Severe TC Vance was one of the most intense cyclones to impact mainland Australia. The observed damage to buildings could be explained in terms of structural performance of those buildings. Combining the structural vulnerability of housing with an estimate of the maximum wind gusts, we can explore the possible impacts that a repeat of Vance would cause in Exmouth, and compare the outcomes with what occurred in 1999. The analysis of the impacts of TC Vance on present-day Exmouth shows that very few houses would be completely destroyed. Not surprisingly, older houses (pre-1980’s construction era, excluding the US Navy block houses) would dominate those destroyed, and most likely the timber-framed style houses, many of which were substantially damaged in TC Vance.

Limited data from emergency services for the April 2015 East Coast Low (ECL) event initially investigated. SES call-out data provides spatial coverage, but does not capture detail of the damage to buildings. EICU data has detailed information, including indicative damage state, but limited spatial coverage. Neither dataset consistently links the damage to the hazard that caused it. Showing that the impact forecasting process adds value beyond the underlying hazard forecasts in this situation is challenging. EICU data can help to calibrate the vulnerability functions applied to model-based hazard forecast data. The SES callout data can help evaluate whether the indicative damage rates for an area are reasonable, through use of a service demand metric. Service demand is the number of callouts compared to the number of buildings for a statistical area (e.g. mesh block, SA1 or local government area). We use the total building count in each area, as the SES callout data does not differentiate between residential and non-residential buildings. It also includes callouts for downed trees or power lines that may not have directly caused structural damage to buildings. Service demand is compared to mesh block-based impact forecast data for the 2015 ECL, using existing heuristic vulnerability functions for severe wind. We recognise these functions are not calibrated against forecast model data, but provide a starting point from which we can establish the workflow while working towards refined vulnerability functions in parallel. The project has sourced EICU and SES post-event survey data, and high-resolution model (reanalysis) data for two additional severe wind and rain events to improve the calibration of the vulnerability functions. Poster presentation at the 2019 AFAC Conference