ASEG Extended Abstracts - ASEG2013 - 23rd Geophysical Conference, 2013

Geological interpretations of aeromagnetic and gravity images are highly subjective but are rarely accompanied by a quantitative confidence assessment, which is a key limitation on the usefulness of the results. This paper outlines a method with which the relative level of data richness can be assessed quantitatively, leading to an improved understanding of spatial variations in interpretational confidence. Simple rules were used to quantify the likely influence of several major sources of uncertainty. These were: 1) the level of geological constraint, using the local abundance of outcropping rock and the scale of geological mapping; 2) the interpretability of the data, considering the strength of edge-like features and the degree of directionality of these features; 3) data collection and processing errors, including gridding errors, and the influence of anisotropic line data collection on the detection of gradients. From these individual sources of uncertainty an overall data richness map was generated through a weighted summation of these grids. Weightings were assigned so as to best match the result to the interpreter’s perception of interpretational confidence. This method produced a map of data richness, which reflects the opportunity that the data provided to the interpreter to make a correct interpretation. An example from central Australia indicated that the data influences were preserved over a moderate range of weighting factors, and that strong bias was required to override these. In addition to providing a confidence assessment, this method also provides a way to test the potential benefits of additional data collection.

Magnetic and gravity data have great potential to inform us about regional-scale features. Major fault zones, basin geometries, basement character, Moho, intra-crustal boundaries etc can all be imaged. However, modelling results are inherently non-unique and often highly uncertain. This feeds into the reliability of the method and its usefulness in resource exploration. Following Occam’s razor, traditional methods have usually sought to find the simplest model possible, through use of maximum smoothness regularization (e.g. UBC-GIF), or through simplifying the model to an analytically unique problem (e.g. Parker-Oldenburg). More recent software packages allow the process to be constrained by lithology and the explicit incorporation of geological knowledge into the process. This increased degree of freedom leads to greater flexibility, but also greater ambiguity in results. A single result is clearly not adequate in these cases. Here we present some examples at regional and continental scales where the inclusion of variability measures has greatly increased the usefulness of the inversion process to understand 1) the features of interest and 2) the robustness of the solution.

Australia, via the efforts of the Government geological surveys, has a program of releasing ever bigger, higher- resolution, continental-scale datasets. The recently released isostatically corrected gravity data images many deep and large-scale crustal features. This is a key dataset for understanding the primary structure of the deep crust across thousands of kilometres. Direct “inversion” of this dataset to a consistent 3D fault surfaces network explains more than 50% of the primary information.

The method of choice relies on multi-scale edge detection or “worming”. This continues to prove effective in the regional mapping domain. Large-scale minerals and oil exploration mapping often make use of this technique. With the current shift to 3D geology modelling, issues arise to improve/generalise the worming technology to produce 3D contacts that can be interpreted, particularly the sub-set that indicates a primary fault network.

The new method allows the gathering of related worms to rapidly compute a consistent 3D fault network for the entire Australian continent by linking the dominant 30 km deep features back to the surface. If measured gravity curvature gradients are available, an improved and more detailed use of the method is now available at the prospect scale.

Potential field geophysical modelling techniques can be applied to better understand the subsurface morphology of volcanoes, and when linked with observations of surface geology can be used to develop a more complete understanding of the volcanic centres eruptive history.

High resolution ground gravity and magnetic data were acquired across several maar volcanoes located within the Newer Volcanics Province (NVP) of Western Victoria. The maar volcanoes surveyed represent a range of the different sizes and styles of eruptions observed within maar volcanoes of the NVP.

Gravity and magnetic data were subject to 2D forward and 3D inverse modelling in order to reveal details on the depth, geometry and petrophysical property distributions within subsurface volcanic structures. Gravity lows with corresponding magnetic highs are observed across the maar craters and were reproduced during modelling with the presence of a diatreme. Smaller wavelength gravity and magnetic anomalies detected in the centre of the more complex volcanic craters can be explained by the presence of intrusive dykes or vents filled with a higher proportion of denser volcanic debris.

Modelling suggests that multiple coalescing diatreme structures exist below the volcanic edifices, some containing intrusive dykes or a denser central vent filled in with volcanic debris. Multiple diatreme structures suggest a complex eruption history involving vent migration, while preserved dykes within the diatreme suggest short-lived fluctuations between phreatomagmatic and magmatic eruption styles.

