From Mantle Convection to Seismic Observations – and Back?: The Impact of Tomographic Resolution and Mineralogical Uncertainty on Reconstructed Mantle Evolution

Add to your list(s) Download to your calendar using vCal

• Bernhard Schuberth (LMU Munich)
• Wednesday 04 March 2026, 14:00-15:00
• Wolfson Lecture Theatre.

If you have a question about this talk, please contact Alice Turner.

Understanding the structure and evolution of Earth’s mantle is fundamental for constraining plate-driving forces, lithospheric stresses, and the long-term behaviour of the geodynamo. While the present-day thermodynamic state of the mantle can be estimated from seismic tomography and high-pressure mineral physics, putting tight constraints on temperature and chemical heterogeneity based on seismic observations still remains a major challenge. Geodynamic simulations, by contrast, provide theoretical predictions for mantle evolution. However, their quality depends on how well input parameters are known, and they are only meaningful when rigorously tested through geodynamic–tomographic comparisons or, ideally, by comparing secondary predictions to a broad range of Earth observations. In this talk, I will discuss recent developments and potential future directions aimed at providing a quantitative, physically consistent link between temperatures predicted by mantle circulation models (MCMs) and the wealth of information contained in seismic recordings. For example, using global 3-D seismic wavefield simulations and full-coupling free-oscillation calculations for the MCM -derived structures, synthetic traveltime residuals and seismic spectra can be computed that accurately capture the various non-linearities in the relation to the underlying temperatures. A critical component in this context is the effect of mineral anelasticity, and it is particularly important to account for the associated uncertainties when comparing synthetic and real data. In addition to this forward approach, geodynamic adjoint inverse modelling can be used to retrodict mantle flow back in geologic time starting from the present-day state derived from tomographic images. Key challenges therein include determining the resolution and uncertainty of the tomography used and how they affect adjoint-state reconstructions, as well as assessing how accurately temperatures can be recovered using uncertain mineralogical information. Addressing the scale discrepancy between fluid dynamic predictions and seismically imaged structures is crucial, as validating reconstructed mantle flow involves surface topography calculations that are highly sensitive to the tomographic input. To improve future retrodictions, we conducted synthetic experiments illustrating the challenges of integrating tomographic and geodynamic models. Using a reference MCM as the “true” structure, we employed the linear SOLA Backus–Gilbert framework to explore spatially optimised averaging kernels and noise-related uncertainties, proposing a workflow to identify ideal SOLA parameters for next-generation adjoint models. Furthermore, synthetic closed-loop experiments demonstrate that tomographic damping, spatial blurring, and simplified mineralogies cause substantial deviations from “true” temperatures, especially near phase transitions. When such temperatures are used to prescribe buoyancy in geodynamic simulations, the errors will amplify non-linearly, potentially activating incorrect phase transitions and significantly altering reconstructed flow trajectories.

This talk is part of the Bullard Laboratories Wednesday Seminars series.

This talk is included in these lists:

• Bullard Laboratories Wednesday Seminars
• Department of Earth Sciences seminars
• Wolfson Lecture Theatre

Note that ex-directory lists are not shown.