PREDICTIVE MODEL FOR THE ASSESSMENT OF VOCS EMISSION IN OPEN DUMPSITE USING GENERALIZED LINEAR MODEL

Ambient air quality assessment and regulations plays an important role in air quality management. Forty-five (45) air samples were collected for both wet and dry season of the year (October 2016- September 2017). The air samples were collected by passive sampler (Drager ORSA 5). The samples were analyzed for VOCs, PM2.5 and some elements according to standard methods. In this study, a method to predict aromatic VOCs concentrations in the urban outdoor environment was developed based on the Generalized Linear Models (GLM). The method is based on the relationship developed between atmospheric concentrations of air pollutants and meteorological variables (i.e. ambient temperature, relative humidity and wind speed). The developed GLM model considers VOCs concentrations as a dependent variable, and PM2.5 pollutants and meteorological parameters as explanatory independent variables. A comparison of obtained prediction with observed values showed that the developed model indicated good performance in estimating VOCs concentration. It should be mentioned that the puissance of the model developed is a function of the covariates. This observation suggest that the model with predicted meteorological variables can be engage in environmental management system were data are not readily available. Keywords: VOC, Generalized Linear Models (GLM), Open Dumpsite, Benin City