Development and Tuning of a 3D Stochastic Inversion Methodology to the European Arctic

Abstract

The development of three-dimensional (3D) seismic models for the crust and upper mantle has traditionally focused on finding one model that provides the best fit to the data, while observing some regularization constraints. Such deterministic models, however, ignore a fundamental property of many inverse problems in geophysics: nonuniqueness. It is likely that if a model can be found to satisfy given datasets, an infinite number of alternative models will exist that satisfy the datasets equally well. Our solution to the inverse problem of developing a seismic model for the Barents Sea, given various datasets, is therefore a probabilistic model, a posterior distribution of models that satisfy the data to the same degree. We use a Markov Chain Monte Carlo algorithm to sample the unknown posterior distribution, which describes the ensemble of models that are in agreement with prior information and the datasets.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2010
Accession Number
ADA569446

Entities

People

  • Hilmar Bungum
  • Jan I. Faleide
  • Juerg Hauser
  • Kathleen M. Dyer
  • Michael E. Pasyanos
  • Stephen A. Clark

Tags

Communities of Interest

  • Energy and Power Technologies
  • Ground and Sea Platforms

DTIC Thesaurus Topics

  • Air Force Research Laboratories
  • Algorithms
  • Barents Sea
  • Bayesian Networks
  • Computational Science
  • Earth Sciences
  • Explosions
  • Ground Based
  • Markov Chains
  • Monte Carlo Method
  • Nuclear Explosions
  • Probabilistic Models
  • Probability
  • Standards
  • Surface Waves
  • Three Dimensional
  • Travel Time

Readers

  • Computational Modeling and Simulation
  • Neural Network Machine Learning.
  • Statistical inference.