In a previous paper, exploratory analysis methods were applied to a dataset of tree-ring width synchronous series to ascertain to what extent the methods could enlighten on the structure of the data, in view to recognize how similar are the tree-ring growth series. In particular, the Evolutionary Principal Component Analysis (EPCA) was proposed as a useful method to describe the variation of the correlation structure of the series through time. In this paper, an analogous methodology is proposed to study a dataset composed of tree-rings synchronous chronologies of Araucaria araucana on the Argentinian Andes. We aim at recognizing the existence of important variations in the correlation structure of the chronologies. Here, we use both Principal Component Analysis (PCA) and Hierarchical Factor Classification (HFC) to give an overall overview of the data structure, and we applied a EPCA methodology to identify the pattern of the correlation structure through time. This allows the identification of both the time-intervals in which the structure keeps more homogeneous and those periods where mayor changes in the correlation structure between series occurs.
Evolutionary analysis applied to tree-ring series / Camiz, Sergio; F. A., Roig. - ELETTRONICO. - (2011), pp. 30-37. (Intervento presentato al convegno E-ICES 6 tenutosi a Malargüe - Argentina nel 4-8/10/2010).
Evolutionary analysis applied to tree-ring series
CAMIZ, Sergio;
2011
Abstract
In a previous paper, exploratory analysis methods were applied to a dataset of tree-ring width synchronous series to ascertain to what extent the methods could enlighten on the structure of the data, in view to recognize how similar are the tree-ring growth series. In particular, the Evolutionary Principal Component Analysis (EPCA) was proposed as a useful method to describe the variation of the correlation structure of the series through time. In this paper, an analogous methodology is proposed to study a dataset composed of tree-rings synchronous chronologies of Araucaria araucana on the Argentinian Andes. We aim at recognizing the existence of important variations in the correlation structure of the chronologies. Here, we use both Principal Component Analysis (PCA) and Hierarchical Factor Classification (HFC) to give an overall overview of the data structure, and we applied a EPCA methodology to identify the pattern of the correlation structure through time. This allows the identification of both the time-intervals in which the structure keeps more homogeneous and those periods where mayor changes in the correlation structure between series occurs.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.