ASEG Extended Abstracts - 25th International Conference and Exhibition – Interpreting the Past, Discovering the Future, 2016

New geophysical and geological results shed light on the tectonic history of the Montresor belt, located on the Rae craton of northern Canada - an Archean terrane that has been reworked by four Proterozoic orogenies. In this contribution we use forward modelling of high-resolution aeromagnetic data to explore the 3D geometry and structural history of the Montresor belt, part of the Rae cover sequence. Previously thought to be a simple syncline, our re-analysis of the aeromagnetic data has outlined a set of earlier structures that provide new insight on the deformation history of the belt. Five cross-sections model discrete magnetic-lithologic units truncated by a series of low-angle faults. Reconstruction of the magnetic map features and forward models reveals a pre-fold geometry analogous to foreland fold and thrust belts, produced by D1 deformation during the Trans-Hudson orogeny, bracketed by available geochronology between 1.94 and 1.864 Ga. The Montresor belt rocks have potential for a variety of mineral deposit types, including precious metals in hydrothermal settings, and are under study as part of the Geo-mapping for Energy and Minerals program in Canada.

Nowadays magnetics, electromagnetics and gravity are among the most abundant airborne surveys. Traditionally they aim at specific depth targets. For instance, Airborne Electromagnetic (AEM) data are known to provide reliable models of a few hundred meters deep; whereas, gravity and magnetic data can reveal geological features below few thousand meters depth. This depth-resolution difference has historically limited the combined interpretation of these data. We, however, hypothesize that there is a commonly sensed depth interval, which could be used to harness the joint inversion of the data and increase the reliability of the models in the wider depth extent. To demonstrate this we designed three inversion experiments using potential and AEM field data acquired in Western Australia. Firstly, we inverted each data set separately using a conventional 2D inversion strategy. Secondly, we jointly inverted the gravity and magnetic data using the cross-gradient constraint. Thirdly, we added a preliminary AEM resistivity model as a cross-gradient constraint for the 2D cross-gradient joint inversion of the gravity and magnetic datasets. Our results show that the three data sets sense a common area of the subsurface and that the vertical resolution of each data set influences in the shallow and deep structures of the joint models.

Estimation of S-wave velocity from P-wave velocity has been widely used in geomechanical applications as both velocity measurements are not always available. In present study, some of previously suggested equations for estimation of shear velocity have been presented. These equations have local validity and new calibration of constants is needed if used in new region. On the other hand, theoretical background in geomechanics shows linear relationship between the squares of these velocities. Measured data points were tested in this linear equation and it was found that good matching is observed only in clean sandstone samples. The idea that alteration in linearity is mainly due to porosity effect, motivated the authors to make a new regression considering porosity as one of input parameters. The results highly confirmed this method and new correlations with previously published data showed correlation coefficient nearly equal to unity.

Airborne gravity tensor (or gravity gradient) surveying is one of the newest techniques for geophysical exploration. The rise in the acquisition of this data type is partially due to the fact that these data provide much more complete, extensive and higher resolution information of rock density distribution than conventional land gravity data. This has positioned Airborne Gravity Gradient (AGG) surveys among the geophysical services requested by mining companies, alongside aeromagnetic and radiometric surveys. The use of these data has been primarily to support geological mapping given the limited access to commercial software for quantitative depth interpretation. In particular, there has been a scarce development of AGG data inversion software, which has limited our understanding of the significance of these data for resolving three-dimensional subsurface targets. In this work we hypothesize that AGG data can provide more detailed information of the multidirectional variations of subsurface density. To prove this hypothesis, we developed conjugate gradient AGG-data inversion software for three-dimensional targets. This software was applied to synthetic data generated by several test assemblages of three-dimensional bodies and used to perform a Singular Value Decomposition (SVD) sensitive analysis to explore the actual resolution power of the different tensor data components and whether they are indeed superior to the conventional vertical gravity.

Big Data techniques have the potential to be paradigm-changing for applied geoscience if they are used widely. A significant number of such techniques, under the umbrella of Earth informatics, involve Machine Learning applied to high dimensional data to create new forms of value. This contribution presents two case studies of successful Earth informatics computation and the communication of the value of results, which provide insight into the uptake of ‘Big Data’ in geosciences.

