The importance of disaster-risk reduction in ensuring long-term sustainability of development and economic growth has gained increased awareness within the international development community, and thereby highlighted a need for a broad assessment of natural hazards across the Asia-Pacific region. A key component of this assessment involves qualifying, and ultimately quantifying, the frequency and potential consequences of large Volcanic Explosivity Index (VEI) of 4 or more volcanic eruptions in this region. The frequencies of large eruptions were determined from frequency-magnitude plots using eruption data provided by the Smithsonian Institution's Global Volcanism Program. However, calculated frequencies represent only minimum values. This is because roughly half of the volcanoes in the region have no eruption chronologies, the eruption record for the most part extends back only 400 years, and good records exist for only the last 180 years. A rough analysis was undertaken to estimate the populations likely to be impacted by large volcanic eruptions, where 'impacted' refers to possible death, injury, building damage, loss of access to basic services, and failure of industrial/agriculture production. The following conclusions were made from frequency-impact plots: - Indonesia and the Philippines have the highest level of risk with respect to volcanic eruptions, in terms of total population impacted. - Volcanic disasters affecting populations of 100,000 or more can be expected at least every decade in Indonesia and once every few decades in the Philippines. - Both Indonesia and the Philippines, at current population levels, have the potential to experience volcanic disasters affecting at least 1 million people, at a rate of once and twice a century, respectively. - All of the countries for which results were obtained - Indonesia, Philippines, Papua New Guinea, Vanuatu and Tonga have the potential for a volcanic disaster that will impact at least 1% of the population, but at different rates: twice a century for Vanuatu, around twice a millennium for Indonesia and the Philippines, and around every millennium for Papua New Guinea and Tonga - Vanuatu has the potential for catastrophic volcanic disaster that seriously affects more than 5% of the population around once a millennium.

Tropical Cyclone Tracy impacted Darwin early on Christmas Day, 1974. The magnitude of damage was such that Tracy remains deeply ingrained in the Australian psyche. Several factors contributed to the widespread damage, including the intensity of the cyclone and construction materials employed in Darwin at the time. Since 1974, the population of Darwin has grown rapidly, from 46,000 in 1974 to nearly 115,000 in 2006. As such it is desirable to know what impact an event similar to Tracy would have on the present day built environment. To assess the impacts in 1974 and the present day, we apply the Tropical Cyclone Risk Model (TCRM) developed at Geoscience Australia. The TCRM is used to produce a wind field consistent with the sparse meteorological observations made in Darwin during the passage of TC Tracy. The wind speed record from the anemometer located at Darwin Airport is used to derive an empirical radial wind profile, which is employed as a control point for the parametric wind profiles generated within TCRM. A parametric wind field is calculated using the Holland radial wind profile coupled with the Kepert boundary layer model. The resulting wind field is further modified by incorporating localised windfield multiplier effects. The resulting wind field predicts maximum wind gusts at Darwin Airport of 72 m/s (sustained over three seconds at ten metres elevation), which matches the estimated maximum wind speed at that location (BoM, 1977). The wind field generated with TCRM is applied to building damage models in an attempt to reproduce the widespread damage to residential structures associated with TC Tracy in 1974. Employing these models yields a mean damage estimate of 36% of replacement cost across all residential building stock in Darwin. This result underestimates the observed damage by around 20%. One possible explanation for this discrepancy is the significant damage attributable to large windborne debris. Based on the satisfactory replication of the damage associated with the historical impacts of TC Tracy, the wind field is then applied to the current day residential building database in order to assess the impact of TC Tracy were it to strike Darwin in 2008. We find that the mean damage to Darwin for the same urban footprint as the 1974 analysis in the present day would be around 5%. This represents an approximately 90% reduction in the modelled damage, and a significant portion of this reduction can be attributed to building code improvements.

The cyclonic wind hazard over the Australian region is determined using synthetic tropical cyclone event sets derived from general circulation models (GCMs). Cyclonic wind hazard is influenced by the frequency, intensity and spatial distribution of tropical cyclones, all of which may change under future climate regimes due to influences such as warmer sea surface temperatures and changes in the global circulation. Cyclonic wind hazard is evaluated using a statistical-parametric model of tropical cyclones - the Tropical Cyclone Risk Model (TCRM) - which can be used to simulate many thousands of years of cyclone activity. TCRM is used to generate synthetic tracks which are statistically similar to the input event set - be it an historical record of other synthetic event set. After applying a parametric wind field to the simulated tracks, we use the aggregated wind fields to evaluate the average recurrence interval wind speeds for three IPCC AR4 scenarios, and make comparisons to the corresponding average recurrence interval wind speed estimates for current climate simulations. Results from the analysis of two GCMs are presented.

