Return Level Estimation of Extreme Rainfall over the Iberian Peninsula: Comparison of Methods

Abstract

Different ways to estimate future return levels (RLs) for extreme rainfall, based on extreme value theory (EVT), are described and applied to the Iberian Peninsula (IP). The study was done for an ensemble of high quality rainfall time series observed in the IP during the period 1961-2010. Two approaches, peaks-over-threshold (POT) and block maxima (BM) with the generalized extreme value (GEV) distribution, were compared in order to identify which is the more appropriate for the estimation of RLs. For the first approach, which identifies trends in the parameters of the asymptotic distributions of extremes, both all-days and rainy-days-only datasets were considered because a major fraction of values of daily rainfall over the IP is zero. For the second approach, rainy-days-only data were considered showing how the mean, variance and number of rainy days evolve. The 20-year RLs expected for 2020 were estimated using these methods for three seasons: autumn, spring and winter. The GEV is less reliable than the POT because fixed blocks lead to the selection of non-extreme values. Future RLs obtained with the POT are greater than those estimated with the GEV, mainly because some gauges show significant positive trends for the number of rainy days. Autumn, rather than winter, is currently the season with the heaviest rainfall for some regions.

Spatial distribution of the level of return of extreme precipitation events in Northeast Brazil

“…In this study, the value of the 95th percentile was used as a threshold, following previous studies (Della-Marta et al, 2009;Santos et al, 2015Santos et al, , 2016Wu et al, 2015;Acero et al, 2018). The 95th percentile was calculated for each seasonal period within each grid point located in NEB, in which only rain records (precipitation greater than 0 mm) were considered.…”

On the Applicability of the Expected Waiting Time Method in Nonstationary Flood Design

“…Recently, many studies have suggested nonstationary approaches to address the aforementioned issue of flood estimation (Yan et al 2017b(Yan et al , 2019Acero et al 2018;Hu et al 2018;Salas and Obeysekera 2014;Cooley 2013;Rootzén and Katz 2013;Parey et al 2010Parey et al , 2007Olsen et al 1998). Among them, two return-period-based methods, i.e., expected waiting time (EWT) (Cooley 2013;Olsen et al 1998) and expected number of exceedances (ENE) (Parey et al 2010(Parey et al , 2007 have drawn considerable attention.…”

Design flood estimation with varying record lengths in Norway under stationarity and nonstationarity scenarios

“…Under the framework of NFFA, if we still follow the design method under stationarity, we can obtain annual design flood for a given return period, which is impractical for engineering design since the relationship between design flood and return period is no longer one-to-one. To address the problem of design flood estimation in NFFA, researchers have carried out many studies and developed several design flood methods in recent years (Olsen et al 1998;Parey et al 2007Parey et al , 2010Cooley 2013;Rootzén & Katz 2013;Acero et al 2017Acero et al , 2018Hu et al 2018;Wang et al 2019;Byun & Hamlet 2020;Lu et al 2020;Yan et al 2020). Yan et al (2019a) comprehensively compared different methods and recommended the use of average design life level (ADLL) and equivalent reliability (Hu et al 2018).…”