Machine Learning techniques split naturally into either supervised or unsupervised approaches. Supervised algorithms, such as Random Forests^™ (RF), support vector machines or neural networks, share the concept of training a classifier using an initial (training) dataset. They are generally applied to predictive tasks, such as our first case study, predicting lithology from remote sensing and airborne geophysical data. Unsupervised algorithms, such as Self-Organising Maps (SOM), allow patterns inherent in the data to emerge without the use of a training dataset. They are generally applied to tasks which seek to explore patterns in data, such as our second case study, which identifies new potentially prospective river catchments. We find that calculating and presenting explicitly the newly extracted value, of the result obtained through computation, is an essential component of the post-compute evaluation.

As strong advocates for the use of a range of Big Data techniques in applied geosciences, we conclude that the benefits to be gained from the way that we ‘compute’ can be lost if we do not also take considerable care with the ways that we ‘communicate’.

Measurement bandwidth is an important feature of a geophysical system. Bandwidth allows detection of very resistive features (such as some kimberlites) and very conductive targets (like massive sulphides). In electromagnetic systems, bandwidth is not simply the sample rate of the data acquisition system or the earliest time channel, but also depends on transmitter spectrum, distance to target and processing. Optimising a system to detect a feature or measure a specific signal requires design considerations and trade-offs for different targets. The choice of excitation waveform in electromagnetic systems is one such trade-off. A square-pulse allows high-frequency energy to be excited, but, because of electronic limitations, has only limited dipole moment. A half-sine waveform efficiently generates energy at the base frequency and first few odd harmonics (low-frequency energy) and less high-frequency energy. Here, we describe the Multipulse configuration, an option on the Helitem system which employs both a half-sine and a trapezoid waveform to efficiently generate high- and low-frequency energy. Using survey data, we show the resolution power of the combined system compared to a single waveform. The combined data is better able to resolve near-surface features and deep structure than data from either waveform alone.

Consider an oil and gas exploration team that has been successfully using seismic attributes such as gradient & intercept or fluid & lithology indicators. Why might this team want to switch from those tried-and-true seismic attributes to seismic inversion? And which type of seismic inversion should they use? We now have access to azimuthal inversion, facies inversion, Bayesian probabilistic inversion and full waveform acoustic and (emerging) full waveform elastic inversion. The evolution and pace of seismic inversion technologies is exciting but present some difficult choices to the uninitiated interpreter. This talk aims to give seismic interpreters an overview of these ‘new’ technologies and how they compare. Topics covered will include: the advantages and challenges of azimuthal AVO inversion in Australia’s high stress environment, Bayesian (probabilistic) inversion and how inverting for facies improves it, synergies and challenges of combining new broad-band / long cable seismic acquisition with AVO inversion, and full waveform seismic inversion - can it replace AVO inversion? A minimum of equations will be shown - and only then to as a guide to illustrate what the different inversion techniques are doing.

A seismic attribute study was conducted over a subset of data located in the onshore area of the Otway Basin, Victoria. Attributes were developed from the original legacy processed seismic data including energy and frequency attributes. With investigation and analysis it was possible to identify an amplitude anomaly with multiple stacked attribute anomalies directly related to reservoir levels with observed hydrocarbon shows from legacy well data close to the structure. This study combined with a new independent evaluation of the exploration potential revealing thickening and extension of an excellent reservoir and indications of trapped hydrocarbons that have previously gone unnoticed on data acquired in the 1980’s.

Evaluation of the well data identified a thick sequence of stacked sandstones representing a facies change of the lower Eumeralla Fm. Identification of anomalous energy and frequency and combined with a neural network chimney model that demonstrates a gas cloud overlying the reservoir interval; presents a clear target for further evaluation.

Seismic mapping identified potential closure from a coarse 2D seismic grid. The nearby Greenslopes-1 well, located in a highly structured zone, represents high risk leak potential from migrated hydrocarbons and was drilled in a zone of low seal potential. This well however, provides an excellent test of the good quality reservoir potential at the level of the interpreted anomalous energy and frequency attributes.

