Skip to Main content Skip to Navigation
Theses

Analyse d'images satellitaires d'inondations pour la caractérisation tridimensionnelle de l'aléa et l'aide à la modélisation hydraulique

Abstract : The thesis aims at deploying methods of flood satellite image analysis beyond 2D flood area delineation in order to estimate water levels and to help hydraulic modelling. Based on Raclot (2003) works with aerial photographs, which provide 20 cm mean uncertainty, the water level estimation method use satellite RADAR images and fine DEM. The method is composed of two steps : i) flood cartography and analysis of image hydraulic relevance for water level estimation, ii) fusion between relevant information resulting from these images with a fine Digital Elevation Model (DEM) and constraining the water levels extracted from images by concepts of coherence with respect to a hydraulic flow through a plain. It provides water level estimations with 38 cm mean uncertainty.To help hydraulic modelling, the PhD aims at reducing equifinality thanks to satellite images of flood. To meet this aim, a traditional step of calibration thanks to hydrographs is completed by comparison between simulation results and flood extends or water levels extracted from images. To deals with calibration uncertainties, GLUE methodology -Beven (1992)- has been used. In perspective, the results of the PhD imply great benefits for flood evolution forecasting after acquisition of flood satellite images because the use of these images as initial conditions or calibration data provide better-constrained models.The main study area is a 28 km section of the Mosel River upstream boundary between France, Germany and Luxembourg. The flood image used has been acquired by RADARSAT-1 on February 02/28/1997.
Document type :
Theses
Complete list of metadata

Cited literature [222 references]  Display  Hide  Download

https://hal.inrae.fr/tel-02588783
Contributor : Migration Irstea Publications <>
Submitted on : Friday, May 15, 2020 - 12:58:25 PM
Last modification on : Tuesday, June 15, 2021 - 2:57:35 PM

File

pub00021115.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : tel-02588783, version 1
  • IRSTEA : PUB00021115

Citation

R. Hostache. Analyse d'images satellitaires d'inondations pour la caractérisation tridimensionnelle de l'aléa et l'aide à la modélisation hydraulique. Sciences de l'environnement. Doctorat Sciences de l'eau, Engref, Montpellier, 2006. Français. ⟨tel-02588783⟩

Share

Metrics

Record views

180

Files downloads

367