B0421
Title: Spatiotemporal modeling of extreme events and analysis of their extent
Authors: Ana C Cebrian - University of Zaragoza (Spain) [presenting]
Erin Schliep - NC State University (United States)
Alan Gelfand - Duke University (United States)
Jesus Asin - University of Zaragoza (Spain)
Jorge Castillo-Mateo - University of Zaragoza (Spain)
Abstract: Modeling for extreme heat events (EHE) is customarily implemented using exceedances of a suitable threshold in temperature series. A space-time Bayesian model is developed that enables the prediction of both the incidence and characteristics of EHEs occurring at any location in a study region. The model employs a two-state model for EHEs with local thresholds to fit daily temperature. The model switches between two observed states, one that defines extreme heat days (those above the temperature threshold) and the other that defines non-extreme heat days. This two-state structure allows temporal dependence of the observations but also that the parameters which control the spatial dependence can differ between the two states. The transition probabilities are driven by a two-state Markovian switching model. Each sub-model includes seasonal terms, covariates and intercepts modelled as Gaussian processes. A formal definition of the spatial extent of an extreme event is also introduced and it is illustrated how it can be calculated using the output from the previous model. For a specified region and day, the spatial extent is calculated as the block average of indicator functions over the region. The risk assessment examines Aragon (NE of Spain) and comparisons are made across decades to reveal evidence of increasing extent over time.
