3‐D multi‐observable probabilistic inversion for the compositional and thermal structure of the lithosphere and upper mantle. II: General methodology and resolution analysis. Issue 4 (8th April 2013)

• Record Type: Journal Article
• Authors: Afonso, J.C.; Fullea, J.; Yang, Y.; Connolly, J.A.D.; Jones, A.G.
• Is Part Of: Journal of geophysical research. Volume 118:Issue 4(2013:Apr.)
• Page Start: 1650
• Page End: 1676
• Publication Date: 2013-04-08
• Subjects: Probabilistic Inversion -- Lithosphere -- Temperature and Composition -- Upper mantle; Geomagnetism -- Periodicals; Geochemistry -- Periodicals; Geophysics -- Periodicals; Earth sciences -- Periodicals; 551.1
• DOI: 10.1002/jgrb.50123
• Languages: English
• ISSNs: 2169-9313

Abstract

Abstract : [1] Here we present a 3‐D multi‐observable probabilistic inversion method, particularly designed for high‐resolution (regional) thermal and compositional mapping of the lithosphere and sub‐lithospheric upper mantle that circumvents the problems associated with traditional inversion methods. The key aspects of the method are as follows: (a) it exploits the increasing amount and quality of geophysical datasets; (b) it combines multiple geophysical observables (Rayleigh and Love dispersion curves, body‐wave tomography, magnetotelluric, geothermal, petrological, gravity, elevation, and geoid) with different sensitivities to deep/shallow, thermal/compositional anomalies into a single thermodynamic‐geophysical framework; (c) it uses a general probabilistic (Bayesian) formulation to appraise the data; (d) no initial model is needed; (e) compositional a priori information relies on robust statistical analyses of a large database of natural mantle samples; and (f) it provides a natural platform to estimate realistic uncertainties. In addition, the modular nature of the method/algorithm allows for incorporating or isolating specific forward operators according to available data. The strengths and limitations of the method are thoroughly explored with synthetic models. It is shown that the a posteriori probability density function (i.e., solution to the inverse problem) satisfactorily captures spatial variations in bulk composition and temperature with high resolution, as well as sharp discontinuities in these fields. Our results indicate that only temperature anomalies of Δ T ⪆ 150°C and large compositional anomalies of ΔMg# > 3 (or bulk ΔAl2 O3 > 1.5) can be expected to be resolved simultaneously when combining high‐quality geophysical data. This resolving power is sufficient to explore some long‐standing problems regarding the nature and evolution of the lithosphere (e.g., vertical stratification of cratonic mantle, compositional versus temperature signatures in seismic velocities, etc) and offers new opportunities for joint studies of the structure of the upper mantle with unprecedented resolution. Key Points: High‐resolution probabilistic inversion of multiple geophysical data in 3D Temperature and composition of the upper mantle Computational implementation of multi‐observable probabilistic inversions