A long-term view of tropical cyclone risk in Australia

Natural hazard risk is assessed by leveraging, among other things, the historical record. However, if the record is short then there is the danger that risk models are not capturing the true envelope of natural variability. In the case of tropical cyclones in Australia, the most reliable observational record spans less than 50 years. Here, we use a much longer (ca. 6000-year) chronology of intense paleo-cyclones and, for the first time, blend this information with a catastrophe loss model to reassess tropical cyclone wind risk in Northeast Australia. Results suggests that the past several decades have been abnormally quiescent compared to the long-term mean (albeit with significant temporal variability). Category 5 cyclones made landfall within a section of the northeast coast of Australia almost five times more frequently, on average, over the late Holocene period than at present. If the physical environment were to revert to the long-term mean state, our modelling suggests that under the present-day exposure setting, insured losses in the area would rise by over 200%. While there remain limitations in incorporating paleoclimate data into a present-day view of risk, the value of paleoclimate data lies in contextualizing the present-day risk environment, rather than complementing it, and supporting worst-case disaster planning.

Automated summary

Natural hazard risk can be assessed using historical records, but short records may fail to capture the true range of variability. This study combines a 6000-year chronology of intense paleo-cyclones with a catastrophe loss model to reassess tropical cyclone wind risk in Northeast Australia. The findings indicate that the past few decades have been less active than the long-term average, and if conditions return to the long-term mean, insured losses would increase by over 200%.