<div>Severe TC Ilsa crossed the Western Australian coastline approximately 120 km east of Port Hedland on Thursday 13 April 2023. Observations at Bedout Island were the highest wind speeds ever recorded on standard BoM instruments (gust wind speed of 289 km/h). In anticipation of the TC, residents in the mining township of Telfer were evacuated, along with a small number of evacuees in other townships (Marble Bar, South Hedland and Nullagine). As a category 5 TC, the threat of widespread destruction was front of mind for emergency managers in Western Australia.</div><div><br></div><div>Geoscience Australia (GA) has established the National Hazard Impact and Risk Service (NHIRS), which provides quantitative modelled impact forecast information for tropical cyclones, large-scale wind events and earthquakes in Australia. NHIRS has been used by the Department of Fire and Emergency Services (DFES) Intelligence Unit to support operational resource planning for TC events.</div><div><br></div><div>In TC Ilsa, DFES Intelligence (and GA) officers reviewed the impact predictions in the days leading up to landfall. Genuine questions were asked about the level of predicted damage, which was almost negligible across northern WA in spite of the predicted landfall intensity. Why was that the case? Was the service operating as expected? This paper highlights the challenge of educating users on the utility of impact forecasting products and communicating the components that are integrated in the impact forecast. Presented at the 30th Conference of the Australian Meteorological and Oceanographic Society (AMOS) 2024

Tropical cyclones present a tangible risk to Australia’s tropical coastal communities, however extratropical transition (ETT) of these storms can result in significant impacts in mid-latitude regions as well. Tropical systems are driven by latent heat release in the inner core of the cyclone. A fully tropical system is highly axisymmetric; with a warm-cored vortex that is readily represented by a simple radial profile (wind speed is a function of distance from the centre in all directions). Extratropical cyclones on the other hand are driven by strong thermal gradients and as a result have a highly asymmetric wind field that cannot be as easily parameterised for use in stochastic models. In order to accurately model the risk of these transitioning storms on communities such as Perth, the wind field of these storms needs to be parameterised for inclusion in stochastic models. These models allow large numbers of storms to be quickly simulated for use in risk modelling applications. Some authors have attempted to develop parameterisations of these wind fields, with some recent success (Loridan et al. 2015), however an implementation for the Australian region has not yet been developed. Geoscience Australia currently undertakes tropical cyclone risk assessments using a parameterised, 2D stochastic model called the Tropical Cyclone Risk Model (TCRM). TCRM uses parameterised wind fields to allow quick generation of thousands of tropical cyclones in order to develop a probabilistic understanding of tropical cyclone risk for Australia. At present, this model is not capable of simulating tropical cyclones undergoing ETT as a parameterisation of the wind field of these storms around Australia is not available. This work aims to explore ETT around Australia using a 3D, dynamical numerical weather prediction model with the ultimate goal of developing a parameterised wind field, suitable for inclusion in TCRM. This would allow risk assessments for these storms to be undertaken, and improve our understanding of the potential impact of such an event on large urban areas, such as Geraldton or Perth. A modified version of the Weather Research and Forecast (WRF) model (Hybrid WRF) was used to simulate a number of hybrid idealised tropical cyclones, and steer them to undergo ETT. Hybrid WRF was developed to facilitate control over the track and location of landfall of a tropical cyclone, by introducing a steering flow to the boundary conditions of the model run. This method was used to steer a number of idealised tropical cyclones from off the northwest coast of Western Australia, south towards Perth, with the intent to force them to undergo ETT. Surface wind fields and other environmental characteristics (minimum pressure, latitude, thermal wind components, geopotential thickness and others) were analysed to determine the phase of ETT. This case study is the first example of Hybrid WRF being used to examine ETT, and while the steering flow did move the tropical cyclones into the extratropics as intended, only one storm was observed to undergo ETT. Further development of the code for Hybrid WRF is underway, with improvements in the initial and boundary conditions identified as a means to improve the representativeness of these experiments. Based on these simulated events, we intend to develop time-evolving, storm-centred wind fields, as well as statistics on cyclone phase space parameters that can be used to determine the stage of transition to be used in a future stochastic-parametric model of tropical cyclones. Abstract submitted to/presented at the 22nd International Congress on Modelling and Simulation 2017 (MODSIM2017) - https://www.mssanz.org.au/modsim2017/