Deterministic volumetrics of the prospect assuming a 70% recovery factor estimate recoverable gas volumes of 108 BCF. If a 10 barrel/MMCF ratio is assumed for condensate then the estimated recoverable condensate volume is estimated at 1.08 MMBBL. The overlying gas cloud potentially indicates significant liquids enrichment and it is likely the reservoir contains a fully saturated hydrocarbon charge.

Due to the current moratorium for onshore oil and gas drilling in Victoria the target will undergo further evaluation to assess the oil potential.

Rapid interpretation of transient electromagnetic (TEM) data sets is highly desirable for timely decision-making in exploration. However, full solution 3D inversion of TEM data sets is often unpractically slow. Therefore, a fast approximate 3D TEM inversion scheme has been developed for time-integrated (resistive limit) data. The resistive limits are amenable to linear 3D magnetic inversion, which is up to 100 times faster than “rigorous” 3D TEM inversion. The resistive limit inversion scheme is suitable for airborne, ground, and downhole TEM, both dB/dt and B-field. Its efficacy is illustrated here via application to a heli-borne sub-audio magnetic (HeliSAM) data set recorded over the Lalor Zn-Cu-Au VMS deposit in Manitoba, Canada. The response from the deposit is clear in the “total field” EM (TFEM) data even though the mineralisation is very deep, extending from depth 575m to over 1100m. A three-stage inversion of resistive limits derived from the TFEM rapidly defined a 3D conductor below the uppermost pyrite-sphalerite lenses, enclosing a volume containing mainly pyrrhotite-chalcopyrite stringer sulphides. Total inversion time was less than one minute on a notebook PC.

Velocity analysis is often necessary in pre-stack seismic processing to produce a good estimate of subsurface velocities. It requires the picking of moveout velocities on the semblance spectra. The semblance spectra is hampered with noise and lack of resolution around the peak representing the best move out velocity approximation. In this paper we introduce a new semblance scheme to reduce spectral noise and increase resolution in the semblance domain. The new scheme is based on a simple amalgamation of previously developed semblance enhancement methods. These methods are; the local-similarity weighted semblance, velocity-sensitivity weighted semblance and boot-strapped differential semblance. Velocity sensitivity semblance weights all traces based on sensitivity in the semblance spectra to changes in velocity, the local similarity weighting accounts for correlation between the stacked gather trace and all traces within the gather on a temporally localised scale and the boot-strapped differential semblance scheme weights the semblance spectra based on enhancing sensitivity to lateral differences in amplitudes. We test the proposed scheme on synthetic and real trace gathers. The test results show that our approach significantly improves the resolution and noise attenuation in the semblance spectra albeit to other semblance schemes.

There are many potential field inversion algorithms available, and all are sufficiently capable of generating a model that explains supplied geophysical observations. The challenge is extracting a model that provides real geological insight. Here we present applications of two different styles of advanced inversions to a deep exploration program at the Platreef PGE-Ni-Cu deposit in the Bushveld Igneous Complex of South Africa. The initial approach was to apply generalised focussing constraints to a 3D magnetic vector inversion, an approach chosen to manage the effect of expected strong remanent magnetisation. This resulted in successful prediction and drill definition of inferred resources within a deep, west-dipping extension to the shallow-dipping “Flatreef” deposit. Later, a detailed 3D model of geological constraints based on drilling and mapping was constructed and used to tightly constrain inversions of gravity data derived from a FALCON airborne gravity gradiometer survey. The resulting 3D density model accurately predicted a continuation of the Flatreef host rocks to shallower levels than previously anticipated. This facilitated further drilldefinition of additional inferred resources within a southern extension of the Flatreef deposit. Key to the success of the inversions at accurately targeting mineralisation at depths of 700-1300 m depth, was the inclusion and integration of all available information to ensure that predictions were consistent with prior observations.

