Transdimensional inversion with more advanced {2D} parameterizations

@inproceedings{Herrero2025RM,
 abstract = {This paper introduces additional geometric parameterizations designed to improve the representation of layered models and make them more consistent with geological observations of the subsurface. These parameterizations are specifically developed to be compatible with transdimensional RJMCMC inversion, allowing the number and geometry of layers to be inferred adaptively from the data. Three parameterizations are presented. The first enables the representation of anticlinal, synclinal, and monoclinal structures using sinusoidal functions to generate a series of 2D points defining geological interfaces. The second incorporates geological faults into the existing parameterizations, introducing displacements across stratigraphic units. The third extends the dipping-angle-based approach of Visser et al. (2019) by allowing interfaces to intersect, forming sedimentary wedges. For each approach, a synthetic transdimensional inversion test is conducted using well log permeability and flow data. The results demonstrate that a large range of geometries can be inferred, even when the number of layers is unknown. While these developments remain preliminary, they pave the way for transdimensional inversion applied to realistic two-dimensional stratigraphic models, with potential extensions to threedimensional applications in the future.},
 author = {Herrero, Julien and Caumon, Guillaume and Bodin, Thomas and Aidahi, Mohammed},
 booktitle = {2025 {RING} meeting},
 language = {en},
 pages = {209--231},
 publisher = {ASGA},
 title = {Transdimensional inversion with more advanced {2D} parameterizations},
 year = {2025}
}