Predictive Ability of Improved Neural Network Models to Simulate Pollutant Dispersion

Khandaker M A Hossain

doi:10.1155/2014/141923

Abstract

This paper describes the ability of artificial neural network (ANN) models to simulate the pollutant dispersion characteristics in varying urban atmospheres at different regions. ANN models are developed based on twelve meteorological (including rainfall/precipitation) and six traffic parameters/variables that have significant influence on emission/pollutant dispersion. The models are trained to predict concentration of carbon monoxide and particulate matters in urban atmospheres using field meteorological and traffic data. Training, validation, and testing of ANN models are conducted using data from the Dhaka city of Bangladesh. The models are used to simulate concentration of pollutants as well as the effect of rainfall on emission dispersion throughout the year and inversion condition during the night. The predicting ability and robustness of the models are then determined by using data of the coastal cities of Chittagong and Dhaka. ANN models based on both meteorological and traffic variables exhibit the best performance and are capable of resolving patterns of pollutant dispersion to the atmosphere for different cities.

Highlights

Air pollution is a major environmental concern in major cities around the world
The emission of pollutants such as carbon monoxide (CO), nitrogen oxides (NOx/NO2), and particulate matter (PM) due to high traffic volumes, congestion, and poor vehicle maintenance has resulted in the transport sector being a major contributor to air pollution in major cities around the world [1]
Previous research studies based on artificial neural network (ANN) modeling concentrate on the prediction of NOx/NO2 and CO concentrations or dispersion characteristics in urban atmosphere taking into account meteorological and vehicular parameters [6, 8,9,10]

Summary

Introduction

Air pollution is a major environmental concern in major cities around the world. The major causes of air pollution include rapid industrialization and urbanization and increased non-environment-friendly energy production. One approach to predict pollutant concentrations is to use a detailed atmospheric diffusion model that requires detailed emissions and metrological data [5] Another approach is the regression modeling based on a statistical approach that has been applied to air quality modeling and prediction. Previous research studies based on ANN modeling concentrate on the prediction of NOx/NO2 and CO concentrations or dispersion characteristics in urban atmosphere taking into account meteorological (excluding the effect of rainfall) and vehicular parameters [6, 8,9,10]. This paper presents the performance of refined ANN models for the prediction of concentrations of carbon monoxide and particulate matters (PM10 and PM2.5) in the atmosphere in an urban setup (Dhaka city, Bangladesh) considering both vehicular emissions and meteorological parameters (including precipitation/rainfall as an additional parameter which was not included in previous research studies). The proposed refined and robust ANN models can be very useful for various government agencies and other organizations involved in the air quality management of urban areas

Development of Artificial Neural Network Models

Performance Evaluation

Conclusions