground level at each sampling location adjacent to the PM2.5 sampler. Data were collected at 5-minute intervals and stored in memory, with data retrieval facilitated by the weather-smart program. Data collection commenced during the dry season of 2021 through wet season of 2022.
Results: The analysis of error evaluation functions revealed that the NGRM exhibited the least deviation from the experimental data compared to the LRM and NPRM. Statistical analysis further demonstrated that the NGRM better represented the experimental data compared to the LRM and NPRM, resulting in the rejection of LRM and NPRM in favour of NGRM for predicting PM2.5 concentration.
Conclusion: The NGRM proved to be the most suitable model for predicting PM2.5 pollutant concentrations in the study area. Temperature and pressure emerged as the most significant predictors of PM2.5 levels.
Keywords: Concentration, mathematical model, metal recycling, pm 2.5, prediction