Modeling jointly low, moderate, and heavy rainfall intensities without a threshold selection

Handle URI:
http://hdl.handle.net/10754/608588
Title:
Modeling jointly low, moderate, and heavy rainfall intensities without a threshold selection
Authors:
Naveau, Philippe; Huser, Raphaël ( 0000-0002-1228-2071 ) ; Ribereau, Pierre; Hannart, Alexis
Abstract:
In statistics, extreme events are often defined as excesses above a given large threshold. This definition allows hydrologists and flood planners to apply Extreme-Value Theory (EVT) to their time series of interest. Even in the stationary univariate context, this approach has at least two main drawbacks. First, working with excesses implies that a lot of observations (those below the chosen threshold) are completely disregarded. The range of precipitation is artificially shopped down into two pieces, namely large intensities and the rest, which necessarily imposes different statistical models for each piece. Second, this strategy raises a nontrivial and very practical difficultly: how to choose the optimal threshold which correctly discriminates between low and heavy rainfall intensities. To address these issues, we propose a statistical model in which EVT results apply not only to heavy, but also to low precipitation amounts (zeros excluded). Our model is in compliance with EVT on both ends of the spectrum and allows a smooth transition between the two tails, while keeping a low number of parameters. In terms of inference, we have implemented and tested two classical methods of estimation: likelihood maximization and probability weighed moments. Last but not least, there is no need to choose a threshold to define low and high excesses. The performance and flexibility of this approach are illustrated on simulated and hourly precipitation recorded in Lyon, France.
KAUST Department:
Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Citation:
Modeling jointly low, moderate, and heavy rainfall intensities without a threshold selection 2016:n/a Water Resources Research
Publisher:
Wiley-Blackwell
Journal:
Water Resources Research
Issue Date:
9-Apr-2016
DOI:
10.1002/2015WR018552
Type:
Article
ISSN:
00431397
Sponsors:
Part of this work has been supported by the ANR-DADA, LEFE-INSU-Multirisk, AMERISKA, A2C2, CHAVANA and Extremoscope projects. The authors acknowledge Meteo France for the Lyon precipitation time series that available to anyone upon request. Part of the work was done when the first author was visiting the IMAGE-NCAR group in Boulder, CO, USA. The authors would also like very much to credit the contributors of the R Core Team [2013]. The data are freely available by sending an email to Philippe Naveau (naveau@lsce.ipsl.fr).
Additional Links:
http://doi.wiley.com/10.1002/2015WR018552
Appears in Collections:
Articles; Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division

Full metadata record

DC FieldValue Language
dc.contributor.authorNaveau, Philippeen
dc.contributor.authorHuser, Raphaëlen
dc.contributor.authorRibereau, Pierreen
dc.contributor.authorHannart, Alexisen
dc.date.accessioned2016-05-08T14:51:39Zen
dc.date.available2016-05-08T14:51:39Zen
dc.date.issued2016-04-09en
dc.identifier.citationModeling jointly low, moderate, and heavy rainfall intensities without a threshold selection 2016:n/a Water Resources Researchen
dc.identifier.issn00431397en
dc.identifier.doi10.1002/2015WR018552en
dc.identifier.urihttp://hdl.handle.net/10754/608588en
dc.description.abstractIn statistics, extreme events are often defined as excesses above a given large threshold. This definition allows hydrologists and flood planners to apply Extreme-Value Theory (EVT) to their time series of interest. Even in the stationary univariate context, this approach has at least two main drawbacks. First, working with excesses implies that a lot of observations (those below the chosen threshold) are completely disregarded. The range of precipitation is artificially shopped down into two pieces, namely large intensities and the rest, which necessarily imposes different statistical models for each piece. Second, this strategy raises a nontrivial and very practical difficultly: how to choose the optimal threshold which correctly discriminates between low and heavy rainfall intensities. To address these issues, we propose a statistical model in which EVT results apply not only to heavy, but also to low precipitation amounts (zeros excluded). Our model is in compliance with EVT on both ends of the spectrum and allows a smooth transition between the two tails, while keeping a low number of parameters. In terms of inference, we have implemented and tested two classical methods of estimation: likelihood maximization and probability weighed moments. Last but not least, there is no need to choose a threshold to define low and high excesses. The performance and flexibility of this approach are illustrated on simulated and hourly precipitation recorded in Lyon, France.en
dc.description.sponsorshipPart of this work has been supported by the ANR-DADA, LEFE-INSU-Multirisk, AMERISKA, A2C2, CHAVANA and Extremoscope projects. The authors acknowledge Meteo France for the Lyon precipitation time series that available to anyone upon request. Part of the work was done when the first author was visiting the IMAGE-NCAR group in Boulder, CO, USA. The authors would also like very much to credit the contributors of the R Core Team [2013]. The data are freely available by sending an email to Philippe Naveau (naveau@lsce.ipsl.fr).en
dc.language.isoenen
dc.publisherWiley-Blackwellen
dc.relation.urlhttp://doi.wiley.com/10.1002/2015WR018552en
dc.rightsArchived with thanks to Water Resources Researchen
dc.titleModeling jointly low, moderate, and heavy rainfall intensities without a threshold selectionen
dc.typeArticleen
dc.contributor.departmentComputer, Electrical and Mathematical Sciences and Engineering (CEMSE) Divisionen
dc.identifier.journalWater Resources Researchen
dc.eprint.versionPublisher's Version/PDFen
dc.contributor.institutionLaboratoire des Sciences du Climat et de l'Environnement, LSCE/IPSL, CNRS-CEA-UVSQ, Université Paris-Saclay; F-91198 Gif-sur-Yvette Franceen
dc.contributor.institutionClaude Bernard University Lyon 1; Franceen
dc.contributor.institutionCNRS; Buenos-Aires Argentinaen
dc.contributor.affiliationKing Abdullah University of Science and Technology (KAUST)en
kaust.authorHuser, Raphaëlen
All Items in KAUST are protected by copyright, with all rights reserved, unless otherwise indicated.