CMStatistics 2018: Start Registration
View Submission - CMStatistics
Title: Bayesian hierarchical vine copula models for the analysis of glacier discharge Authors:  Mario Gomez - Universidad Carlos III de Madrid (Spain)
Carmen Dominguez - Universidad de Salamanca (Spain)
Concepcion Ausin - Universidad Carlos III de Madrid (Spain) [presenting]
Abstract: Glacier discharge is the loss of liquid water produced by melting ice. The aim is to analyze the relationship among the glacier discharge and other meteorological variables such as temperature, humidity, solar radiation and precipitation. We propose a Bayesian hierarchical vine copula model where we assume that the dependence structure among variables is different in each season, although governed by common hyperparameters. Further, we also assume a hierarchical structure in each particular season such that the relationship among variables is different for the cases of positive and/or zero precipitation and/or discharge, but with common hyperparameters. Bayesian inference is implemented and compared using both ABC and MCMC techniques. The proposed methodology is applied to a real data set of glacier discharge measured in King George Island in the Antarctica.