Attribution of climate records in a multi-model max-stable framework
| Quand ? |
Le 22/09/2016, de 10:45 à 12:00 |
|---|---|
| Où ? | CBP room 1027 |
| Participants |
Philippe Naveau |
| Ajouter un événement au calendrier |
vCal
iCal
|
Both climate and statistical models play an essential role in the process of demonstrating that some atmospheric variable has changed over time and in establishing the most likely causes for the detected change. One difficulty in the research field of Detection and Attribution resides in defining events that can be easily compared and accurately inferred from reasonable sample sizes. As most impacts studies focus on extreme events, the inference of small probabilities and the computation of their associated uncertainties quickly becomes challenging.
In the particular context of event attribution, we address the question of how to compare records between the so-called world as "it may have been been without antropogenic forc- ings" and the "world that is". Records are often the most important events in terms of impact and we will show that choosing records as the event of interest can improve the inference of the ratio of two small probability events. The gain is particularly substantial if we leverage the tools and hypotheses used in Extreme Value Theory. To illustrate our approach, the Fraction of Attributable Risk, an often used indicator in event attribution studies, is modified and tailored to handle records. We illustrate the advantages of our method through theoretical results, simulation studies, temperature records in Paris and outputs from a numerical climate model.