A sequential calibration strategy for an operational semi-distributed river flow model. Implementation all over France - INRAE - Institut national de recherche pour l’agriculture, l’alimentation et l’environnement Accéder directement au contenu
Communication Dans Un Congrès Année : 2016

A sequential calibration strategy for an operational semi-distributed river flow model. Implementation all over France

Stratégie de calibration séquentielle pour la mise en oeuvre opérationnelle d'un modèle hydrologique sur toute la France

Alban de Lavenne
Guillaume Thirel
Vazken Andréassian
Charles Perrin
Maria-Helena Ramos

Résumé

Developing modeling tools that help to understand the spatial distribution of water resources is a key issue for better management. The dynamic of streamflow depends on (i) the spatial variability of rainfall (which can be successfully handled by distributed or semi-distributed hydrological models), (ii) the heterogeneity of catchment behaviors (which can be dealt with spatially-variable model parameters), and, increasingly, (iii) localized human regulations (reservoirs). Since calibration is only based on discharge measurements at the outlet of the catchment, semi-distributed hydrological models are often difficult to parameterize. Those difficulties are due to the smoothing effect of catchment and to dispersive effects of flow routing combined with numerical issues and measurements uncertainty. We analyzed the question of calibration for a semi-distributed daily model applied to continental France. We developed a sequential calibration, i.e. from upstream to downstream interiors points, in the aim of making an efficient use of interior streamflow data. Instead of calibrating each sub-catchment independently, the strategy relies on constraining downstream parameters sets according to upstream parameters sets. The model used is a recently developed daily semi-distributed model, which is based on a spatial distribution of the lumped GR5J model. The model has five parameters to calibrate for each sub-catchment and a streamflow vitesse parameter for flow routing between them. It implements two buckets, one for production and one for routing, and estimates inter-catchment groundwater flow according to the level of the routing reservoir. It is demonstrated that this new calibration strategy helps to (i) address sensibility issues that can usually lead to unrealistic parameters sets, (ii) facilitate the coherence of the spatial distribution of the parameters. Impact of dams are indirectly considered by using observed discharge time series (instead of simulation) for catchments directly influenced by dam (located upstream). This is done to prevent calibration to try to fit influenced streamflow with unrealistic parameters set. This work is done in comparison with the lumped GR5J model on more than 1100 catchments all over France.

Mots clés

Fichier non déposé

Dates et versions

hal-02605873 , version 1 (16-05-2020)

Identifiants

Citer

Alban de Lavenne, Guillaume Thirel, Vazken Andréassian, Charles Perrin, Maria-Helena Ramos. A sequential calibration strategy for an operational semi-distributed river flow model. Implementation all over France. GEWEX events: Including Water Management in Large Scale Models, Sep 2016, Gif-sur-Yvette, France. ⟨hal-02605873⟩

Collections

IRSTEA INRAE HYCAR
10 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More