Volume rendering has useful applications with emerging technologies such as virtual and augmented reality. The high frame rate targets of these technologies poses a problem for volume rendering because of its very high computational complexity compared with conventional surface rendering. We developed an efficient empty space skipping algorithm for accelerating volume rendering. A distance map is generated which indicates the Chebyshev distance to the nearest occupied region (with non-transparent voxels) within a volume. The distance map is used to efficiently skip empty regions while volume ray casting. We show improved performance over state-of-the-art empty space skipping techniques.

Severe marine heatwaves have recently become a common feature of global ocean conditions due to a rapidly changing climate. These increasingly severe thermal conditions are causing an unprecedented increase in the frequency and severity of mortality events in marine ecosystems, including on coral reefs. The degradation of coral reefs will result in the collapse of ecosystem services that sustain over half a billion people globally. Here, we show that marine heatwave events on coral reefs are biologically distinct to how coral bleaching has been understood to date, in that heatwave conditions result in an immediate heat-induced mortality of the coral colony, rapid coral skeletal dissolution, and the loss of the three-dimensional reef structure. During heatwave-induced mortality, the coral skeletons exposed by tissue loss are, within days, encased by a complex biofilm of phototrophic microbes, whose metabolic activity accelerates calcium carbonate dissolution to rates exceeding accretion by healthy corals and far greater than has been documented on reefs under normal seawater conditions. This dissolution reduces the skeletal density and hardness and increases porosity. These results demonstrate that severe-heatwave-induced mortality events should be considered as a distinct biological phenomenon from bleaching events on coral reefs. We also suggest that such heatwave mortality events, and rapid reef decay, will become more frequent as the intensity of marine heatwaves increases and provides further compelling evidence for the need to mitigate climate change and instigate actions to reduce marine heatwaves.

It is well established that small-scale heterogeneities can have a significant impact on the extent of the plume movement within a CCS storage complex. Small-scale heterogeneities cannot be directly incorporated in static and dynamic simulation models because of limitations in computer speed. Therefore, the scale-transgressive effects of small-scale heterogeneities in large-scale numerical simulations (static and dynamic models) must be accounted for through upscaling. Current upscaling techniques do not capture all relevant scales of heterogeneity and only give reliable results for a limited set of scales and flow scenarios. Accurate predictions of multiphase flow require dynamic reservoir models that capture geological heterogeneities and that are populated with representative equivalent static and dynamic flow properties.

Here we describe the development of a multiscale work program supporting the both the CTSCo Surat and CO2CRC Otway CCS demonstration projects. The study focusses on likely injection intervals in thePrecipice and Paaratte Formation Sandstones which both exhibit complex internal thinly laminated sedimentary features. We describe a workflow based on multiscale 3D imaging calibrated with laboratory and experimental probe analysis which seeks to address the impact of small-scale geological heterogeneity on the static and dynamic rock properties from pore to whole core (foot) scales and beyond. The workflow requires one to extrapolate static and dynamic properties measured at small volumes into a larger volume. The methodology is based on measurement, prediction and anchoring of data at multiple scales. The results of the program illustrate the importance of incorporating realistic geological structures at multiple scales to offer greater confidence in static and multiphase flow predictions at larger – from pore to whole core to geo-model/dynamic model – scales. The multiscale workflow illustrated in the figure includes:

1. Imaging – improved whole core acquisition to provide a quantitative bridge from pore scale imaging and plug scale measurements to whole core properties and log scale responses. Smaller samples also imaged to characterise pore structure at the highest resolution.

2. Classification – at each scale a novel morphological description of voxelised data is used in a multi-variate clustering analysis (MVCA) to identify the 3D distribution of rock types. This data can be used to locate representative volumes of multiple rock types. Classification at the coarsest (whole core) scale assists in sample selection for experimental measurement or imaging at higher resolution. The process can be repeated at descending scales until the elementary pore structures (individual pores/grains) are captured in 3D.

