Historical Geodynamics

By historical geodynamics we mean that we explicitly link time-dependent geodynamic models to the geological record. These approaches complement those of present day geodynamics in that we compare the models against different kinds of the data than traditional geophysically orientated geodynamic models. The complementary data sets are time-dependent plate motions, the record of stratigraphic deposits onshore and offshore, and the record of volcanism. The historical and present-day models each have with their strengths and weaknesses. On this page, we look at several studies that show the deversity of our work and the important role they are playing in pure and applied earth science. We will at some regional models of anomalies vertical motions. This is followed by explicit work on global sea level change over millions of years. Finally, we look at the initiation of a new subduction zone.

Much of the work linking geodynamic models with plate reconstructions has been completed by closely working with Dietmar Müller and his EarthByte Team at the University of Sydney. We've worked closely to develop GPlates, a open source Paleogeographic system.

Figure 1. Results for an inverse geodynamic model starting with scaled seismic velocity (vs) scaled to temperature and integrated backward to 100 Ma from Liu, Spasojevic & Gurnis. The vertical cross sections show the scaled temperature field and the horizontal surface maps show dynamic topography. The cold slab (blue in cross section), and associated negative dynamic topography (green and blue in map) are migrating from west to east. This image was created from our data by Bernhard Steinberger (then at the Norwegian Geological Survey) who created this visualiztion for a Perspective he wrote for Science Magazine

Regional Models and the North-America Cretaceous seaway

Inverse geodynamic models exploited the seismically resolved Farallon slab below the eastern-half of the United States and the well studied sedimentary section of the Cretaceous interior seaway that has long been recognized as forming in response to mantle flow. The global models employed an adjoint method, which backward reverses mantle convection, with the present-day mantle structure constrained by seismic tomography and the time-dependent evolution by plate motions and stratigraphic data. The application of the adjoint method to mantle convection was an important part of Lijun Liu's Ph.D. these at Caltech and he wrote up details of the methods in a JGR paper.

By synthesizing proxies for North American vertical motions from the late Cretaceous to the present (e.g. paleoshorelines and borehole tectonic subsidence curves), we reconstructed the geometry of the Farallon slab (Fig. 1). The dynamic topography associated with the subduction of the Farallon slab is localized in western North America over the late Cretaceous, representing the primary factor controlling widespread flooding. The dynamic model was later shown to be consistent with late Mesozoic uplift of the Colorado Plateau, and the migration of the depocenter of the hydrocarbon bearing, late Cretaceous succession across central Utah, Colorado, and southern Wyoming. While previous simplified geodynamic concepts did not predict a migrating depocenter, similar predictions are common in our dynamic models and provide an opportunity to understand such patterns commonly seen in the stratigraphic record.

Former geophysics graduate students Lijun Liu and Sonja Spasojevic collaborated on the application of this adjoint method to North American Cretaceous seaway. Read our People's Page to read about the current and former graduates of the geodynamics group.

The models of North America are examples of regional geodynamic models. Interestingly enough all of our regional models are within a global domain, but we have formulated the problems so as to avoid problems with "edge effects". Besides, application to North America, we have applied our models to a variety of other locations, Including Australia, New Zealand, South America and South East Asia. Here are just some examples, with some links to the relevant papers. There are many more examples found by going to Michael Gurnis's Publications page.

Sonja completed a study of the anomalous subsidence of the Campbell Plateau with Rupert Sutherland of GNS Science.

As a visiting student at Caltech, Lydia DiCaprio completed a study on the Cenozoic tilting and subsidence of Australia.

Sydney graduate student Grace Shephard applied the adjoint models to the reversal of the Amazon River.

North China Craton

With Shaofeng Liu, Professor at China University of Geoscieces in Beijing, we are working on linking the evolution of the North China Craton with geodynamic models. With constraints from structural geology, sedimentology, geochoronology, and geodynamic modeling, we are deciphering the relations between the subduction history of paleo-Pacific plate and the response of shallow basin and range above the North China Craton (NCC) during its destruction process. By combining the global tectonics, the oceanic and continental geomagnetic fields, regional seismic tomography, and the constraints from the shallow geological structure, we are constructing the 3-D subduction model of western paleo-Pacific plate since 200 Ma, including to simulate the subduction history and to study the deep mantle convection and evolution of shallow structures. Based on the above results, we shall be able to explore the geodynamic explanations between the destruction of North China Craton and the subduction of paleo-Pacific plate.

Global Sea Level

With the volume of ocean water assumed constant over millions of years, global sea level on geologic time scales results from a change of: ocean bathymetry due to variations in ocean floor age, dynamic topography in oceanic and flooded continental areas, emplacement of oceanic plateaus, and sedimentation.

Figure 2. Components of the hydrid dynamic earth models used to compute region and global sea level change. Both maps are at 80 Ma. Above: age of the oceanic lithosphere from the work of Maria Seton of the EarthByte group, vectors of plate motion and observed marine inundation onto the continents. Below: Sonja Spasojevic's predicting flooding and shorelines from the hydrid dynamic earth models.

