Slow slip preceding earthquake rupture. Where to expect it and how large is it?


Deformation on major tectonic faults is accommodated by co-seismic slip (earthquakes), and aseismic slow slip (creep) during the earthquake cycle. Acceleration of creep preceding earthquakes is predicted by theoretical models and in laboratory experiments, and recently observed before natural large magnitude earthquakes (e.g. Tohoku 2011, Iquique 2014). However the magnitude and the prevalence of pre-slip remain unclear. The aim of this project is to place bounds on the magnitude of the expected pre-slip in earthquakes, based on realistic fault structures which are strongly heterogeneous. Uncovering the physics of pre-slip would allow a deeper understanding of the earthquake cycle; understanding its role in earthquake triggering would improve its integration in probabilistic forecasting and contribute to build resilience.

Click on an image to expand

Image Captions

Fig 1. Conceptual representation of a megathrust according to the current paradigm. Earthquakes usually nucleate in domains B or C, and for the largest ones propagate upwards into A, except for “Slow Tsunami” earthquakes which are limited to region A. Within domain D the only seismic activity appears to be tremor, which is believed to arise from the failure of small asperities within a stable slipping domain under close to velocity-neutral friction dependence. B: Schematic illustration of the experimental samples. The slip surface will be prepared with a bimodal distribution of patches: the orange background represent velocity-strengthening material which undergoes stable sliding (creeping patches), red patches represent higher-friction, velocity-weakening material (seismic asperities). A similar set of faults will be prepared with the opposite structure (red-strengthening and orange-weakening). (i) homogeneous fault; (ii) single asperity; (iii) array of similar asperities. The radius r of the single asperity and the minimum distance R between asperities will be varied to produce different dimensionless ratios r/R and r/h*. The value of r/R will relate to the relative density of asperities on the surface. (iv) patches of clustering asperities (toward a fractal distribution). The asperity radii r, the distance R between asperities within clusters, and the distance R” between clusters will result in two dimensionless parameters r/R and R/R’. A similar structure will be reproduced on the numerically simulated faults.

Fig 2. Experimental setup used to obtain pre-slip in the triaxial press. P is confining pressure and F axial load.

Fig 3. Instability and pre-slip recorded on an experimental fault. Loading phase (L), followed by stable sliding (S), growth of slip instability (dashed rectangle G), and final dynamic rupture (R). Three full cycles are represented. C Zoom of dynamic slip velocity V and the dynamic stress drop (red curve). Westerly granite under 75 MPa of normal stress.


The PhD will conduct carefully controlled and monitored laboratory experiments on pre-slip and fault failure at the cm scale, and numerical simulations at the hundred-km scale, using them jointly to develop and constrain new models of earthquake nucleation. Large earthquakes are frequent enough to cause substantial damage and life loss, but not frequent enough to provide a statistically-significant catalogue. Experiments can provide a large number of observations.

Project Timeline

Year 1

The activity in the first year of the PhD will be devoted to bibliographic research, laboratory training, adapting existing laboratory machines, design/realization of new mechanical parts, perfecting of sample preparation procedures, and preliminary tests.

Year 2

The second year will be devoted to experimental activity, field excursions to obtain natural fault samples, analysis of post-experimental microstructure and interpretation of experimental observations in terms of processes.

Year 3

Third year will be devoted in small part to further experimental activity, but mainly, to the interpretation, extrapolation of results to real earthquakes and writing of publications and PhD thesis.

Year 3.5

Finish writing thesis, dissemination (submit publications to journals, conferences)

& Skills

The student will be trained in different disciplinary fields as field structural geology, earthquake mechanics, laboratory mechanical experiments and fluid flow modelling. Importantly, he or she will learn to use high pressure rock deformation apparatuses and techniques which are widespread not only in the world of academic research but also that of technical expertise and industry. In addition, he or she will develop skills to undertake microscopic analysis of rock formations and the interpretation of the deformation microstructures.

References & further reading

Guerin-Marthe, S., Nielsen, S., Bird, R., Giani, S., & Di Toro, G. (2019). Earthquake nucleation size: Evidence of loading rate dependence in laboratory faults. Journal of Geophysical Research: Solid Earth, 124, 689-708.
Christopher W.A. Harbord, Stefan B. Nielsen, Nicola De Paola, Robert E. Holdsworth; Earthquake nucleation on rough faults. Geology ; 45 (10): 931-934. doi:

Further Information

Apply Now