Skip to Main content Skip to Navigation

Méthodes de traitement de données géophysiques par transformée en ondelettes.

Abstract : This thesis concerns the development of new processing and interpretation methods of potential field data. First of all, we have elaborated a fast algorithm to den oise geophysical data with the help of orthogonal wavelets transform. The method takes into account non-stationarities of the signal-to-noise ratio. It is based on statistical criteria which allow to distinguish the wavelet coefficients generated by the signal and these by the noise. The signal is then reconstructed with the preserved coefficients by an inverse transform. The second method was conduct to develop a structural imaging of the substratum from a field measured at the surface. We have first inserted the potential theory into the wavelet theory owing to the homogeneity of fields produced by multipolar sources. The introduction of wavelets satisfying the Poisson equation allow to define precisely the location and the homogeneity of the existing sources. The use of the ridge function of the wavelet transform makes the method robust in regards of noise. We have generalized the method to extended sources in order to localize and characterize edge structures. Their slopes is then determined by the use of complex wavelets. Finally, this formulation of the inverse problem gives a tool to develop a filtering from structural criteria (type and depth of sources). We give sorne applications on gravimetric data. The feasability of the method is then illustrated from examples on fracture zones and subduction zones.
Complete list of metadata
Contributor : Isabelle Dubigeon Connect in order to contact the contributor
Submitted on : Tuesday, January 3, 2012 - 2:35:33 PM
Last modification on : Friday, March 4, 2022 - 2:39:06 PM
Long-term archiving on: : Wednesday, April 4, 2012 - 2:45:21 AM


  • HAL Id : tel-00656040, version 1


Frédérique Moreau. Méthodes de traitement de données géophysiques par transformée en ondelettes.. Géophysique [physics.geo-ph]. Université Rennes 1, 1995. Français. ⟨tel-00656040⟩



Record views


Files downloads