Marine towed streamer acquisition has been largely restricted to just a few geometries which are determined by the streamer separation and number of sources. Most acquisition uses two sources with 100m streamer separations, or sometimes 75m and 50m for higher resolution surveys. Greater flexibility can be gained by using more than two sources and this leads to a range of design options more commonly associated with land acquisition. The aim here is to allow greater tuning of the acquisition to meet efficiency, quality or time restraints of the survey so that so a better match is obtained between the actual and desired survey.

The use of multiple sources is in part enabled by the commercialisation of interfering shot energy removal which has long been practiced in land seismic, but is much more challenging with marine streamer data due to the lack of azimuthal and offset variation of the source positions relative to the receivers. The ability to remove interfering shot energy means shotpoint intervals can be reduced, enabling multi-source designs by increasing inline fold, which is needed for successful processing of the data in domains such as 2D CMP and common-trace.

With triple sources we show that the options are greatly increased to either improve efficiency or quality and sometimes both, but we can also consider 5 or penta sources with which we obtain very high density data - 6.25m cross-line cells –with acquisition efficiency.

A 400km² survey was acquired on behalf of Quadrant Energy using the 5 source method and in addition a smaller 50km² using a conventional geometry. Direct comparisons can be made between the geometries and the simple fast-track processing shows the benefits of decreasing the cross-line sampling to fully realise the benefits of broadband data as high-frequencies are fully sampled in the cross-line domain.

The past ten years have been marked by dramatic advances in four seemingly isolated research fields: Thermodynamic modelling of minerals and rocks at high PT conditions, numerical simulation of the thermomechanical behaviour of the Earth's interior, efficient decomposition techniques to solve complex simulation-based problems, and probabilistic data analysis and inversion methods. All these disciplines/techniques have individually created true "revolutions" in the way we understand and model natural systems, including the interior of the Earth. Nevertheless, a more profound understanding is still ahead of us from the formal combination of these disciplines/techniques into a single operational framework to study the physical state of the Earth's interior.

In this contribution, I will present and discuss the concept of multi-observable probabilistic tomography or ''thermochemical tomography". This new kind of tomography is particularly designed for studies of the fundamental thermodynamic variables of the Earth's lithosphere, namely temperature, pressure and chemical composition. Once these variables are known, all physical parameters of interest (e.g. seismic velocities, density, viscosity, conductivity, etc), as well as traditional tomography images, are also retrieved in a thermodynamically-consistent way. The method is built on a simulation-based inversion technique where multiple satellite (e.g. gravity gradients, geoid height, etc) and land-based (e.g. seismic, plate motions, heat flow, etc) datasets can be jointly inverted to maximize the physical consistency of the resulting Earth model. Assembling this large problem required a collaborative effort between thermodynamicists, mineral physicists, geophysicists and geochemists, and marks the first step towards a full coupling between geophysics, geodynamics, thermodynamics, and geochemistry. I will present results for both synthetic and real case studies, which serve to highlight the advantages and limitations of this approach.

Two deep crustal seismic profiles, centred on the Olympic Dam Cu-Au-U deposit, were reprocessed by HiSeis Pty Ltd. using a proprietary method. This processing method aimed to enhance subtle variations in signal strength and highlight upper crustal discontinuities. The resulting images enable interpretation of steep structures and regions of enhanced/reduced reflectivity possibly associated with large-scale alteration zones. This work highlights additional information within these deep crustal seismic lines that illuminate different aspects of the geology.

The Coompana airborne survey is a large-scale pre-competitive aeromagnetic and radiometric survey in western South Australia flown in 2015. The survey was undertaken in order to provide greater geological controls on the basement geology in this area, which lies beneath the cover sequences of the Eucla, Bight, Denman and Officer Basins. This survey covers much of the South Australian part of the Coompana Province, which includes the Coompana Magnetic Anomaly. In this preliminary interpretation of the aeromagnetic survey, we have distinguished five basement domains based on the magnetic data and attempted to differentiate where domain boundaries are structural or intrusive. Integrating the aeromagnetic interpretation with existing drillhole and outcrop constraints suggests that the western part of South Australia had a long-lived, dynamic geological history with at least four major rock-forming events identified in the Coompana Province: c. 1610 Ma, c. 1500 Ma, c. 1180 Ma and c. 860 Ma. Detailed interpretation of the 200 m line-spaced infill region has enabled contact and relative age relationships of the intrusive bodies causing the Coompana Magnetic Anomaly. We suggest a multi-phase intrusive complex, with an early deep disc-like intrusion, followed by pipe-like satellite bodies intruding higher into the crust. The relative age of this intrusive complex is interpreted to be between c. 1120 Ma and c. 860 Ma.

