RETURN LEVEL ESTIMATION FOR EXTREME RAINFALL OVER THE IBERIAN PENINSULA: COMPARING METHODOLOGIES.

Different ways to estimate future return levels for extreme rainfall are described and applied to the Iberian Peninsula (IP), based on Extreme Value Theory (EVT). This study is made for an ensemble of high quality rainfall time series observed in the Iberian Peninsula over the period 1961-2010. Both, peaks-over-threshold (POT) approach and block maxima with the Generalized Extreme Value (GEV) distribution will be used and their results compared when linear trends are assumed in the parameters: threshold and scale parameter for POT and location and scale parameter for GEV. Both all-days and rainy-days-only data sets were considered, because rainfall over the IP is a special variable in that a large number of the values are 0. Another methodology is then tested, for rainy days only, considering the role of how the mean, variance, and number of rainy days evolve. The 20-year return levels (RLs) expected in 2020 were estimated using these methodologies for three seasons: fall, spring and winter. GEV is less reliable than POT because fixed blocks lead to the selection of non-extreme values. Future RLs obtained with POT are higher than those estimated with GEV, mainly for some observatories showing significant positive trend for the number of rainy days. Fall becomes the season with heaviest rainfall, rather than winter nowadays, for some regions.