<div>The Australian National Seismic Hazard Assessment (NSHA) was updated in 2023 to coincide with the revision of the Australian Standard for the structural design actions for earthquake (AS1170.4). This updates the 2018 NSHA (NSHA18), which applied globally accepted, best-practice probabilistic approaches and yielded considerably lower hazard estimates at the 1/500 annual exceedance probability relative to previous national assessments and the AS1170.4–2007 seismic design values. Whilst the science underpinning the NSHA18 has withstood on-going peer review and critique, Geoscience Australia has reflected on this model and has advocated for adjustments to modelling choices where warranted. The use of structured expert elicitation was a major advance in the development of the NSHA18 and promoted ownership of the model amongst the Australian seismological community. However, some consequences of the modelling choices made during that process were not fully appreciated at the time. The 2023 NSHA (NSHA23) was intended to be a modest update to the 2018 model and again is characterised through structured expert elicitation. However, there are some key changes to the earthquake catalogue, which include: 1) the inclusion of events from mid-2017 through to the end of 2022; 2) the recalculation of local magnitudes for recent events (2010 onwards) to resolve recently discovered inconsistent observatory practice by different Australian monitoring agencies and errors with station metadata, and; 3) the subsequent revision of local magnitude to moment magnitude conversions. All seismic source models in the updated NSHA23 that rely on the updated earthquake catalogue (pre-instrumental and instrumental) are updated to reflect these changes. Incremental updates to the fault-source model and changes to the relative weights between different source-model class types have also been implemented. Another significant advance is the augmentation of the Australian Ground-Motion Database with new and legacy data. High-quality data acquired from Australian earthquakes since 2018 were used to enable more informed choices for the ground-motion characterisation model. In summary, the 2023 updates to the instrumented earthquake catalogue have led to reduced earthquake rates and seismic hazard. This change is counteracted using an updated suite of ground-motion models. Adjustments to weights in the source characterisation model yield spatially variable changes in hazard, most commonly leading to modest increases in hazard. Presented at the 2024 18th World Conference on Earthquake Engineering, Milan, Italy

<div>Geoscience Australia, together with contributors from the wider Australian seismology community, have produced a new National Seismic Hazard Assessment (NSHA23), recommended for inclusion in proposed updates to Standards Australia’s&nbsp;AS1170.4. NSHA23 builds on the model framework developed for NSHA18, and incorporates scientific advances and stakeholder feedback received since development of that model. Key changes include: further refinement and homogenisation of the earthquake catalogue; revisions to the fault source model through inclusion of newly identified faults and revised activity rates on some faults; assessment of ground motion models through quantitative comparison against observations; and inclusion of a specific ground motion model for shaking from plate-boundary earthquakes in northern Australia. Expert elicitation was used to capture epistemic uncertainty surrounding model choices. The elicitation focused on decision points that sensitivity analysis had shown were more important for hazard, where new models had been developed, and where model choices had been controversial in NSHA18. Key questions included which catalogue to use as the basis for calculating hazard, the weighting of different source model classes (background, regional, seismotectonic, smoothed seismicity and smoothed seismicity with faults), and the selection and weighting of ground motion models for different tectonic regions. NSHA23 hazard results for capital cities show minor changes compared with NSHA18, with the exception of Darwin. Here the ground motion with a 10% probability of exceedance in 50 years increases significantly, a result that is attributed to inclusion of a new, more realistic ground motion model for plate-boundary earthquakes in this unique tectonic setting.</div><div><br>This paper was presented to the 2023 Australian Earthquake Engineering Conference 23-25 November 2023 (https://aees.org.au/aees-conference-2023/)</div>