Premium
Changes in extreme precipitation over Spain using statistical downscaling of CMIP5 projections
Author(s) -
Monjo Robert,
Gaitán Emma,
Pórtoles Javier,
Ribalaygua Jaime,
Torres Luis
Publication year - 2016
Publication title -
international journal of climatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.58
H-Index - 166
eISSN - 1097-0088
pISSN - 0899-8418
DOI - 10.1002/joc.4380
Subject(s) - gumbel distribution , climatology , downscaling , precipitation , coupled model intercomparison project , environmental science , quantile , extreme value theory , generalized extreme value distribution , weibull distribution , climate change , climate model , return period , meteorology , statistics , mathematics , geography , geology , oceanography , archaeology , flood myth
The Mediterranean coast of Spain often experiences intense rainfall, sometimes reaching remarkable amounts of more than 400 mm in one day. The aim of this work is to study possible changes of extreme precipitation in Spain for this century, simulated from several Coupled Model Intercomparison Project Phase 5 ( CMIP5 ) climate models. Eighteen climate projections (nine models under RCP4.5 and nine RCP8.5 scenarios) were downscaled using a two‐step analogue/regression statistical method. We have selected 144 rain gauges as the rainiest of a network by using a threshold of 250 mm in one day for a return period of 100 years. Observed time‐series have been extended using the ERA40 reanalysis and have subsequently been used to correct the climate projections according to a parametric quantile–quantile method. Five theoretical distributions (Gamma, Weibull, Classical Gumbel, Reversed Gumbel and Log‐logistic) have been used to fit the empirical cumulative functions (entire curves, not only the upper tail) and to estimate the expected precipitation according to several return periods: 10, 20, 50 and 100 years. Results in the projected changes for 2051–2100 compared to 1951–2000 are similar (in terms of sign and value) for the four return periods. The analysed climate projections show that changes in extreme rainfall patterns will be generally less than the natural variability. However, possible changes are detected in some regions: decreases are expected in a few kilometres inland, but with a possible increase in the coastline of southern Valencia and northern Alicante, where the most extreme rainfall was recorded. These results should be interpreted with caution because of the limited number of climate projections; anyway, this work shows that the developed methodology is useful for studying extreme rainfall under several climate scenarios.