From 1 - 10 / 20186
  • This paper presents a model to assess bushfire hazard in south-eastern Australia. The model utilises climate model simulations instead of observational data. Bushfire hazard is assessed by calculating return periods of the McArthur Forest Fires Danger Index (FFDI). The return periods of the FFDI are calculated by fitting an extreme value distribution to the tail of the FFDI data. The results have been compared against a spatial distribution of bushfire hazard obtained by interpolation of FFDI calculated at a number of recording stations in Australia. The results show that climate simulations produce a similar pattern of bushfire hazard than the interpolated observations but the simulated values tend to be up to 60% lower than the observations. This study shows that the major source of error in the simulations is the values of wind speed. Observational wind speed is recorded at a point-based station whilst climate simulated wind speed is averaged over a grid cell. On the other hand FFDI calculation is very sensitive to wind speed and hence to improve the calculation of FFDI using climate simulations it is necessary to correct the bias observed in the simulations. A statistically-based procedure to correct the simulation bias has been developed in this project. Bias-corrected calculation of FFDI shows that the major bushfire hazard in south-eastern Australia is in the western parts of SA and NSW; and in south-western Tasmania.

  • 22-2/D52-09/2-6 Contour interval: 25

  • The sensitivity of the Jaiswal and Wald (Earthq. Spectra, 2010) empirical earthquake fatality model is evaluated relative to the model space for a suite of macroseismic intensity prediction methods. The relative difference between intensity prediction methods is shown through the use of self-organizing maps to visualize high-dimensional ground shaking data in a two-dimensional space. Among all the macroseismic intensity prediction methods evaluated, there is significant variability in the resulting loss estimates for an earthquake of given source parameters with losses being most sensitive to those intensity models that predict high near-source ground shaking. Because the empirical fatality models evaluated herein are based on a consistent suite of ground-motion model inputs, application of the fatality models with other intensity prediction methods may result in undesirable outcomes. Consequently, it is recommend that empirical loss models be calibrated directly with hazard inputs used in the proposed loss assessment methodology.

  • 200 m line spacing coverage to the east 22-3/J54-4/3-2

  • This study brings together a wide range of datasets to provide a comprehensive assessment of the Pandurra Formation sedimentology and geochemistry in 3D. This record is associated with both the GA Record and the digitial data release. Sedimentology and geochemistry datasets generated this study are combined with pre-existing data to generate a 3D interpretation of the Pandurra Formation and improve understanding of how the Pandurra Formation as we see it today was deposited and subsequently post-depositionally mineralised. The digital release incorporates the underlying digital data generated this study, the final gOcad objects generated, and reference datasets from Wilson et al., 2011 as required. Study extent in eastings and northings: SW Corner (444200, 6263000) NE Corner (791409, 6726000).

  • This study demonstrates that seabed topography and geodiversity play key roles in controlling the spatial dynamics of large fish predators over macro-ecological scales. We compiled ten years of commercial fishing records from the Sea Around Us Project and developed continental-scale catch models for an assemblage of large open-water fish (e.g. tuna, marlins, mackerels) for Western Australia. We standardised catch rates to account for the confounding effects of year, gear type and species body mass using generalised linear models, from which relative indices of abundance were extracted. We combined these with an extensive array of geophysical, oceanographic, biological, and anthropogenic data to (1) map the location of pelagic hotspots and (2) determine their most likely mechanistic drivers. We tested whether submarine canyons promote the aggregation of pelagic fish, and whether geomorphometrics (measures of seafloor complexity) represent useful surrogate indicators of their numbers. We also compared predicted fish distributions with the Australian network of Commonwealth Marine Reserves to assess its potential to provide conservation benefits for highly mobile predators. Both static and dynamic habitat features explained the observed patterns in relative abundance of pelagic fish. Geomorphometrics alone captured more than 50% of the variance, and submarine canyon presence ranked as the most influential variable in the North bioregion. Seafloor rugosity and fractal dimension, salinity, ocean energy, current strength, and human use were also identified as important predictors. The spatial overlap between fish hotspots and marine reserves was very limited in most parts of the EEZ, with high-abundance areas being primarily found in multiple use zones where human activities are subject to few restrictions.

  • 75% coverage, SW area is missing 22-3/J54-8/3-1/4

  • 22-1/H52-08/4 Vertical scale: 200