3. Calibration of Static properties – probe the static properties (permeability, resistivity) at the millimetre scale along continuous slabbed whole core surfaces along meter scales. Comparing the measured values to the ones obtained/predicted from the 3D digital rock modelling and propagation to similar rock types enables one to QC the static property predictions at the whole core scale.

4. Calibration of Dynamic Properties – undertake experimental 3D pore scale SCAL imaging studies on plugs acquired from core samples and image drainage and imbibition at aquifer conditions using time-lapse 3DCT imaging. Generate anchored SCAL data for important reservoir rock types including the generation of trapping curves and tensor-based flow properties – permeability and relative permeability data – on chosen subsamples acquired from whole core samples. Identify potential uncertainties with plug scale data – investigate e.g., impact of wettability, Swi.

5. Populate/Propagate – at a “representative” scale, petrophysical property correlations and saturation dependent properties derived from digital or laboratory data are defined on rock types.

6. Upscaling – extend measurements of static and dynamic properties to the whole core scale – assessing and accommodating the impact of small-scale geological heterogeneity on the static and dynamic rock properties along meter scale lengths of core material. Additionally lithofacies models are developed and populated using object-based geo-modelling tools – this offers a preliminary assessment of impact on larger lateral scales.

Digital rock analysis, an emerging technology driven by rapid advances in 3D pore scale imaging and computation, allows an unprecedented quantitative understanding of the pore scale at which all reservoir processes operate. Industry is seeking the extension of this technology to reliably derive and predict petrophysical & SCAL data along continuous lengths of core material and to integrate the data derived at the pore scale to predictions at increasingly larger scales (log characterization, geomodels and ultimately reservoir simulators). In this paper we describe an integrated pore to whole core scale upscaling method for a 6 meter section of the Precipice sandstone cores from the CTSCo Wandoan project. The workflow undertaken involves imaging at pore, plug and whole core scales coupled with experimental measures including mineralogy, gamma, mini-permeametry, RCA and SCAL data at plug and whole core scales. At each scale a novel morphological description of voxelised data is used in a multi-variate clustering analysis (MVCA) to identify the 3D distribution of rock types. This data is used to locate representative volumes of each rock type to be cored and imaged at a higher resolution. This process is repeated at descending scales of imaging until the elementary pore structures (intergranular pores and microstructure in clay) are captured in 3D. At this scale, petrophysical property correlations such as porosity:permeability, porosity:resistivity, and multi-phase flow properties such as capillary pressure and relative permeability are calculated using direct numerical methods on different rock types. The properties of each rock type calculated at the fine scale are used to populate the coarser image which is described by the geometric distribution of rock types and upscaled. The properties are propagated by taking into account the local variations observed within rock types. Depending on the complexity of the rock, a hierarchical propagation of up to 4 scales of imaging and upscaling can be undertaken for complete characterisation of the core material.

The objective of this study was to develop low-cost, robust and accurate sensor systems on a quadrotor for the implementation of a sensor-based vehicle formation control algorithm. The algorithm was recently developed at the ANU and was only ever tested in simulation.

A new quadrotor was assembled and mounted with a custom omnidirectional vision system to acquire bearings measurements of other vehicles. A custom omnidirectional lens was used which had a better vertical field of view compared with tested commercial solutions. The vision system, coupled with an efficient marker detection algorithm, detected bearings of other vehicles at 55Hz with a typical error of less than 5◦ in azimuth and elevation.

A widely used driver package for the Vicon motion capture system was modified to provide robust velocity measurements over a poor wireless channel. This was experimentally evidenced by a significant reduction in variability of velocity measurements when compared with a previously used method. An adaptation to the formation control algorithm was proposed which could function without velocity measurements for system damping and its performance was verified in simulation.

The robustness and accuracy of these key sensor measurements was demonstrated in a flight test where the quadrotor was flown under manual control. Support for offboard control was added to the flight controller so that the vehicle could be controlled by the vision system. Unfortunately, the quadrotor was not flown under the control of the formation control algorithm because of time constraints.