The global demand for clean energy alternatives is constantly increasing, creating significant interest for more sustainable energy resources such as uranium and geothermal. Australia is host to over 25% of the world’s known uranium resources as well as having significant geothermal potential.

The Mount Painter Domain, in the Northern Flinders Ranges in South Australia, is in a region of anomalously high heat flow generated by radiogenic decay of uranium and thorium rich granites. Two distinct uranium deposits have formed from dissolved uranium carried from the ranges by fluids, being deposited where reduction in sediment pH causes uranium precipitation.

In May 2012 a magnetotelluric profile was collected, extending from the Northern Flinders Ranges to the Lake Frome embayment to help constrain existing resistivity models. Precipitation of uranium at the Beverley Mine site is anomalous as no surface water flow is present, suggesting the presence of subsurface processes. A conductive pathway extending to the surface at Beverley is linked to a 50 Ω m conductive body at the brittle- ductile boundary of the mid-crust, directly under the Paralana geothermal prospect.

To better characterise the eruptive histories and subsurface structures of several maar volcanoes from the Newer Volcanics Province, forward and inverse geophysical modelling is combined with a study on the geology of the volcanic centres. The maar volcanoes under investigation include the Red Rock Volcanic Complex (RRVC), Ecklin maar, the Mount Leura Volcanic Complex (MLVC).

High resolution gravity and magnetic data were acquired across each of the maars and the data was modelled in two and three dimensions to understand the subsurface morphology of the volcanoes vents.

Varied geophysical responses are observed across each of the maars surveyed, indicating the complex and variable nature of the subsurface volcanic vent, even when they present similar surface morphology. Where corresponding gravity and magnetic lows are detected across a maar crater, it is suggested that all the available magma was erupted and the maar diatreme (subsurface collapse structure) was not intruded by any dykes. The gravity low arises because of lower density lake sediments and pyroclastic debris infilling the crater. The lack of any intrusive dykes or remnant vents within the diatreme suggest that groundwater was available for phreatomagmatic explosions.

Maars with corresponding gravity and magnetic highs indicate a large volume of subsurface basalt, resulting from the ponding of magma at the surface of the vent. This results from a lack of groundwater for magma to interact with during the eruption, which facilitates magma rising upwards through the diatreme where it is fragmented at shallower levels.

As a part of the Airport Tunnel project in Brisbane, an MASW survey was carried out to investigate the nature of the fill site of an old rock quarry. The main target of the survey is the depth of the landfill after the quarrying operation to avoid the tunnelling penetrate the fill material. The site is currently used as a shopping centre car park in an inner suburb of Brisbane along one of the main roads. To avoid excessive noise, the survey took place at night.

Some part of the seismic data presented an unusual but consistent noise pattern. External sources of this unusual noise, such as buried electric cables and drainage were considered, but no source could account for it. Therefore the noise was determined to be due to composition of the materials underground.

The analysis of the seismic data outlined the depth and shape of the quarry. The noisy area was identified as a shallow hard rock left in the quarry, perhaps as an access ramp. This was verified a historic photography of the old quarry.

The depth of the quarry was found shallower than the proposed depth of the tunnel to ensure the safety of boring.

This paper is inspired by an image of a time slice from the prolific hydrocarbon state of Texas showing a stacked channel system. Similar channel systems also exist in Australia’s Cooper Basin but are often difficult to see with legacy sparse seismic acquisition geometries. Due to reasons of cost and environmental and cultural heritage protection, relatively wide (‘sparse’) line intervals have been used, though these also allow coverage of larger areas than would otherwise be achieved. These sparse designs can combine low environmental impact with reasonable images at target. Lack of traces at medium and near offset ranges may result in strong amplitude artefacts in the final image - “acquisition footprint”. This case study is from the 2012 acquisition where, to reduce these artefacts, we deployed sources in a smooth “wavy” sinusoidal pattern, modified as necessary to follow natural features in the terrain. This methodology results in acquisition with a minimal visual and environmental impact and provides significant benefits in reducing the acquisition footprint. Innovative survey design and data processing techniques which accommodate non-uniform sampling resulted in the dataset where channel features are now clearly visible on the migrated volumes. The acquisition technique also features broadband point- source vibroseis using a non-linear Maximum Displacement sweep of 2 to 100 Hz, broadband digital point-receivers and dense sampling along both the source and receiver lines. Using the described technologies the acquired survey not only met, but by far exceeded, the initial expectations, and within the specified time frame. This survey has shown that exploration seismic surveys can be tailored to minimize environmental and acquisition foot print, and provide a high quality seismic dataset suitable for seismic attributes extraction.

