The ANU, RSES, Seismology & Mathematical Geophysics, Research

 



Seismic structure of the lithosphere and upper mantle
Kennett, Rawlinson, Tkalcic, Pozgay, Salmon, Arroucau

The earthquake belt through Indonesia, New Guinea, Fiji, Tonga and New Zealand provides an excellent set of seismic sources for studying the mantle under Australia, and is occasionally complemented by events from the south on the Australian-Antarctic ridge. In consequence Australia is well placed for a wide variety of studies on the nature and three-dimensional structure of the lithosphere and upper mantle.

Portable array experiments have been carried out since the mid 70s, particularly in northern Australia. In the earlier short-period studies the emphasis was on structure in the upper mantle and transition zone. Since 1992 most experiments have used broad-band seismometers and have been used for a range of P and S wave studies.

From 1993-1996 a reconnaissance survey of the whole continent was carried out in the SKIPPY experiment. A set of 5-6 month deployments of typically 10 instruments were made, with an interstation spacing of approximately 400 km. Since 1997 a number of more focussed experiments have built on the SKIPPY results so that now more than 120 broad-band sites have been occupied, this provides a very significant data set.

The broad band studies have provided a detailed model of the three-dimensional structure for shear waves across the whole continent and the neighbouring oceanic regions and this model continues to be refined using new observations and inversion techniques. The emphasis has been on surface waveform tomography using Rayleigh waves, but some work has been carried out using Love waves that indicates significant variations in polarisation anomalies with depth. Body wave studies provide an independent check on the surface wave results and also allow delineation of attenuation structure for the northern part of the continent, with a significant attenuation beneath the very low-loss lithosphere.

Receiver function studies have been carried out for our past broadband deployments and are being exploited as our new stations continue to emerge, to enhance knowledge of lithospheric structure for the continent. We are currently combining receiver functions, surface wave velocity dispersion data, P-wave polarization data, as well as information from microseismic noise to improve the general understanding of the lithospheric structure of Australia, but also of various regions of the world. Joint modeling of these various types of data by means of various methods is used to place better constraints on lithospheric structure, including features such as the crustal thickness, upper mantle low velocity zone and transverse isotropy (polarization anisotropy). The improved structural models will be used to analyze regional moment tensors and to better understand the seismicity and hazard assessment for the region. We recently developed an interactive forward receiver function modeller (IRFFM), which can be used either as a stand alone tool or as a useful complement to inversion. The package with a manual can be freely downloaded here.

Present experiments are directed at understanding major contrasts in continental structure particularly the transition from the ancient core in western and central Australia with high wavespeeds in the mantle lithosphere to the slower wavespeeds beneath the Phanerozoic belts in the east. In the TIGGER experiment in Tasmania in 2002, a large array of short-period instruments was deployed along with broad-band instruments to elucidate the crustal and upper mantle structure. The Tasman Line experiment from 2003-2005 deployed 20 broadband instruments to bracket the transition from the craton to the younger fold belts. The object is to obtain as high a resolution as possible of the character of the transition as a function of depth. Field work in 2006/2007 (CAPRA) concentrated in northwestern Australia and focussed on the Archean Pilbara Craton and its borders. SOC (Structure of cratons) (2007/2008) concentrated on imaging crustal structure beneath southern Australia, while BILBY (2008/2009) is focused on seismic investigations of lithospheric transitions between the northern and southern Australian cratons. At the same time, we are continuing with an unprecedented coverage of the southeastern part of Australia (including Tasmania) by short period deployments. Notably, average spacing between our stations is about 40 km in the mainland and about 20 km in Tasmania. Recent deployments include SEAL, EVA, SETA, SEAL2, SEAL3 and SEAL4.

The studies of the Australian continent are complemented by a deployment of 7 broadband instruments in the Australian Antarctic Territory in the SSCUA experiment. Australia and Antarctica were joined in East Gondwanaland, and an object of this work is to try to relate structures in the two continents.


Geophysical parameterization and interpolation of irregular data using natural neighbours
Sambridge

We, in collaboration with Jean Braun and Herbert McQueen of the geodynamics group, have developed an approach for interpolating a property of the Earth (e.g. temperature, or seismic velocity) specified at a series of arbitrarily distributed `reference' points in two or three dimensions. The method makes use of some powerful algorithms from the field of computational geometry and is applicable to a wide range of problems including numerical modelling and building complex parameterizations in 3-D. Read more ...


Neighbourhood Algorithm
Sambridge, Bodin

The Neighbourhood algorithm, a general Monte Carlo method for non-linear inversion has been developed at RSES. The web page contains a description of the method and its application to the inversion of receiver functions for crustal seismic structure. Also access is provided to postscript versions of related papers, related websites, and information on how to obtain the software. See applications for seismic event locations and seismic source inversion. Read more ...


Inversion for Regionalised Upper Mantle models: - The RUM model
Sambridge

Oli Gudmundsson and Malcolm Sambridge have performed a series of inversions of global travel time data for a regionalised seismic model (RUM) of the upper mantle. This work also provides geometrical descriptions of the major subduction zones. Read more ...


Seismic Tomography
Rawlinson, Sambridge, Kennett, Tkalcic, Arroucau

The development of seismic velocity models for the Earth's interior depends on our ability to invert different classes of seismic data and current studies exploit both seismic travel times and long-period seismic waveforms for regional studies.

Studies using travel times exploit the joint use of P and S wave data to obtain images on both global and regional scales. An effective parameterisation is in terms of bulk-sound speed and shear wavespeed to isolate the influence of the bulk and shear modulus. Joint global inversions using 2x2 degree cells have demonstrated the dominance of S structure in the mid mantle, so that the narrow slab-like features in P wave images are controlled by S structure. At a regional scale variations in the balance of the bulk-sound and shear wavespeed have a strong correlation with the age of subducting lithosphere. Lithosphere older than around 85 Ma normally displays stronger shear anomalies.

