We consider a class of semiparametric models for univariate air pollutant time series which is able to incorporate some stylized facts usually observed in real data, such as missing data, trends, conditional heteroschedasticity and leverage effects. The inference is provided by a semiparametric approach in a two step procedure in which the cycle-trend component is firstly estimated by a local polynomial estimator and then the parametric component is chosen among several ``candidate models'' by the Model Confidence Set and estimated by standard procedures. An application to PM10 concentration in Torino area in the North-Italian region Piemonte is shown.
Volatility Modelling for Air Pollution Time Series
Albano G.Membro del Collaboration Group
;La Rocca M.Membro del Collaboration Group
;Perna C.
Membro del Collaboration Group
2020
Abstract
We consider a class of semiparametric models for univariate air pollutant time series which is able to incorporate some stylized facts usually observed in real data, such as missing data, trends, conditional heteroschedasticity and leverage effects. The inference is provided by a semiparametric approach in a two step procedure in which the cycle-trend component is firstly estimated by a local polynomial estimator and then the parametric component is chosen among several ``candidate models'' by the Model Confidence Set and estimated by standard procedures. An application to PM10 concentration in Torino area in the North-Italian region Piemonte is shown.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
