Prediction of ozone concentration in ambient air using multivariate methods

Lengyel, Ákos József; Héberger, Károly; Paksy, L.; Bánhidi, O.; Rajkó, Róbert

doi:10.1016/j.chemosphere.2004.07.043

Cited by 68 publications

(40 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The obtained regression equation (6) shows that the combined contribution of ozone precursors and meteorological data account for up to 78% of the concentration of this pollutant. The other two equations (4) and (5) show lower values, but these are significant for the clear differentiation of the relative participation of each pollutant in overall ozone concentration.…”

Section: Results From Multiple Linear Regression Analysismentioning

confidence: 99%

See 1 more Smart Citation

Multivariate Analysis for Modeling of Air Pollutants and Ozone Concentration in Dimitrovgrad, Bulgaria

Gocheva-Ilieva¹,

Ivanov²,

Iliev³

2016

IJESD

View full text Add to dashboard Cite

Abstract-Air pollution is one of the key problems in urban areas and its investigation is vital both for people's health but also for the environment as a whole. In particular, ozone is a secondary air pollutant with concentrations dependent mainly on changes in the levels of other pollutants and meteorological conditions within a given region. This paper presents a multivariate analysis of hourly data on 9 air pollutants and 6 meteorological variables in the town of Dimitrovgrad, Bulgaria over a period of 7 years and 3 months. Yeo-Johnson power transformation is applied to each air pollutant variable to improve normality and symmetry. The dominant patterns in the considered data are examined with the help of Principal Component Analysis (PCA) and factor analysis. Furthermore, particular focus is given to determining the concentration levels of ozone in relation to the other air pollutants and 6 meteorological parameters using multiple linear regression. The fitting of the obtained models with coefficients of determination R2 over 78% are obtained. The results are interpreted. Index Terms-Air pollution modeling, factor analysis, multiple linear regression, ozone concentration, principal component analysis (PCA). I. INTRODUCTIONThe monitoring, investigation and control of ambient air quality is a topical issue, very important for preserving the human health and the environment. Directives and regulatory restrictions are established in all European countries and worldwide for permissible concentrations of air pollutants [1]-[3]. In Bulgaria, 12 types of pollutants are systematically monitored by more than 36 automated stations run by the Executive Environment Agency which manages and coordinates activities related to the control and environmental protection of the country. The availability of a huge amount of collected data allows their statistical examination and makes it possible to find significant patterns, as well as dependencies within the data enabling the prediction of future states.In The objective of this paper is to investigate the relationships between large number of air pollutants in Dimitrovgrad, a town in Bulgaria, over an extended period of 7 years and 3 months. The specific goals of the study are: 1) Establishing the existence and determining the type of correlations between the investigated air pollutants; 2) Defining patterns and possible groups of pollutants which are closely related, and classifying the pollutants; 3) Obtaining multivariate regression models which describe the changes in ozone pollution levels in relation to the other pollutants and meteorological data. II. MATERIALS AND METHODSWe examine air quality in the town of Dimitrovgrad, a typical urban region in South-central Bulgaria, 220 km away from the capital city of Sofia. The town is located in the Thracian valley on the banks of the Maritza river at an altitude of 125 m above sea level. It has about 40 000 inhabitants.The study was carried out based on hourly data about concentration of air pollutants between 1st January 2007...

show abstract

Section: Results From Multiple Linear Regression Analysismentioning

confidence: 99%

“…The optimal values of  were chosen to give the smallest possible value of the Jarque-Bera test of normality [14], defined as 22 6 24…”

Section: Multivariate Analysis For Modeling Of Air Pollutants Andmentioning

confidence: 99%

Multivariate Analysis for Modeling of Air Pollutants and Ozone Concentration in Dimitrovgrad, Bulgaria

Gocheva-Ilieva¹,

Ivanov²,

Iliev³

2016

IJESD

View full text Add to dashboard Cite

show abstract

“…To investigate the influence of meteorological parameters on ozone volume fractions, a PCA model has been developed (Lengyel et al 2004;Abdul-Wahab et al 2005;Kovač-Andrić et al 2009;Gvozdić et al 2011). PCA, followed by varimax rotation, yields the results given in Fig.…”

Section: Pca and Influence Of Meteorologymentioning

confidence: 99%

Assessment of ozone variations and meteorological influences in a tourist and health resort area on the island of Mali Lošinj (Croatia)

Kovač-Andrić

Gvozdić

Herjavić

et al. 2013

Environ Sci Pollut Res

View full text Add to dashboard Cite

The purpose of this study was to investigate ozone, variations, and its correlation with meteorological parameters at a remote location on the Mali Lošinj Island, which has been a tourist and health resort area in the northern Adriatic. The measured data are discussed in relation to the EU guidelines (Directive 2002/3/EC; Directive 2008/50/EC). In order to characterize ambient air with respect to ozone vegetation injury and photochemical pollution, we calculated accumulated dose over a threshold of 40 parts per billion index and two photochemical pollution indicators. The influence of local meteorological parameters on the measured ozone volume fractions was also investigated. We used the multivariate technique principal component analysis to trace correlations between measured ozone concentration and meteorological parameters.

show abstract

“…Shively and Sager (1999) extended the work of Cox and Chu (1992) as well as Bloomfield et al, (1996) by using nonparametric regression models to model ozone. The use of multivariate methods was further supported by Lengyel et al, (2004) who analyzed air quality data of which the hidden structure was uncovered by factor analysis and modeled ozone concentration using MLR, PLS, and PCR. While PCA is a very useful tool for selection of properties and different qualities processes leading to a linear model of the data, MLR and PCR or PLS can predict ozone concentration with an error below 2, 5, and 1ppb levels, respectively.…”

Section: Introductionmentioning

confidence: 99%

“…One common approach is the use of a parametric regression model to link some characteristics of ozone, such as the mean level of ozone to meteorological variables. Other scientists have equally used PCA to pattern the spatial and temporal variations of ozone and to identify the important factors influencing ozone concentration [Klaus et al, 2001;Lengyel et al, 2004;Pissimanis et al, 2000]. Different subregions have, however, been identified where ozone concentration exhibited characteristic spatial and temporal patterns based on the differences arising from the interaction of their respective meteorological conditions with anthropogenic effects [Alvarez et al, 2000].…”

Section: Introductionmentioning

confidence: 99%