The term "Smartline" refers to a GIS line map format which can allow rapid capture of diverse coastal data into a single consistently classified map, which in turn can be readily analysed for many purposes. This format has been used to create a detailed nationally-consistent coastal geomorphic map of Australia, which is currently being used for the National Coastal Vulnerability Assessment (NCVA) as part of the underpinning information for understanding the vulnerability to sea level rise and other climate change influenced hazards such as storm surge. The utility of the Smartline format results from application of a number of key principles. A hierarchical form- and fabric-based (rather than morpho-dynamic) geomorphic classification is used to classify coastal landforms in shore-parallel tidal zones relating to but not necessarily co-incident with the GIS line itself. Together with the use of broad but geomorphically-meaningful classes, this allows Smartline to readily import coastal data from a diversity of differently-classified prior sources into one consistent map. The resulting map can be as spatially detailed as the available data sources allow, and can be used in at least two key ways: Firstly, Smartline can work as a source of consistently classified information which has been distilled out of a diversity of data sources and presented in a simple format from which required information can be rapidly extracted using queries. Given the practical difficulty many coastal planners and managers face in accessing and using the vast amount of primary coastal data now available in Australia, Smartline can provide the means to assimilate and synthesise all this data into more usable forms.

A 'shake-map' represents the spatial distribution of macroseismic intensity resulting from an earthquake. These maps are often used to determine potential humanitarian consequences from scenario earthquakes, or in near-real time following the detection of an event. In the absence of dense strong-motion networks to calibrate real-time ground-shaking in many of the most vulnerable regions of the world, shake-maps are commonly generated using either Intensity Prediction Equations or Ground-Motion Prediction Equations combined with Ground-Motion to Intensity Conversion Equations. There are several empirical models available to estimate the spatial distribution of intensity for an earthquake of given magnitude and location. However, these models can predict very different estimates of shaking intensity given the same input parameters; particularly at near-source distance ranges - the most critical distances for impact assessments. Consequently, the application of different shaking hazard model inputs can result in significantly different impacts. High-dimensional information visualisation techniques are used to study the mutual differences among different empirical intensity prediction models. We applied the Self-Organising Map (SOM) and Principle Component Analysis (PCA) techniques to project empirical prediction models onto a two-dimensional 'map' to visually compare the similarities and differences between models. The results clearly demonstrate the sensitivity of ground shaking to the selection of intensity prediction models. The effect of these sensitivities to earthquake impact assessments are investigated using a scenario event in Sumatra region, Indonesia.

We report on an assessment of severe wind hazard across the Australian continent, and severe wind risk to residential houses (quantified in terms of annualised loss). A computational framework has been developed to quantify both the wind hazard and risk due to severe winds, based on innovative modelling techniques and application of the National Exposure Information System (NEXIS). A combination of tropical cyclone, synoptic and thunderstorm wind hazard estimates is used to provide a revised estimate of the severe wind hazard across Australia. The hazard modelling utilises both 'current-climate information and also simulations forced by IPCC SRES climate change scenarios, which have been employed to determine how the wind hazard will be influenced by climate change. We have also undertaken a national assessment of localised wind speed modifiers including topography, terrain and the built environment (shielding). It is important to account for these effects in assessment of risk as it is the local wind speed that causes damage to structures. The effects of the wind speed modifiers are incorporated through a statistical modification of the regional wind speed. The results from this current climate hazard assessment are compared with the hazard based on the existing understanding as specified in the Australian/New Zealand Wind Loading Standard (AS/NZS 1170.2, 2002). Our analysis has identified regions where the design wind speed depicted in AS/NZS 1170.2 is significantly lower than 'new' hazard analysis. These are regions requiring more immediate attention regarding the development of adaptation options including consideration by the wind loading standards committee for detailed study in the context of the minimum design standards in the current building code regulations.

