The Earth’s lithosphere is fragmented into plates that move rigidly and deform only within narrow boundary zones. My research focuses on strain localization in the brittle upper part of the lithosphere, i.e., the formation and evolution of major fault systems. I am particularly interested in the interactions between faulting and geological phenomena such as erosion, sedimentation, magmatism, and hydrothermal circulation. To better understand the physics that underlie these interactions on a range of time scales, I develop models that relate the morphology and structure of plate margins to the rheological properties of the lithosphere and the forces that act on it. While my primary focus is on extensional faulting at continental rifts and mid-ocean ridges, I am also involved in a number of projects relative to the long-term evolution of convergent plate boundaries.
Major types of extensional and convergent plate boundaries, where shallow strain localization manifesting as seismogenic faults interacts with a range of geological phenomena. These include magmatism, hydrothermal circulation, and surface processes.
Geological controls on extensional faulting at rifts and ridges
Extensional plate boundaries are characterized by a spectrum of tectonic styles from steep, short-offset normal faults to low-angle, large-offset detachments exhuming lower crust and mantle units. To what extent does this diversity reflect variability in the thermo-mechanical state of the lithosphere and in the geological forcings that may act on divergent margins?
How do surface processes modulate the localization of extensional strain?
In sub-aerial environments, surface processes constantly rework fault-induced topography, which in turn affects the stress state of the lithosphere and the long-term evolution of faults. In extensional settings in particular, alleviating the topographic load of a footwall crest and hanging wall basin by erosion and sedimentation can enhance the life span of a normal fault, and affect subsequent strain localization. By combining numerical modeling and tectonomorphic analyses of continental rifts, I seek to assess the relative effect of surface processes and lithospheric strength on the modes of extensional faulting.
Surface processes rework the topography induced by slip on the Teton fault (Wyoming).
Numerical simulation of extensional faulting strongly affected by surface processes, from Olive et al. [2014b].
Abyssal hills form when a normal fault uplifts newly accreted volcanic seafloor at the axis of a mid-ocean ridge, and make up the most common landform on the surface of the solid Earth. With colleagues at the University of Hawaii, WHOI, and IPGP, we combine numerical modeling and seafloor observations to untangle magmatic and tectonic contributions to the morphology of abyssal hills, and to better understand the mechanics of normal faulting in relation to crust emplacement. I am particularly involved in the study of oceanic detachments, which are large-offset normal faults that are thought to form when the magma supply of a ridge segment is locally reduced.
Processes linking mid-ocean ridge magma supply to seafloor morphology, from Olive et al. .
A simple model (gray shading) for the characteristic spacing of abyssal hills as a function of mid-ocean ridge spreading rate, from Olive et al. . Data from Goff et al. , and refs. therein.
Connecting seismic and geological time scales of deformation
Earthquakes are the primary manifestation of fault activity on human time scales. But how central are earthquakes to the build-up of permanent deformation and topography at extensional plate boundaries?
Does seafloor topography grow seismically or aseismically?
The measured rates of seismic moment release at mid-ocean ridges typically account for less than ~10–20% of the fault slip rates documented geologically. My research seeks to understand how the (long-term) thermo-mechanical state of the oceanic lithosphere might affect its behavior on seismic cycle time scales. To do so, I search for potential correlations between mid-ocean ridge seismicity, seafloor morphology, and tectonic styles along the axis of mid-ocean ridges. Mechanical models then provide a conceptual framework to explain these correlations. For example, a solution to the seismic "deficit" of mid-ocean ridges may pertain to the inhibited nucleation of frictional instabilities on normal faults that have a limited spatial extent.
Seismic energy released along the Northern Mid-Atlantic Ridge, where accretion occurs either along symmetric (red) or asymmetric (detachment fault-bearing, blue) sections.
Related publication: Olive and Escartín .
How do seismic cycles affect the evolution of submarine landscapes?
In the absence of geomorphic drivers such as rivers or glaciers, submarine extensional terrains are mainly reworked by rockslides that are likely influenced by the occurrence of large earthquakes. I am currently investigating potential relations between mass wasting and seismicity along submarine normal faults of the Chile Ridge, the Mid-Atlantic Ridge, and the Lesser Antilles arc.
