CFE 2019: Start Registration
View Submission - CMStatistics
Title: Patterns in spatio-temporal extremes Authors:  Marco Oesting - University of Siegen (Germany) [presenting]
Alexander Schnurr - University Siegen (Germany)
Abstract: In many applications in environmental sciences, extreme events exhibit a complex spatial-temporal structure which needs to be described in a compressed way. To this end, we propose to investigate patterns originating from the temporal course of various characteristics of the extreme event. Examples include the magnitude of the event, its centroid or the size of the affected area. We verify the existence of the corresponding limit distributions in the framework of regular variation, develop non-parametric estimators and show their asymptotic normality under appropriate mixing conditions. The finite-sample behaviour will be demonstrated in simulated and real data examples.