Logo image
Modélisation conceptuelle de la transformation pluie-débit dans le contexte méditerranéen
Thèses et HDR   Open Access

Modélisation conceptuelle de la transformation pluie-débit dans le contexte méditerranéen

Antoine Hreiche
Doctoral, Université de Montpellier
09/10/2003

Résumé

Hydrologie bassins versants méditerranéens relation pluie-débit modélisation conceptuelle
MEDOR, a daily lumped conceptual rainfall-runoff model with four parameters was conceived for Mediterranean catchments. It was applied on six Mediterranean basins, French and Lebanese. Its calibration is affected by the equifinality issue. Systematic scanning of the Nash criterion objective function shows the existence of an equifinality relationship among the two loss function parameters (PER) independently of the transfer parameters. The agglomerated Nash criterion with variable terms determines the PER using only the annual balance of rainfall-runoff and the daily stochastic structure of rainfall.<br />Coupled to a stochastic model of rainfall of a given region, MEDOR generates equifinality relations between runoff coefficients (CR) defining a surface in the parameters space. Five zones have been identified in the Mediterranean region having a single reference CR surface (e.g., East cost of the Mediterranean Sea). The runoff coefficient of a given watershed located in one of these areas fixes the specific PER of the catchments. The property of spatio-temporal additivity is demonstrated on MEDOR. Model parameters for a catchment made of several sub-catchments are given by specific laws using relative parameters of each sub-catchment. One of the loss parameters is related to the useful depth of the soil, which allows the transformation of MEDOR from a lumped model into a semi-distributed one. The use of more catchments would allow to specify the significance of the parameters, and thus to consider the application of the model on ungauged basins.

Fichiers et liens (1)

url
Find in HALAfficher

Indicateurs

1 Consultations de la notice

Détails

Logo image