Interpreting magnetotelluric (MT) models requires solid modelling of the data as well as good knowledge from other geophysical data and geological constraint in the particular tectonic setting of the survey area. MT measurements, relating the natural variations of electric and magnetic field to obtain the electrical resistivity distribution of the crust and mantle appear to show that enhanced electrical conductivity zones are more abundant at certain depths. Models show that frequently enhanced conductivity zones are topping out in the upper crust at depths of about 10-15 km. These features are discrete and extend usually over a few km to tens of km laterally, and can be found across the Delamerian Orogeny, in zones of high heat flow east of the Northern Flinders Ranges and also in the central Eyre Peninsula. We interpret this to be related to recent findings on dynamic interactions between brittle and ductile layers leading to mid- to upper crustal detachment faults. A second zone of higher conductivity occasionally appears in the lower crust, as imaged east of the Flinders Ranges at depths of around 25-35 km. Thirdly, at 80 km depth mantle conductors appear in stable Archean and Proterozoic terranes around the world, such as in the Slave Craton, Kaapvaal Craton and the Gawler Craton. In summary, information from geodynamic modelling helps to understand the processes in the earth in regards to fluid movement and potential mapping of heat flow and corresponding shift in depth of brittle-ductile boundaries.

We have investigated techniques for the efficient early detection of landslides using time series analysis incorporating synthetic aperture radar (SAR) images. The study area, in the Miyazaki, Nagasaki, and Saga prefectures in Kyushu, was determined based on interference fringes detected during interference SAR (InSAR) analysis. We used ALOS/PALSAR data acquired from 2006-2011 to detect early warning signs of landslides that were poorly expressed geomorphologically by conducting time series analysis of InSAR data acquired periodically. Moreover, in order to remove the noise caused by geographical feature stripes or phase retardation, we applied median filtering, histogram extraction processing, and clearisation of the displacement with a Laplacian filter. We evaluated the validity of each filter separately and in combination with other filters and assigned a gradient vector to each pixel value of the SAR picture using altitude data. In order to confirm the assumption that surface-of-the-earth displacement proceeds in the dip direction, we conducted direction analysis and a technique and confirmed the results of InSAR analysis by field survey. Our results prove the effectiveness of InSAR analysis in hazard monitoring over a wide area through the detection of local landslides.

Geoscientific data interpretation is a highly subjective and complex task as human intuition and biases play a significant role. Based on these interpretations, however, mining and petroleum industries make decisions with paramount of financial implications. As a first step towards understanding and improving the interpretation process, we carried out two experiments to monitor the human-data interactions during the process of identifying ‘targets’ (porphyry-style intrusive systems) within the aeromagnetic imagery. This is achieved by capturing the eye gaze position using an eye tracker system and the brain responses using electroencephalography (EEG).

The first experiment was intended to analyse the target spotting performance and the data observation patterns. For this experiment participants performed exercises, where the same magnetic image was presented in different orientations. Some key findings include: inconsistencies in target spotting performance within and between the interpreters; an improvement performance when the data were viewed in multiple orientations; and a strong correlation between the target spotting performance and efficient (systematic) data observation pattern. There was no correlation between success in identifying targets and the participants’ perception of their expertise.

The second experiment was designed to identify the characteristics of the targets that are easier to detect using EEG. For this experiment images with targets and without targets were presented in a rapid visual display. The analysis on the image characteristics based on the human visual attention model show a strong correlation between target spotting difficulty and dispersion of the visual attention.

Accurate and efficient identification and mapping of geological structures has broad application across the minerals industry. Recent advances in data acquisition technologies using Unmanned Aerial Vehicles (UAV), have led to a growing interest in capturing high- resolution rock surface images and analysing those datasets remotely. However due to the large volumes of data that can be captured in a short flight, efficient analyses of these data brings new challenges.