A research focus is on the improvement of methods for extracting 3-D seismic shear structure using the records of fundamental and higher mode surface waves. The path-by-path inversions have been improved by using multiple starting models so that better estimates can be made for model error. A three-stage approach to 3-D model construction via the intermediary of multi-mode phase speed maps as a function of frequency provides the opportunity to include both ray-tracing corrections and the effects of finite frequency propagation.

Experiments are underway in southeastern Australia using dense deployments of short-period instruments with the objective of mapping uppermost mantle structure using delay-time tomography. From a study using 70+ instruments in Tasmania a new approach to arrival time-picking using adaptive stacking and tomography using fats-marching methods for travel time calculation has been developed which is both flexible and efficient.

The group is involved in travel-time tomography at both regional and global scales for both P and S waves particularly in joint inversions for bulk-sound speed and shear wavespeed for the whole mantle. The web pages currently provide access to images. Models can be made available on request to Brian Kennett. The group is also involved in mapping deep Earth structure, in particular the core-mantle boundary region using PKP, PcP and other body waves sensitive to the core structure. This web page provides access to images and compressional models of the D'' region. Newly installed short period and broadband deployments in Australia provide excellent coverage and promise further improvements in imaging Earth's deep interior.

On a regional scale there is an major effort to map 3-D shear wavespeed structure beneath the Australian region using waveform tomography for surface waves. A summary of results from the Skippy experiment can be found by clicking here. Current work focusses on inversion for both shear wavespeed and azimuthal and polarisation anisotropy. Read more ...


Structure of the Earth's core and lowermost mantle
Tkalcic, Pozgay

Ongoing research on the Earth's core and lowermost mantle investigates topics such as anisotropy and structure of the inner core, structure of the outer core, and the characterization of the core-mantle and inner-core boundaries. Inadequate spatial sampling of the central inner core by PKP waves in all directions makes further advances in understanding anisotropic properties (especially anisotropy's radial dependence and hemispherical pattern) very difficult. One of the reasons for this incomplete sampling lies in the fact that, in order to pass through the central regions of the inner core, PKP waves must be nearly antipodal. With the spatial distribution of large earthquakes and current configuration of the seismographic stations worldwide, this is difficult to achieve, except for the paths nearly parallel to the equatorial plane. We are, however, working toward improving the spatial sampling of the core and the lowermost mantle by exploiting other core phases, such as PKPPKP waves. There are three major approaches that could be pursued to achieve a better spatial samling of the deep Earth:
1) Observation and analysis of seismic phases with more complex geometry, such as PKPPKP or PnKP, which must be employed as a necessary supplement to PKP measurements, when using seismic travel times to study deep Earth structure (because of their unique sampling of the core that cannot be achieved by PKP waves only);
2) Installation of seismic stations at extreme geographic latitudes and ocean islands, in order to increase the coverage of the inner core by polar paths. The Australian deployments of the seismographic stations across Australia and in Antarctica greatly help in achieving this objective;
3) Development and application of new techniques (e.g. array signal processing techniques, or travel time measurements by non-linear inversion instead of direct time picking), which will allow us to use PKP and other data that were previously discarded.


Seismic Wave Propagation
Kennett

Current research is directed towards understanding of the influence of three-dimensional structures on the seismic wavefield on both small scales through stochastic methods and larger scales through direct numerical simulation.

Limited numerical simulation of wave propagation through the Australian region using the Spectral Element method have been made as a means of testing analysis algorithms and waveform inversion methods. Such 3-D numerical studies are expected to be developed further.


Body wave synthetics
Surface waves in 3-D media


Mathematical Geophysics
Sambridge, Rawlinson, Kennett

Research in mathematical geophysics is primarily focused on developing new methodologies for mathematical and data analysis problems in seismology, geodynamics and geophysical inverse theory. The ability to extract reliable information on Earth structure from seismic data, e.g. travel times or digital waveforms, or other classes of data, depends on methods of inverse theory. Fully non-linear (stochastic) inversion methods can provide valuable insight into the character of the solution. Genetic algorithms and variants are of particular interest. A new method of data space exploration termed the Neighbourhood Algorithm (NA) has been developed and has had a number of applications in different areas of seismology. A number of links have arisen with other groups in the School in the development of new data analysis tools; one example is in separating mixtures of different age components in geochronology.

The study of mathematical aspects of data analysis supports the observation work of the group. Strong links also exist with other areas in the school, notably with the geodynamics group where the common interest is in the development of new computational techniques for geodynamic modelling, which involves the solution of partial differential equations using finite difference and finite element techniques in two and three dimensions. Significant progress has been made in the application of these techniques to unstructured numerical grids, an approach which also has potential for studies of seismic wave propagation.


Seismic sources
Tkalcic

Current research is directed towards observing and studying seismic sources in Australia and on global scale, since moderate size or large earthquakes are rare in Australia. We have been using regional moment tensor inversion for recent significant Australian earthquakes, although the lack of knowledge of crustal and upper mantle structure in regions without historical seismicity has hindered the progress so far. We are using receiver functions in conjunction with teleseismic waves and other established techniques to better understand structural effects and calibrate Green's functions for the moment tensor inversion. We have also been working on improving continental-wide Green's functions for reliable source parameter estimation of earthquakes surrounding Australia. This is a work in progress. Apart from that, we are interested in anomalous seismic radiation and puzzling focal mechanisms, such as those occurring in volcanic environment. We have been using a full moment tensor representation and different computational methods to reveal statistically significant non-double-couple components and model complex finite sources. Australia and its surroundings is an exciting environment for such studies, because of a great variety of physical mechanisms responsible for earthquakes.