Tropical Cyclone (TC) 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. If TC Tracy were to strike Darwin in 2008, the impacts could be catastrophic. We perform a validation of Geoscience Australia's Tropical Cyclone Risk Model (TCRM) to assess the impacts TC Tracy would have on the 1974 landscape of Darwin, and compare the impacts to those determined from a post-impact survey. We then apply TCRM to the present-day landscape of Darwin to determine the damage incurred if a cyclone identical to TC Tracy impacted the city in 2008. In validating TCRM against the 1974 impact, we find an underestimate of the damage at 36% of replacement cost (RC), compared the survey estimate of 50-60% RC. Some of this deficit can be accounted for through the effects of large debris. Qualitatively, TCRM can spatially replicate the damage inflicted on Darwin by the small cyclone. The northern suburbs suffer the greatest damage, in line with the historical observations. For the 2008 scenario, TCRM indicates a nearly 90% reduction in the overall loss (% RC) over the Darwin region. Once again, the spatial nature of the damage is captured well, with the greatest damage incurred close to the eye of the cyclone. Areas that have been developed since 1974 such as Palmerston suffer very little damage due to the small extent of the severe winds. The northern suburbs, rebuilt in the years following TC Tracy, are much more resilient, largely due to the influence of very high building standards put in place between 1975 and 1980. Article published in the Australian Journal of Emergency Management

We present the formulation of an open-source, statistical–parametric model of tropical cyclones (TCs) for use in hazard and risk assessment applications. The model derives statistical relations for TC behaviour (genesis rate and location, intensity, speed and direction of translation) from best-track datasets, then uses these relations to create a synthetic catalogue based on stochastic sampling, representing many thousands of years of activity. A parametric wind field, based on radial profiles and boundary layer models, is applied to each event in the catalogue that is then used to fit extreme-value distributions for evaluation of return period wind speeds. We demonstrate the capability of the model to replicate observed behaviour of TCs, including coastal landfall rates which are of significant importance for risk assessments. <b>Citation: </b>Arthur, W. C.: A statistical–parametric model of tropical cyclones for hazard assessment, Nat. Hazards Earth Syst. Sci., 21, 893–916, https://doi.org/10.5194/nhess-21-893-2021, 2021.

The first step in understanding risk is understanding the hazard. This means knowing the likelihood of the hazard event and its intensity. During 2018, Geoscience Australia updated the Tropical Cyclone Hazard Assessment (TCHA) to better calculate the likelihood of tropical cyclones in Australia.

This dataset contains a collection of ESRI geodatabases that hold hazard and impact data derived as part of the Severe Wind Hazard Assessment for Western Australia (2017-2020) project. There are separate geodatabases for each community examined in the project. Within each community, multiple TC scenarios were analysed for each community. The list of scenarios is included below. Geodatabase structure --------------------- Within each geodatabase, the data is structured as set out below. The structure is repeated for each available scenario in that community. Note scenario id numbers have the hyphen ('-') removed in the <scenario id> string below. - Shapefiles |-- TCs within 50 km |-- Cat<X> <scenario id>_Impact [Polygon shape file of SA1-level mean damage state for residential housing] |-- Cat<X> <scenario id>_regionalwind [Polygon shape file of categorised regional wind speed] |-- Cat<X> <scenario id>_track_line [Line shape file of scenario track line segments] |-- Cat<X> <scenario id>_track_point [Point shape file of scenario track points] - Cat<X>_<scenario id>_localwind [Raster format local wind data] Scenarios --------- Scenairo Id number, TC intensity, Location 000-01322,3,Exmouth 013-00928,3,Exmouth 000-06481,5,Exmouth 003-03693,3,PortHedland 000-08534,5,PortHedland 012-06287,3,Broome 012-03435,5,Broome 006-00850,3,Karratha-Roebourne 009-07603,5,Karratha-Roebourne 011-01345,1,Carnarvon 003-05947,3,Carnarvon 011-02754,1,Geraldton 001-08611,3,Geraldton 007-05186,1,Perth bsh291978,1,Perth

<div>Severe wind from tropical cyclones (TC) can cause significant damage to property and infrastructure, and accurately predicting these impacts is essential for ensuring community safety. At Geoscience Australia (GA), the Tropical Cyclone Risk Model (TCRM) is a statistical-parametric tool designed to estimate the severe wind hazards posed by TCs.</div><div><br></div><div>To assess the performance of TCRM, all TCs impacting Australia from 1998 to 2023 were simulated using various wind profile settings. Since TCRM currently lacks adjustments for local landscape effects on wind speed (e.g., topographic enhancements or variations due to vegetation or built environments), GA’s wind multiplier was applied to convert TCRM’s regional wind to local wind for each TC. Additionally, a new multiplier, derived from the Parallelized Large-Eddy Simulation Model (PALM), was used to calculate local wind. The local winds from these two multipliers were then compared and validated against 1-minute Automatic Weather Station (AWS) observations from the Bureau of Meteorology (BoM). </div><div><br></div><div>The results indicate that the Willoughby wind profile outperforms other wind profiles, with a mean sea level pressure error of just 0.2 hPa compared to AWS observations over the mainland. Although the local wind from both multipliers aligns closely across 82 AWS stations (a mean peak wind gust error of -1 m/s for the GA multiplier and 0.2 m/s for the new multiplier), significant discrepancies may occur in regions with steep mountain ranges, urban areas with high-rise buildings, and other areas with complex topography.</div><div><br></div>

