Skip to Main content Skip to Navigation
Theses

Estimation des conditions océanographiques par inversion de données issues d'un radar imageur non calibré

Abstract : The statistical characterization of « sea clutter » is an important challenge in microwave radar remote sensing. It mostly relies on purely empirical models which cannot be interpreted in terms of sea state and do not comply to an inversion process. To model the statistical distribution of the backscattered intensity we use a Two-Scale approach parametrized by the radar resolution cell and the mean square slopes of waves. This model allows an analytical description of the normalized intensities with help of a single slope parameter. It matches perfectly the experimental observation when the latter is tuned in each polarization channel. However, inconsistent values are found for this tuning parameter obtained in fifferent polarizations. This is explained by a poor estimation of the polarisazatio ratio of the radar cross-section predicted by the model. To correct this shortcoming we developed an original analytical formulation of the Two-Scale model which allows to restore in a simple way the correct order of magnitude of the polarization ratio, regardless of the radar frequency and sea state. This improved version has veen validated with several experimental comparisons in various configurations ; it leads to a consistent modelling of the sea clutter statistics in the different polarization channels. In the end we have devised a inversion process to estimate a sea state from non-calibrated intensities on an imaging radar. The relevance of this model has been established with the help of various data sets including those with calibration issues.
Complete list of metadata

https://hal.archives-ouvertes.fr/tel-03200071
Contributor : Véronique Soullier <>
Submitted on : Friday, April 16, 2021 - 10:57:38 AM
Last modification on : Wednesday, July 21, 2021 - 3:14:43 AM
Long-term archiving on: : Saturday, July 17, 2021 - 6:26:50 PM

File

DEMR21016.1618305179.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : tel-03200071, version 1

Collections

Citation

Floriane Schreiber. Estimation des conditions océanographiques par inversion de données issues d'un radar imageur non calibré. Traitement du signal et de l'image [eess.SP]. Université de Toulon, 2020. Français. ⟨tel-03200071v1⟩

Share

Metrics

Record views

63

Files downloads

87