We propose a semi-automated method that allows efficient mapping of geological structures using photogrammetry of rock surface data collected by UAV. Our method harnesses advanced automated image analysis techniques and human data interactions to identify structures and calculate dip and dip angles of structures. Geological features were detected in two dimensional (2D) images and the corresponding three dimensional (3D) features were automatically identified from 3D surface models. The location, dip and dip angle of geological features were then calculated.

A feature map generated by our semi-automated method correlates well with a fault map resulting from visual interpretation by an expert. Some advantages of our semi-automatic method include the following: Firstly; it generates results in few minutes whilst manual interpretation took around an hour, thus contributing significantly in time saving. Secondly; unlike manual interpretation, our software technology provides objective and consistent results that can be reproduced.

We evaluate the use of airborne electromagnetic data from the SPECTREM2000 system flown for ore body detection, regolith mapping and assessment of aquifers. Since the position and orientation of the receiver bird are not measured, the primary field at the bird cannot be known and removed precisely. In order to successfully invert the AEM data, and produce conductivity-depth models, we first reinstate the removed primary field estimate and convert the data from ppm units to Teslas. We then simultaneously inverte the X and Z component data, to solve for a 1D layered conductivity model and receiver position.

The SPECTREM system has flown many line kilometres in other parts of the world but substantially less in Australia. Through further processing and inversions we have resolved conductivity-depth structures very similar to those previously obtained from other well-established AEM systems flown under Australian conditions. We also present a section of AEM data with logged drilling core data as a means of assessment of our inversion models against an independent data set.

This paper demonstrates recent advances in grid-based reflection tomography model building in conjunction with anisotropic pre-stack depth migration. These advances significantly benefit the imaging and resolution of 3D seismic data in a number of structurally complex basins.

Using 3D seismic imaging examples from Petroleum Exploration Permit PEP 51558 in New Zealand’s Taranaki Basin, this paper demonstrate how reflection tomography techniques can utilize implicit geological constraints to resolve complex velocity variations and, thereby, reduce structural and imaging uncertainty.

One challenge associated with velocity modelling in the Taranaki Basin is to adequately resolve the large lateral velocity contrasts across major faults. This work shows how the use of steering filters during the reflection tomography process significantly improves the resolution and delineation of these fault-constrained velocity contrasts. This approach is compared with the conventional gridded tomography approach of velocity model building.

Using a detailed velocity model from the Taranaki Basin, comparisons were made between time and depth migration, as well as alternate depth migration techniques.

Geophysical data interpretation is largely an anomaly detection task which involves recognising and synthesising anomalous patterns within single or multiple datasets. The accuracy and efficiency of these interpretations heavily relies on the skills and practices of interpreters, thus the greatest challenge is to minimise personal biases to produce objective and consistent interpretation outcomes. We present an innovative data visualisation method which can empower interpreters to effectively delineate anomalies of varying frequency scales within aeromagentic data using a single image display. This is achieved by harnessing the power of image enhancement and visualisation techniques to assist interpretation.

We adapted and extended the use of colour composite techniques to present different frequencies presented in potential field data. Aeromagnetic data from an area in Kirkland Lake, Ontario, Canada is used for our experiment. long wavelength and short wavelength anomalies are identified from the data using low pass- and high pass filters respectively. These two different frequency enhanced images and the original image are represented as separate colour channels which are then combined to generate a composite image. The luminance of the composite image is scaled to highlight high frequency signals as they hold the key for detailed structure interpretation. We use a technique called dynamic range compression, which preserves the integrity of the phase component of the signal while performing high pass filtering. The resulting display is compared to the geological map of the area to validate the effectiveness of the method. The proposed technique is widely adaptable for different types of datasets.

In 2011 a review of an historical TEMPEST airborne electromagnetic survey in the Murchison Region of Western Australia identified a number of discrete bedrock conductor anomalies potentially associated with base metal mineralisation. One of the anomalies identified was in close proximity to a known gossan at Hollandaire, within rocks of the Archaean Mt Eelya Complex. The Hollandaire gossan had been previously investigated during the mid-late 1970’s using a variety of geological, geochemical and geophysical methods, including ground magnetics, induced polarisation, magnetic induced polarisation and time-domain electromagnetics.

