Assessing causality from multivariate time series
- 25 August 2005
- journal article
- research article
- Published by American Physical Society (APS) in Physical Review E
- Vol. 72 (2) , 026222
- https://doi.org/10.1103/physreve.72.026222
Abstract
In this work we propose a general nonparametric test of causality for weakly dependent time series. More precisely, we study the problem of attribution, i.e., the proper comparison of the relative influence that two or more external dynamics trigger on a given system of interest. We illustrate the possible applications of the proposed methodology in very different fields like physiology and climate science.Keywords
This publication has 64 references indexed in Scilit:
- Corrections to the Mann et. al. (1998) Proxy Data Base and Northern Hemispheric Average Temperature SeriesEnergy & Environment, 2003
- Direction of coupling from phases of interacting oscillators: An information-theoretic approachPhysical Review E, 2003
- Global Climate Models Violate Scaling of the Observed Atmospheric VariabilityPhysical Review Letters, 2002
- Evolution of the Sun's Spectral Irradiance Since the Maunder MinimumGeophysical Research Letters, 2000
- Northern hemisphere temperatures during the past millennium: Inferences, uncertainties, and limitationsGeophysical Research Letters, 1999
- Solar modulation of cosmogenic nuclide production over the last millennium: comparison between 14C and 10Be recordsEarth and Planetary Science Letters, 1997
- Natural and anthropogenic changes in atmospheric CO2 over the last 1000 years from air in Antarctic ice and firnJournal of Geophysical Research: Atmospheres, 1996
- Measuring statistical dependences in a time seriesJournal of Statistical Physics, 1993
- Phase space topography and the Lyapunov exponent of electrocorticograms in partial seizuresBrain Topography, 1990
- Investigating Causal Relations by Econometric Models and Cross-spectral MethodsEconometrica, 1969