As part of her Ph.D., Sonja Spasojevic generated dynamic earth models to better understand the origin of global and regional sea level change. Predicted global sea level by DEMs shows an overall decrease from the Late Cretaceous to the present, with a maximum at 80 Ma.

She used dynamic earth models are used to better understand the impact of mantle dynamics on vertical motion of continents and regional and global sea-level change since the Late Cretaceous. Her hybrid approach combined inverse and forward models of mantle convection, and accounted for the principle contributors to long-term sea-level change: the evolving distribution of ocean floor age, dynamic topography in oceanic and continental regions, and the geoid. She inferred relative importance of dynamic versus other factors of sea-level change, determined time-dependent patterns of dynamic subsidence and uplift of continents, and derived a sea-level curve. She found that both dynamic factors and the evolving distribution of sea floor age are important in controlling sea level. By tracking the movement of continents over large-scale dynamic topography by consistently mapping between mantle and plate frames of reference, she found that this movement resulted in dynamic subsidence and uplift of continents. The amplitude of dynamic topography in continental regions is larger than global sea level in several regions and periods, so that it controls regional sea level in North and South America and Australia since the Late Cretaceous, North Africa and Arabia since the Late Eocene, and Southeast Asia in the Oligocene-Miocene period. East and South Africa experience dynamic uplift over the last 20-30 million years, while Siberia and Australia experience Cenozoic tilting. The dominant factor controlling global sea level is a changing oceanic lithosphere production that results in a large amplitude sea-level fall since the Late Cretaceous, with dynamic topography offsetting this fall.

Initiation of Subduction

Subduction initiation, although only a transient phenomenon, is a vital phase of the plate tectonic cycle. Long-lived and well-developed subduction zones disappear and new subduction zones form. Indeed, nearly half of all presently active subduction zones initiated during the Cenozoic. This record provides a rich geological history that can be used to validate dynamic models. As such, we have an active program to try and better understand the dynamics of subduction initiation and link such models explicitly to the geological record. Over the years, Gurnis has worked on this problem with a number of students and post-docs and that work cointinues with Wei Leng.

Theoretical studies, and interpretation of the Mesozoic and later plate tectonic history of the Pacific, suggest that subduction initiation alters the force balance on plates. If we hope to make fundamental advances in understanding the forces driving and resisting plate motions, then a detailed picture of subduction initiation is needed. After a hiatus of a number of years, computational and synthesis models of subduction initiation are now being advanced.

The best examples where we know subduction started and has since evolved into fully self-sustaining subduction zones, include the Eocene initiation of the Izu-Bonin-Marianas (IBM) and Tonga-Kermadec. Earlier, we have formulated models with fault or shear zone development in a visco-elastoplastic material. We use a simple geometry to illustrate the tectonic precursors to the transition from forced to self-sustaining subduction. When the plate is compressed with a weak structure, there will be uplift on one side of the shear zone; as compression continues, a paired topographic feature results, with a high ridge adjacent to a trough. With further compression, the dip of the shear zone becomes shallower and the two sides slide past each other. The force required to compress the system can increase during this period as the underthrust plate bends. As more cold oceanic lithosphere is thrust below the overriding plate, the ridge subsides. Depending on the amount of convergence, the overriding plate will uplift and subside over a period of ~5-10 Myr. This reversal in vertical motion heralds the transition from forced to self-sustaining subduction.

Figure 3. A new study of subduction initiation was recently published by Wei Leng.

With Wei Leng, we have started to better understand the factors of subduction initiation. Changes of plate motion may have induced subduction initiation (SI), but the tec tonic history of SI is different from one subduction zone to another. Izu-Bonin-Mariana (IBM) SI, accompanied by strong backarc spreading and voluminous eruption of Boninites, contrasts with the Aleutians which shows neither. Using finite element models, Leng & Gurnis explored visco-elasto-plastic parameters and driving boundary conditions for SI evolution. With an imposed velocity, they found three different evolutionary modes of SI: continuous without backarc spreading, continuous with backarc spreading and a segmented mode. With an increase in the coefficient of friction and a decrease in the rate of plastic weakening, the amount of convergence needed for SI increases from ~20 to ~200 km, while the mode changes from segmented to continuous with backarc spreading and eventually to continuous without backarc spreading. If the imposed velocity boundary condition is replaced with an imposed stress, the amount of convergence needed for SI is reduced and backarc spreading does not occur. These geodynamic models provide a basis for understanding the divergent geological pathways of SI: First, IBM evolution is consistent with subduction of an old strong plate with imposed velocity which founders causing intense backarc spreading and Boninitic volcanism. Second, the New Hebrides SI is in the segmented mode due to its weak plate strength. Third, the Puysegur SI is in the continuous without backarc spreading mode with no associated volcanic activities. Fourth, the Aleutians SI has neither trench rollback nor backarc spreading because the slab is regulated by constant ridge push forces.