Heat and mass transfers associated with hydrothermal systems
Where there are faults, there are (generally) fluids. At mid-ocean ridges, fluid circulation in the crust is a primary means of heat extraction, which feeds back onto the thermo-mechanical state of the lithosphere. Hydrothermal convection is also central to many geochemical cycles. My work aims at refining estimates of hydrothermal fluxes and identifying the geological parameters that control these fluxes.
Quantifying thermo-chemical fluxes between the lithosphere and ocean
Hydrothermal vents are the most spectacular manifestation of fluid circulation through the oceanic crust. Along with diffuse venting, focused hydrothermal plumes transport heat and chemicals from the lithosphere into the deep ocean. Over the past few years, I have developed a simple methodology to quantify the flow rate of individual plumes using only a few measurements of temperature and passive tracer concentrations within a few meters above the vent. My colleagues and I are now using this technique to infer the thermo-chemical anatomy of a hydrothermal plume, and develop detailed maps of heat output at selected hydrothermal sites.
The thermo-chemical output of hydrothermal systems is primarily modulated by the permeability of the oceanic crust and the availability of a heat source to drive and sustain hydrothermal convection. In order to better quantify these controls, Thibaut Barreyre (WHOI) and I are developing novel methods to measure the depth-dependent permeability of the crust beneath hydrothermal discharge zones. With Tim Crone (LDEO), I investigate the dynamics of the boundary layer that defines the base of the hydrothermal convective system. Finally, with Eric Mittelstaedt (U. Idaho), we combine numerical and analog modeling to understand the self-organization of fluid pathways in heavily faulted oceanic crust, where sharp permeability gradients are the norm.
Many of the physical concepts developed in my work on extensional plate boundaries apply to the dynamics of subduction zones and orogenic wedges. I am currently involved in a number of projects on specific aspects of strain localization and mantle flow at convergent plate boundaries:
• Dynamics of orogenic wedges (Led by J. Weiss and G. Ito, U. Hawaii) Goal: understanding the conditions that lead to the localization of the frontal thrust of an orogenic or accretionary wedge.
• Mechanics of outer rise faulting (Led by J. Lin, WHOI) Goal: identifying the controls on the occurrence and pervasiveness of normal faults in the outer-rise region of a subducting plate. Related publication: Zhou et al. .
• The role of compositional buoyancy on slab behavior (Led by B. Klein and O. Jagoutz, MIT) Goal: deciphering the role of thermal and chemical buoyancy in the dynamics of a sinking slab in the present and early Earth (and the consequences for layered mantle convection).
• Mantle flow around a subducting slab (in collaboration with S. Rondenay, U. Bergen) Goal: extracting quantitative information on mantle flow around the retreating Hellenic slab by mapping sharp spatial transitions in seismic anisotropy over the Western Hellenic subduction zone. Related publication: Olive et al. [2014a].
Numerical geodynamic modeling & inverse problems
Numerical modeling of long-term lithosphere and mantle dynamics poses numerous challenges. A geodynamic modeling code typically handles very large, very localized deformation, non-linear rheologies with sharp contrasts, an evolving free surface... and must be able to do so in two or three-dimensions in a relatively fast manner.
Developing and distributing numerical tools for geodynamics
Along with colleagues at WHOI, MIT, and the Universities of Hawaii and Idaho, I have developed a simple Matlab code for modeling 2-D long-term deformation of the lithosphere. This code, aptly named SiStER, for Simple Stokes solver with Exotic Rheologies, is based on the finite-difference / marker-in-cell technique described by Gerya (2010), and largely builds on Matlab's capabilities for rapid vector and sparse matrix operations. SiStER is now widely distributed on Github, and can be used as both a research and a teaching tool. We are currently working on the implementation and optimization of 3-D numerical codes for a range of geodynamic applications.
Another important aspect of computational geodynamics consists of making sure that the basic assumptions of our models are valid on a fundamental level. I specifically work on the importance of elasticity in long-term lithosphere evolution models, and seek to develop novel techniques to implement the elasto-plastic rheology into viscous flow solvers.
I am generally interested in inverse problem theory and its applications in the geosciences. I the last few years, I have worked with Mike Krawczynski on developing a code that performs mass balance calculations and uncertainty assessment on compositional datasets from the realm of experimental petrology. This code, named LIME, for Logratio Inversion of Mixed Endmembers, estimates the proportions of the various minerals that make up a rock of known bulk composition. Its main advantage on standard linear regression techniques is that it provides error bounds on the results and never gives a phase proportion that is either negative or greater than 100%.