8 March 2013 – Seminar
When: 4pm on Friday, 8 March 2013
Where: DIAS, Geophysics Section, 5 Merrion Square, Dublin 2, (library)
Speaker: Prof. Kerry Gallagher (University of Rennes, France)
Title: Transdimensional (Inverse) Modelling in Earth Sciences
Abstract:
In many modelling problems in Earth Science, the number of parameters is fixed in advance. While suitable for deterministic forward modelling, this can be limiting for inverse modelling. In this case, arguments can be used to bound the maximum number of parameters that need to be considered or the number of unknowns is set at some arbitrary value. Ideally we would like to treat the number of unknowns itself directly as an unknown. This leads to a transdimensional problem, that is, one where the dimension of the parameter space is a variable to be solved for. I will present a brief overview of an approach used to solve such problems, and illustrate its application to a range of Earth Science problems such as the inference of past climate from geochemical data in boreholes, construction of seismic tomography models, and the decomposition of geochronological data to identify mixture components.
Leave a Comment
Last Updated: 22nd March 2016 by Louise Collins
2013-03-08 – Seminar: Kerry Gallagher
8 March 2013 – Seminar
When: 4pm on Friday, 8 March 2013
Where: DIAS, Geophysics Section, 5 Merrion Square, Dublin 2, (library)
Speaker: Prof. Kerry Gallagher (University of Rennes, France)
Title: Transdimensional (Inverse) Modelling in Earth Sciences
Abstract:
In many modelling problems in Earth Science, the number of parameters is fixed in advance. While suitable for deterministic forward modelling, this can be limiting for inverse modelling. In this case, arguments can be used to bound the maximum number of parameters that need to be considered or the number of unknowns is set at some arbitrary value. Ideally we would like to treat the number of unknowns itself directly as an unknown. This leads to a transdimensional problem, that is, one where the dimension of the parameter space is a variable to be solved for. I will present a brief overview of an approach used to solve such problems, and illustrate its application to a range of Earth Science problems such as the inference of past climate from geochemical data in boreholes, construction of seismic tomography models, and the decomposition of geochronological data to identify mixture components.
Category: Geophysics Section News & Events
Recent Posts
Internationally acclaimed scientific integrity expert Elisabeth Bik to deliver DIAS Day Lecture
Earth Day 2022: DIAS Contributing to the sustainability of our world
DIAS Cosmic Physics marks 75 years of ground-breaking discoveries about our planet and universe
In a world first, Irish scientists observe cosmic explosion with unprecedented detail
DIAS celebrates success of female researchers for International Women’s Day 2022
Language switcher
Tweets by DIAS_Dublin