<div>Tropical cyclones (TCs) frequently encompass multiple hazards, including extreme winds, intense rainfall, storm surges, flooding, lightning, and tornadoes. Accurate methods for forecasting TC tracks are essential to mitigate the loss of life and property associated with these hazards. Despite significant advancements, accurately forecasting the paths of TCs remains a challenge, particularly when they interact with complex land features, weaken into remnants after landfall, or are influenced by abnormal satellite observations. To address these challenges, we propose a generative adversarial network (GAN) model with a multi-scale architecture that processes input data at four distinct resolution levels. The model is designed to handle diverse inputs, including satellite cloud imagery, vorticity, wind speed, and geopotential height, and features an advanced center detection algorithm to ensure precise TC center identification. Our model demonstrates robustness during testing, accurately predicting TC paths over both ocean and land, while also identifying weak TC remnants. Compared to other deep learning approaches, our method achieves superior detection accuracy, with an average error of 41.0 km for all landfalling TCs in Australia from 2015 to 2020. Notably, for five TCs with abnormal satellite observations, our model maintains high accuracy with a prediction error of 35.2 km, a scenario often overlooked by other approaches. <b>Citation:</b> Huang, H.; Deng, D.; Hu, L.; Sun, N. Anomaly-Aware Tropical Cyclone Track Prediction Using Multi-Scale Generative Adversarial Networks. Remote Sens. 2025, 17, 583. https://doi.org/10.3390/rs17040583

<div>The Severe Wind Hazard Assessment for South East Queensland (SWHA-SEQ) analysed risk from severe wind events in a marginal tropical cyclone (TC) region with a large exposed population, and historical severe thunderstorm and TC impacts. SWHA-SEQ was a collaborative effort bringing together 15 partners across government, academia and the insurance sector to improve the collective understanding of wind risk in the region and inform future strategies to reduce this risk, in the context of climate change, urban planning and socio-economic status of the population. </div><div>The project involved enhancing the understanding of hazard, exposure and physical vulnerability to strengthen the comprehension of risk, including local-scale wind hazard from thunderstorm and TC wind gusts, and a semi-quantitative analysis of future wind hazard. Structural characteristics of residential housing stock were updated through a combination of street surveys, national databases of built assets and insurance portfolio statistics. Vulnerability models for residential houses including retrofitted models for 5 common house types were developed, alongside identification of key vulnerability factors for residential strata buildings.</div><div>Local governments are building on the outcomes of the project, with the City of Gold Coast using the project outcomes as the key evidence base for a A$100m investment over 7 years to advocate for uplift of building design criteria, targeted community engagement and resilience of City-owned infrastructure. Other local governments have conducted specific exercises exploring how they would manage a severe TC impact. The investments and activities directly flowing from SWHA-SEQ are testament to the partner engagement through the project. Presented at the 2024 Symposium on Hurricane Risk in a Changing Climate (SHRCC2024)

<div>Tropical cyclone wind hazard across Queensland and the Coral Sea was evaluated using Geoscience Australia's Tropical Cyclone Risk Model (TCRM), which provides a spatial representation of the AEP wind speeds arising from tropical cyclones, as part of the <em>Severe Wind Hazard Assessment for South East Queensland</em> (SWHA-SEQ) project. The project is a collaboration between Queensland Fire and Emergency Services (QFES), Geoscience Australia, James Cook University (JCU) and six coastal local governments in South East Queensland.</div><div><br></div><div>This regional wind hazard data underpins a local-scale risk assessment for South East Queensland, which included the effects of synoptic and severe thunderstorms on wind hazard. The assessment aslo provides a baseline for projections of future TC wind hazard in the Queensland region. Details of the hazard assessment are provided in the <em>Severe Wind Hazard Assessment for South East Queensland – SWHA-SEQ Technical Report</em> (https://pid.geoscience.gov.au/dataset/ga/147446).&nbsp;</div><div><br></div>