Title: Patterns in spatio-temporal extremes with an application to red sea surface temperature data
Authors: Marco Oesting - University of Stuttgart (Germany) [presenting]
Raphael Huser - King Abdullah University of Science and Technology (Saudi Arabia)
Abstract: In many applications in environmental sciences, extreme events exhibit a complex spatial-temporal structure that needs to be described in a compressed way. To this end, we propose to investigate not only the duration of the event, but also patterns originating from the temporal course of characteristics describing the event and its spatial structure. Examples include the magnitude of the event or the size of the affected area. In the mathematical framework of regular variation of stochastic processes, we verify the existence of meaningful limit distributions for the patterns of interest and develop non-parametric estimators. These estimators are applied to analyze extreme sea surface temperature in the Red Sea.