Detailed petrologic and petrophysic data obtained from surface rock chip and diamond core of the Mt Leahy Tenement, Ekuti Range porphyry copper-gold exploration project area, Morobe Province, Papua New Guinea have been applied to interpretations of geologic setting and hydrothermal environment of mineralisation, and interpretations of airborne magnetic field data in defining a causative intrusion framework for mineralisation. Petrologic data confirm and refine a model of magmatic hydrothermal fluid sources for structurally confined mesothermal to epithermal style Cu, Au, Mo, Ag, Pb, Zn and Bi mineralisation within eroded composite granodiorite, quartz monzodiorite and diorite/andesite intrusion centres and hornfelsed metasedimentary rock, with potential for delineation of "disseminated" porphyry and hornfels styles of mineralisation. The application of petrophysic data, derived from petrology, to the interpretation of airborne magnetic field data defines three intrusion centres within the Leahy tenement: a northern and relatively deeply eroded Otibanda-Weke/Waikanda composite intrusion zone, and higher-level Kopekio and Ekoato intrusion domains, the latter with greater potential for discovery of low-grade high-tonnage style copper-gold mineralisation.

GPR in clay rich environments is often assumed to yield poor results due to signal attenuation, however it can still be possible to penetrate deep enough to reflect off targets of interest. Finding the bedrock level in an old quarry filled with clay and building rubble was successful using GPR reflection image profiling primarily due to the fact the fill material had previously been dewatered. Distinctive layers were observed in the fill material which when drilled corresponded to specific dominant material types. It appears the compaction level of the clay and the moisture content of the fill were important factors in the success of the GPR profile.

Edge detection is an essential task in the interpretation of potential field data. In this paper, we present a new method to delineate the edges of the sources, which is based on the windowed correlation coefficients of the average and the standard deviation of vertical derivatives. The zero value of the correlation coefficients is used to delineate the geological edges. This method can clearly give resolution of the edges of the deeper and shallower sources. The measure is initially applied on the synthetic gravity data. The test of theoretical models indicates that this method could detect the geological edges in different depths and the result is in accordance with model edges. Finally, this method is applied to gravity data from a portion of Vientiane Basin, Laos. As a result, the method can recognize geologic fractures more clearly. Moreover, it can recognize more geologic details when the window size is small and give superior results when the data are relatively smooth.

Comprehensive borehole seismic surveys hold the key to unlocking the knowledge contained within the long-offset full-azimuth surface seismic surveys that are fast becoming common in land exploration. This paper presents a case study of acquisition, processing and use of such survey for validation and calibration of processing parameters and inversion results of a recent point-source/point-receiver 3D surface seismic dataset acquired in the Cooper Basin.

The 3D surface seismic data have been initially processed without significant borehole seismic data input. However, as the processing revealed gaps in knowledge, the need for borehole calibration was realised. This led to acquisition of a complex borehole seismic survey in a gas-discovery well comprising Multi-Azimuthal Walkaway (MAZ WVSP), Walkaround and Zero Offset Vertical Seismic Profiles (ZVSP). The acquired dataset shed light into the peg-leg multiple mechanisms as well as the VTI and azimuthal anisotropy. Advanced processing techniques such as calibrated piece-wise VTI inversion and azimuthal travel time fitting were applied to the MAZ WVSP data to validate the processing steps of the 3D surface seismic data and calibrate the results of its AVOaz inversion.

Apart from showing some of the results of this study, this paper documents the various contact points between VSP and surface seismic datasets and how the results of processing complement each other. The final result comprised a calibrated seismic map of drilling targets.