We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Trans‐Dimensional Surface Reconstruction With Different Classes of Parameterization.
- Authors
Hawkins, Rhys; Bodin, Thomas; Sambridge, Malcolm; Choblet, Gaël; Husson, Laurent
- Abstract
The use of Bayesian trans‐dimensional sampling in 2‐D and 3‐D imaging problems has recently become widespread in geophysical inversion. Its benefits include its spatial adaptability to the level of information present in the data and the ability to produce uncertainty estimates. The most used parameterization in Bayesian trans‐dimensional inversions is Voronoi cells. Here we introduce a general software, TransTessellate2D, that allows 2‐D trans‐dimensional inference with Voronoi cells and two alternative underlying parameterizations, Delaunay triangulation with linear interpolation and Clough‐Tocher interpolation, which utilize the same algorithm but result in either C0 or C1 continuity. We demonstrate that these alternatives are better suited to the recovery of smooth models, and show that the posterior probability solution is less susceptible to multimodalities which can complicate the interpretation of model parameter uncertainties. Key Points: We present a new software for trans‐dimensional surface reconstruction incorporating hierarchical error estimation, Hamiltonian Monte Carlo, and parallel temperingWe propose two alternative parameterizations to the ubiquitous Voronoi cellsThese alternate parameterizations may open up the application of trans‐dimensional inversion to a wider variety of geophysical problems
- Subjects
SURFACE reconstruction; PARAMETERIZATION; HAMILTON'S equations; CENTROIDAL Voronoi tessellations; MONTE Carlo method
- Publication
Geochemistry, Geophysics, Geosystems: G3, 2019, Vol 20, Issue 1, p505
- ISSN
1525-2027
- Publication type
Article
- DOI
10.1029/2018GC008022