15 Pages Posted: 25 Oct 2007
Date Written: September 1, 2006
The 2003 heat wave dramatically affected France. The aim of this paper is to get an accurate estimate of its return period. Using minimal temperature in Paris (which has been abnormally high for a dozen days), two models are considered here: either short memory process, with heavy tailed noise (since temperature were quite high), or long memory process with light tailed noise (since heat wave can be interpreted as a persistent problem). As we will see, depending on the definition of the 'heat wave event' in Paris, the results are substentially different for the two models.
Keywords: Heat wave, long range dependence, return period, GARMA processes
Suggested Citation: Suggested Citation