We report on an assessment of severe wind hazard across the Australian continent, and severe wind risk to residential houses (quantified in terms of annualised loss). A computational framework has been developed to quantify both the wind hazard and risk due to severe winds, based on innovative modelling techniques and application of the National Exposure Information System (NEXIS). A combination of tropical cyclone, synoptic and thunderstorm wind hazard estimates is used to provide a revised estimate of the severe wind hazard across Australia. The hazard modelling utilises both 'current-climate' information and also simulations forced by IPCC SRES climate change scenarios, which have been employed to determine how the wind hazard will be influenced by climate change. We have also undertaken a national assessment of localised wind speed modifiers including topography, terrain and the built environment (shielding). It is important to account for these effects in assessment of risk as it is the local wind speed that causes damage to structures. The effects of the wind speed modifiers are incorporated through a statistical modification of the regional wind speed. The results from this current climate hazard assessment are compared with the hazard based on the existing understanding as specified in the Australian/New Zealand Wind Loading Standard (AS/NZS 1170.2, 2002). Our analysis has identified regions where the design wind speed depicted in AS/NZS 1170.2 is significantly lower than 'new' hazard analysis. These are regions requiring more immediate attention regarding the development of adaptation options including consideration by the wind loading standards committee for detailed study in the context of the minimum design standards in the current building code regulations.

The National Exposure Information System (NEXIS) is a capability developed by Geoscience Australia, an agency within the portfolio of the federal Department of Resources, Energy and Tourism. NEXIS is a nationally consistent database of building assets, essential infrastructure, economic activity and demographic information. All these community elements are at risk to natural hazards and will be exposed to the unavoidable, long term influences of climate change. The system collects and collates a broad range of information for research and policy development in Australia, including that associated with climate change adaptation. The development of NEXIS has been undertaken in parallel to ongoing national assessments of climate change risk for hazards such as storm surge, severe wind, bushfire and extreme temperature NEXIS employs a largely statistical approach to developing a national definition of exposure using a number of existing databases maintained by others. These include the Geocoded National Address File (GNAF), the Property Cadastre, the Business Registry, and census datasets from the Australian Bureau of Statistics. Costing modules developed by quantity surveyors have also been incorporated to provide estimates of building replacement costs across Australia. State Government departments have supplied data on local building information in Tasmania and South Australia. The Census of Land Use and Employment (CLUE) has also been made available by local government for comprehensive information about land use, employment and economic activity across the entire Greater Melbourne area.

The short historical record of tropical cyclone activity in the Australian region is insufficient for estimating return period wind speeds at long return periods (greater than 100 years). Utilising the auto-correlated nature of tropical cyclone behaviour (forward speed and direction, intensity and size), Geoscience Australia has developed a statistical-parametric model of tropical cyclone behaviour to generate synthetic event sets that are statistically similar to the historical record. The track model is auto-regressive, with lag-1 auto-regression used for forward speed and bearing, and lag-2 auto-regression applied to the intensity and size characteristics. Applying a parametric wind field and a linear boundary layer model to the synthetic tropical cyclone tracks allows users to generate synthetic wind swaths, and in turn fit extreme value distributions to evaluate return period wind speeds spatially. The model has been applied to evaluate severe wind hazard across Australia and neighbouring regions. In conjunction with statistical models of synoptic (mid-latitude storms) and thunderstorm wind hazard, we have been able to generate a national assessment of severe wind hazard, which is comparable to existing wind loading design standards. Using tropical cyclone-like vortex tracks directly detected from regional climate models, it is also possible to project cyclonic wind hazard into future climate conditions, accounting for both changes in frequency and intensity of tropical cyclones.

We describe a weighted-average approach for incorporating various types of data (observed peak ground motions and intensities, and estimates from ground motion prediction equations) into the ShakeMap ground motion and intensity mapping framework. This approach represents a fundamental revision of ShakeMap technique, particularly as it pertains to processing ground motion and intensity data. Combining ground motion and intensity data onto composite ShakeMaps proves invaluable for loss calibration of historical events as well as for loss estimation in near-real time applications. In addition, the increased availability of near-real-time macroseismic intensity data, the development of new relationships between intensity and peak ground motions, and new relationships to directly predict intensity from earthquake source information, have facilitated the inclusion of intensity measurements directly into the ShakeMap computations. Our approach allows for the possible combination of all of the following data sources and estimates: 1) nearby observations (ground motion measurements and reported intensities), 2) converted observations from intensity to ground motion (or vice-versa), and 3) estimated peak ground motions from prediction equations (or numerical estimates).