skip to primary navigationskip to content

E322: High resolution imaging on the core-mantle boundary (Lead Supervisor: Sanne Cottaar, Earth Sciences)

Supervisors: Sanne Cottaar (Earth Sciences), Andrew Curtis (University of Edinburgh), and David Al-Attar (Earth Sciences)

Importance of the area of research:

Over the past decades, strong heterogeneity has been discovered right above the core-mantle boundary. Studies have identified the scattered occurrence of patches of extremely slow velocities. These patches are suggested to be tens of kilometres thick and hundreds of kilometres across and have been named Ultra-Low Velocity Zones (ULVZs). Potentially these structures are high density roots to broad whole-mantle upwellings that cause (mid-plate) volcanism at the surface.

ULVZs have never before been mapped in detail through inversion techniques. They are invisible to current global tomographic techniques as these models are limited to include long period waveform data and the ULVZs are too thin. Constraining the extent, morphology and velocity of the ULVZs will further our understanding of their density, composition, and their role in global mantle dynamics.

Project summary:

The goal of this project is to provide high resolution mapping of anomalous structures  on the core-mantle boundary using core-diffracted seismic phases and non-linear Monte Carlo inversion techniques. Shorter period core-diffracted shear waves are highly affected by these structures (aka ULVZs), with post-cursory waveforms appearing over 30 seconds delayed to the predicted arrival time.  We will use these core-diffracted phases in a non-linear Monte Carlo inversion technique, conducting the first high resolution mapping on the core-mantle boundary. Initially, we will apply this technique to the base of a whole-mantle upwelling beneath the Hawaiian hotspot.

What the student will do:

The student will collect a data set of shear diffracted waves, starting with the ULVZ at the base of the mantle beneath Hawaii. The travel-times of the observed direct and postcursory diffracted phases provide the constraints to a reversible-jump Markov chain Monte Carlo  (rj-McMC) inversion. This technique will sample a large family of suitable models, which will also highlights the eventual trade-offs and uncertainties at the edges of the ULVZ structure. The student will test samples and means of this family of models by forward modelling full waveforms and comparing these to the observations. Andrew Curtis has experience of using the rj-McMC method for lithospheric tomography and will actively help the student to implement and apply the new lowermost-mantle code.

After applying this to Hawaii the student will scout other areas of interest on the core-mantle boundary that have suitable earthquake- station geometries for diffracted phases (e.g. beneath Iceland).

Please contact the lead supervisor directly for further information relating to what the successful applicant will be expected to do, training to be provided, and any specific educational background requirements.


Cottaar, S. & Romanowicz, B. 2012. An unusually large ULVZ at the base of the mantle near Hawaii. Earth Planetary Science Letters, 355, pp.213–222.

Galetti, E., Curtis, A., Meles, G. A.,  Baptie, B. 2015. Uncertainty Loops in Travel-Time Tomography from Nonlinear Wave Physics. Physical Review Letters, 114(14), p.148501. Available at:

Sambridge, Malcolm, et al. "Transdimensional inference in the geosciences." Phil. Trans. R. Soc. A 371.1984 (2013): 20110547.

Follow this link to find out about applying for this project

Other projects involving the supervisor(s) listed can be viewed here:

Filed under: