Probabilistic surface reconstruction from multiple data sets: An example for the Australian Moho
Open Access
- 19 October 2012
- journal article
- seismology
- Published by American Geophysical Union (AGU) in Journal of Geophysical Research
- Vol. 117 (B10)
Abstract
Interpolation of spatial data is a widely used technique across the Earth sciences. For example, the thickness of the crust can be estimated by different active and passive seismic source surveys, and seismologists reconstruct the topography of the Moho by interpolating these different estimates. Although much research has been done on improving the quantity and quality of observations, the interpolation algorithms utilized often remain standard linear regression schemes, with three main weaknesses: (1) the level of structure in the surface, or smoothness, has to be predefined by the user; (2) different classes of measurements with varying and often poorly constrained uncertainties are used together, and hence it is difficult to give appropriate weight to different data types with standard algorithms; (3) there is typically no simple way to propagate uncertainties in the data to uncertainty in the estimated surface. Hence the situation can be expressed by Mackenzie (2004): “We use fantastic telescopes, the best physical models, and the best computers. The weak link in this chain is interpreting our data using 100 year old mathematics”. Here we use recent developments made in Bayesian statistics and apply them to the problem of surface reconstruction. We show how the reversible jump Markov chain Monte Carlo (rj‐McMC) algorithm can be used to let the degree of structure in the surface be directly determined by the data. The solution is described in probabilistic terms, allowing uncertainties to be fully accounted for. The method is illustrated with an application to Moho depth reconstruction in Australia.Keywords
This publication has 48 references indexed in Scilit:
- Inference of abrupt changes in noisy geochemical records using transdimensional changepoint modelsEarth and Planetary Science Letters, 2011
- Three‐dimensional Moho topography in Italy: New constraints from receiver functions and controlled source seismologyGeochemistry, Geophysics, Geosystems, 2011
- Climate driven glacial–interglacial variations in the osmium isotope composition of seawater recorded by planktic foraminiferaEarth and Planetary Science Letters, 2010
- A RJMCMC Algorithm for Object Processes in Image ProcessingMonte Carlo Methods and Applications, 2001
- Joint Bayesian model selection and estimation of noisy sinusoids via reversible jump MCMCIEEE Transactions on Signal Processing, 1999
- Marine gravity anomaly from Geosat and ERS 1 satellite altimetryJournal of Geophysical Research, 1997
- Reversible jump Markov chain Monte Carlo computation and Bayesian model determinationBiometrika, 1995
- Estimating the Dimension of a ModelThe Annals of Statistics, 1978
- A new look at the statistical model identificationIEEE Transactions on Automatic Control, 1974
- Nouvelles applications des paramètres continus à la théorie des formes quadratiques. Deuxième mémoire. Recherches sur les parallélloèdres primitifs.Journal für die reine und angewandte Mathematik (Crelles Journal), 1908