Please use this identifier to cite or link to this item: http://hdl.handle.net/2445/69730
Title: Trends in ozone concentrations in the Iberian Peninsula by quantile regression and clustering
Author: Monteiro, Alvaro A. N.
Ribeiro, I.
Scotto, M.
Barbosa, S.
Alonso, A.
Baldasano Recio, José M. (José María)
Pay Pérez, María Teresa
Miranda, A.I.
Borrego, C.
Carvalho, A.
Keywords: Ozó atmosfèric
Anàlisi de conglomerats
Anàlisi de regressió
Península Ibèrica
Atmospheric ozone
Cluster analysis
Regression analysis
Iberian Peninsula
Issue Date: Sep-2012
Publisher: Elsevier Ltd
Abstract: In this paper, 10-years of ozone (O3) hourly concentrations collected over the period 2000-2009 in the Iberian Peninsula (IP) are analyzed using records from 11 background sites. All the selected monitoring stations present an acquisition efficiency above 85%. The changes in surface ozone over the Iberian Peninsula are examined by means of quantile regression, which allows to analyse the trends not only in the mean but in the overall data distribution. In addition, the ozone hourly concentrations records are clustered on the basis of their resulting distributions. The analysis showed that high altitude stations (>900 m) have higher background O3 concentrations (∼80 μg m−3). The same magnitude of background O3 concentrations is found in stations near the Mediterranean Sea. On the other hand, the rural stations near the Atlantic coast present lower background values (∼50-60 μg m−3) than those of Mediterranean influence. The two sub-urban stations exhibit the lowest background concentrations (∼45 μg m−3). The results of the quantile regression show a very distinct behaviour of the data distribution, the slopes for a fixed quantile are not the same over IP, reflecting the spatial dependence of O3 trends. Hence the rate of temporal change is not the same for all parts of the data distribution, as implicitly assumed in ordinary regression. The lower quantile (percentile 5) presents higher rates of change than the middle (percentile 50) and the upper quantile (percentile 95). The clustering procedure reveals what has been already detected in the quantile regression. The station with highest rates of decrease on the O3 concentrations (easternmost station of IP) is isolated and then other clusters are formed among the moderately positive/negative O3 trends around the IP. The clustering procedure highlighted that the largest trends are found for the lower ozone O3 values, with largest negative trend at the easternmost station of IP, and also in northern and mainland stations, and an opposite behaviour, with positive O3 trends, is observed at the Atlantic coast stations.
Note: Versió postprint del document publicat a: http://dx.doi.org/10.1016/j.atmosenv.2012.03.069
It is part of: Atmospheric Environment, 2012, vol. 56, p. 184-193
URI: http://hdl.handle.net/2445/69730
Related resource: http://dx.doi.org/10.1016/j.atmosenv.2012.03.069
ISSN: 1352-2310
Appears in Collections:Articles publicats en revistes (Genètica, Microbiologia i Estadística)

Files in This Item:
File Description SizeFormat 
657156.pdf1.37 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.