– D. Vibert (LAM) : Solar Rotational Tomography : reconstruction of the electronic density in the Solar Corona.

Carte non disponible

Date/heure
Date(s) - 30/05/2013
14 h 00 min - 15 h 00 min

Catégories Pas de Catégories


Solar Rotational Tomography: reconstruction of the electronic density in the Solar Corona. By D. Vibert, LAM Abstract: I will describe the concept of SRT applied to white light coronographic images and then focus on the time variation problem. The usual assumption is that the corona is almost stable during one rotation so that a static tomography can be achieved. This is strongly not valid near the sun and I will show different techniques we developed to address this issue. This problem is largely ill-posed by nature and taking into account the time variation without increasing the size of the observed data set leads to a massively undetermined problem needing a strong amount of regularization. The emergence of efficient L1-norm minimization algorithm targeting large scale problems without loosing too much accuracy let us apply a proximal splitting method to achieve a TV spatio-temporal regularization of the least-square solution.

– D. Vibert (LAM) : Solar Rotational Tomography : reconstruction of the electronic density in the Solar Corona.

Carte non disponible

Date/heure
Date(s) - 30/05/2013
14 h 00 min - 15 h 00 min

Catégories Pas de Catégories


Solar Rotational Tomography : reconstruction of the electronic density in the Solar Corona.\n\nBy D. Vibert, LAM\n\nAbstract :\nI will describe the concept of SRT applied to white light coronographic images\nand then focus on the time variation problem. The usual assumption is that the\ncorona is almost stable during one rotation so that a static tomography can be\nachieved. This is strongly not valid near the sun and I will show different\ntechniques we developed to address this issue. This problem is largely\nill-posed by nature and taking into account the time variation without\nincreasing the size of the observed data set leads to a massively\nundetermined problem needing a strong amount of regularization. The\nemergence of efficient L1-norm minimization algorithm targeting large scale\nproblems without loosing too much accuracy let us apply a proximal\nsplitting method to achieve a TV spatio-temporal regularization of the\nleast-square solution.[