A single line of time-domain in-loop transient electromagnetic data was collected at Hollandaire during 2011, in order to follow up the TEMPEST bedrock conductor. A strong time-domain EM anomaly was identified, with a very high time-constant of 107 ms. Plate modelling of the in-loop data resulted in a target at depth-to-top 100 m, dipping at 45 degrees to the west- northwest, and with conductance 5000 S. The nine initial holes drilled to test the electromagnetic target all intersected sulphide mineralisation. The inferred resource now totals 2.8 million tonnes at 1.6% Cu, 0.4 g/t Au and 5 g/t Ag, with the supergene zone averaging 4.7% Cu. Subsequent downhole electromagnetic (DHEM) surveys have identified an off-hole electromagnetic conductor to the south of the known mineralisation, which has not yet been tested by drilling.

Seismic methods can be utilized in unconventional resources characterization studies to achieve an improved understanding of the entire reservoir heterogeneity, structure and stress orientation. This assists in an identification of production “sweet spots” and more efficient well placement. To enable this type of study with surface seismic, we need to analyse and invert the data not only against offset but also azimuth. This requirement places greater demands on the seismic than would be the case for a purely structural image.

In this paper we will describe the use of a potential “best-practice” solution based on experience in Australia and elsewhere for the design and implementation of the high specification “Winnie 3D” seismic survey. This survey featured broad-band point-sources and point-receivers using a non-linear maximum displacement sweep of 1.5 to 110 Hz. The omni-directional symmetrical dense sampling, in combination with long offsets, resulted in uniform azimuthal coverage and extremely high trace density. This design combined with a broad-band acquisition enables azimuthal analysis, inversion and seismic attributes extraction.

We will demonstrate how this 3D design, tailored for unconventional targets, allowed extraction of seismic attributes even in the early stages of data processing, enabling detection of anomalies that could be related to shallow igneous intrusions.

Over the last two decades 3D reflection seismic has been applied for mine planning in South Africa. For coal exploration, the mining targets are at shallower depths of less than 400 m. The challenge set by the mining companies is to obtain high-resolution seismic data with maximum vertical resolution at this depth of investigation. The cost of geophysics must also be more attractive than that of the ‘total drilling’ alternative. Seismic source trials were therefore conducted at one mine site. The ICIS (Internal Combustion Impulse Source) has been developed by CGG to provide a solution for shallow target surveys and for in-fills where access for conventional sources is restricted. The ICIS and Mini-vibrator performance were tested in shallow coal prospects, where existing 2D data have been acquired by Nomad 65 heavy vibrator. The tests demonstrated that ICIS is well suited to image target depths less than 500 m with frequencies up to 100 Hz. The comparison with the vibroseis sources showed that ICIS I) Delivered very good near offset data (> 50 m) and for up-hole/low velocity layer survey ii) Required more shots per shot point in the medium offset range (1000 m) due to its low energy iii) Not designed for far offset data (3000 m). ICIS was found to be a viable geophysical and logistic solution for certain high-resolution mineral surveys.

Full Tensor Gradient (FTG) gravity data measures the derivatives of the Earth’s gravitational field. Such variations in the gravitational field may be due to the presence of bodies of higher or lower density relative to the surrounding rock.

As the gravity tensor contains 5 independent components, effective visualisation of this high-dimensional dataset is advantageous for efficient processing of the FTG data. We present two aspects of visualising mass anomalies in FTG gravity datasets. First, we create a textured image where the orientations of the resulting texture reflect local lateral orientations encoded in the FTG data. It uses a colour map to highlight geologically significant structures such as linear features and radially symmetric points by identifying different geological features and using colour components to represent different feature types. This visualisation method is shown to be robust to significant levels of modelled noise, and we demonstrate its applicability to a field FTG survey.

Second, we present an algorithm for estimating the depths of mass anomalies in FTG data. A voxel representation of the subsurface is created and voxels are voted for according to gravitational curvature properties encoded in the FTG tensor. A visualisation of the volume at successive depths highlights 3D locations of mass anomalies at local maxima of the volume. The algorithm is evaluated on a forward-modelled FTG dataset where the depths of mass anomalies are known. The depths of mass anomalies are shown to be accurately located in the presence of noise.