Premium
Fitting copulas to bivariate earthquake data: the seismic gap hypothesis revisited
Author(s) -
Nikoloulopoulos Aristidis K.,
Karlis Dimitris
Publication year - 2008
Publication title -
environmetrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.68
H-Index - 58
eISSN - 1099-095X
pISSN - 1180-4009
DOI - 10.1002/env.869
Subject(s) - copula (linguistics) , bivariate analysis , tail dependence , econometrics , joint probability distribution , marginal distribution , statistical physics , range (aeronautics) , mathematics , multivariate statistics , statistics , physics , random variable , materials science , composite material
The seismic gap hypothesis assumes that the intensity of an earthquake and the time elapsed from the previous one are positively related. Previous works on this topic were based on particular assumptions for the joint distribution implying specific type of dependence. We investigate this hypothesis using copulas. Copulas are flexible for modelling the dependence structure far from assuming simple linear correlation structures and, thus, allow for better examination of this controversial aspect of geophysical research. In fact, via copulas, marginal properties and dependence structure can be separated. We propose a model averaging approach in order to allow for model uncertainty and diminish the effect of the choice of a particular copula. This enlarges the range of potential dependence structures that can be investigated. Application to a real data set is provided. Copyright © 2007 John Wiley & Sons, Ltd.