Non-stationary flood frequency analysis in continental Spanish rivers, using climate and reservoir indices as external covariates

摘要

Recent evidences of the impact of persistent modes ofregional climate variability, coupled with the intensification of humanactivities, have led hydrologists to study flood regime without applying thehypothesis of stationarity. In this study, a framework for flood frequencyanalysis is developed on the basis of a tool that enables us to address themodelling of non-stationary time series, namely, the "generalized additivemodels for location, scale and shape" (GAMLSS). Two approaches tonon-stationary modelling in GAMLSS were applied to the annual maximum floodrecords of 20 continental Spanish rivers. The results of the first approach,in which the parameters of the selected distributions were modelled as afunction of time only, show the presence of clear non-stationarities in theflood regime. In a second approach, the parameters of the flooddistributions are modelled as functions of climate indices (ArcticOscillation, North Atlantic Oscillation, Mediterranean Oscillation and theWestern Mediterranean Oscillation) and a reservoir index that is proposed inthis paper. The results when incorporating external covariates in the studyhighlight the important role of interannual variability in low-frequencyclimate forcings when modelling the flood regime in continental Spanishrivers. Also, with this approach it is possible to properly introduce theimpact on the flood regime of intensified reservoir regulation strategies.The inclusion of external covariates permits the use of these models aspredictive tools. Finally, the application of non-stationary analysis showsthat the differences between the non-stationary quantiles and theirstationary equivalents may be important over long periods of time.