The problem of river flow time series clustering isdiscussed. In particular, the Autoregressive metric, originallydefined for ARIMA processes, is considered and theasymptotic distribution of the AR metric is derived. The comparison of timeseries is set in an hypotheses testing framework and a clusteringtechnique is illustrated. Finally, the proposed technique is applied for the analysis of unregulatedstreamflow river series observed in the main basins in Oregon.
The AR metric for the classification of hydrological time series / Corduas, Marcella. - (2007). (Intervento presentato al convegno Statistica e Gestione delle Risorse Ambientali tenutosi a Università di Napoli-Facoltà di Scienze Politiche nel 28 settembre 2007).
The AR metric for the classification of hydrological time series
CORDUAS, MARCELLA
2007
Abstract
The problem of river flow time series clustering isdiscussed. In particular, the Autoregressive metric, originallydefined for ARIMA processes, is considered and theasymptotic distribution of the AR metric is derived. The comparison of timeseries is set in an hypotheses testing framework and a clusteringtechnique is illustrated. Finally, the proposed technique is applied for the analysis of unregulatedstreamflow river series observed in the main basins in Oregon.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.