Dresden 2014 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
DY: Fachverband Dynamik und Statistische Physik
DY 41: Poster - Pattern/ Nonlinear Dyn./ Fluids/ Granular/ Critical Phen.
DY 41.35: Poster
Thursday, April 3, 2014, 17:00–19:00, P3
Contemplating coincidences: Rigorous testing for synchrony between event-related time series — •Reik V. Donner1,2 and Jonathan F. Donges2,3 — 1Max Planck Institute for Biogeochemistry, Jena, Germany — 2Potsdam Institute for Climate Impact Research, Potsdam, Germany — 3Stockholm Resilience Centre, Stockholm University, Sweden
Studying temporal point processes like the timings of extreme events is a common problem in various fields of sciences. When series of distinct types of events are available in a common time frame, assessing their statistical interrelationships can be valuable for testing theories proposing specific causal relationships. Here, we introduce a simple statistical framework for computing the probability that the observed number of simultaneous events in two series is due to chance. While in the case of rare events and no serial correlations, this probability can be well approximated analytically, introducing short- or long-range correlations as typical in real-world data can result in substantial deviations from this theory. We systematically study these deviations based on numerical results for some generic stochastic processes as well as real-world examples related to climate and ecosystem variability. Formal links to other techniques like Ripley's cross-K function from spatial statistics, or event synchronization and cross-recurrence rate from nonlinear time series analysis are briefly addressed.