In large urban areas, where many activities occur due to a big number of citizens, physical polluting agents should be carefully assessed in order to protect human health. The monitoring and control of air pollution, acoustical noise, electromagnetic fields, etc., represent a relevant problem to be considered. Therefore, the development of mathematical models able to predict the air pollution behaviour is a very important field of research. In this paper, the authors will use a model based on Time Series (TS) analysis. A large set of field measurements is reconstructed by an appropriate functional form. The proposed model can evaluate the trend and the periodic pattern in the calibration dataset, and can forecast the slope in any future time range. The TS analysed in this work is composed by hourly CO concentrations observed in the urban site of San Nicolas de los Garza, Nuevo Leon, Mexico. To calibrate the model's parameters, a large set of one year field measurements will be analysed: this procedure will highlight a periodicity of 24 hours in the dataset and a substantial absence of any relevant trend. The authors will perform also a validation of the model on two different months, using data not adopted in the calibration phase. This procedure will show that the model is able to reproduce the overall trend and the periodic behaviour present in the large part of the dataset but, at the same time, it cannot predict isolated peaks or sudden fall of CO concentration. In addition, a Principal Component Analysis will be applied to a data set for one year period in order to investigate the relation among CO concentrations and other measured parameters (meteorological conditions and air criteria pollutants). Resulting principal components can be used in future multiple regression analysis in order to mitigate multicollinearity.

An Application of Time Series Analysis for Forecasting and Control of Carbon Monoxide Concentrations

GUARNACCIA, CLAUDIO;QUARTIERI, Joseph;TEPEDINO, CARMINE;
2014-01-01

Abstract

In large urban areas, where many activities occur due to a big number of citizens, physical polluting agents should be carefully assessed in order to protect human health. The monitoring and control of air pollution, acoustical noise, electromagnetic fields, etc., represent a relevant problem to be considered. Therefore, the development of mathematical models able to predict the air pollution behaviour is a very important field of research. In this paper, the authors will use a model based on Time Series (TS) analysis. A large set of field measurements is reconstructed by an appropriate functional form. The proposed model can evaluate the trend and the periodic pattern in the calibration dataset, and can forecast the slope in any future time range. The TS analysed in this work is composed by hourly CO concentrations observed in the urban site of San Nicolas de los Garza, Nuevo Leon, Mexico. To calibrate the model's parameters, a large set of one year field measurements will be analysed: this procedure will highlight a periodicity of 24 hours in the dataset and a substantial absence of any relevant trend. The authors will perform also a validation of the model on two different months, using data not adopted in the calibration phase. This procedure will show that the model is able to reproduce the overall trend and the periodic behaviour present in the large part of the dataset but, at the same time, it cannot predict isolated peaks or sudden fall of CO concentration. In addition, a Principal Component Analysis will be applied to a data set for one year period in order to investigate the relation among CO concentrations and other measured parameters (meteorological conditions and air criteria pollutants). Resulting principal components can be used in future multiple regression analysis in order to mitigate multicollinearity.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11386/4516261
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 15
  • ???jsp.display-item.citation.isi??? ND
social impact