Skip to Main content Skip to Navigation
Theses

Assimilation variationnelle d'observations de télédétection dans les modèles de fonctionnement de la végétation : utilisation du modèle adjoint et prise en compte de contraintes spatiales

Abstract : Environment and resources management requires characterizing vegetation status. Crop models simulate canopy functioning from climate, cultural practices and plant / soil properties. The ADAM project, in which this thesis fits, aims at combining remote sensing data with information provided by agronomic models, in order to take part in crops reasoning and environment assessment. Through a study model, we proposed an original approach of simultaneous assimilation on a scene, by applying constraints on model parameters according to their variability linked to the pixel, the field, the cultivar. . . The dimension of the control space is thus decreased, as compared to a pixel by pixel solved problem, the latter being ill-posed when few observations are available. This technique allowed both improving strongly retrievals of canopy biophysical variables and reducing the satellite temporal frequency. We then studied the feasibility of applying such a method for a complex mechanist crop model (STICS), which differentiation presented theoretical and practical difficulties. Developing a sensitivity analysis using the STICS adjoint model, we could perform a hierarchical sort of its parameters according to their influence on crop growth. We proved the feasibility of an assimilation technique with spatial constraints in STICS model, which makes it possible to consider many applications in agronomic research.
Complete list of metadatas

Cited literature [257 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00010443
Contributor : Claire Lauvernet <>
Submitted on : Thursday, October 6, 2005 - 4:22:01 PM
Last modification on : Tuesday, February 16, 2021 - 2:04:02 PM
Long-term archiving on: : Friday, April 2, 2010 - 10:41:42 PM

Identifiers

  • HAL Id : tel-00010443, version 1

Collections

Citation

Claire Lauvernet. Assimilation variationnelle d'observations de télédétection dans les modèles de fonctionnement de la végétation : utilisation du modèle adjoint et prise en compte de contraintes spatiales. Modélisation et simulation. Université Joseph-Fourier - Grenoble I, 2005. Français. ⟨tel-00010443⟩

Share

Metrics

Record views

612

Files downloads

2225