2024 – The Calibration of LifeSim to the 2011 Grantham Life Loss Disaster

Fatima Azhar, Erin Hughes, Hench Wang, Chris Nielsen

The 2011 Grantham flood stands as one of Australia’s most catastrophic flood events in recent history which resulted in 12 fatalities. This flood event has seen various predictive methods applied to it to estimate potential loss of life (PLL). Despite their utility, these methods have only offered indicative estimates. LifeSim, an agent-based system developed by the US Army Corps of Engineers (USACE), simulates population redistribution during an evacuation to estimate PLL. Its enhanced simulation capabilities have made it a primary tool in Australia for assessing PLL due to flooding. However, LifeSim had not been calibrated to an Australian disaster involving flood fatalities until now. This study presents the calibration of LifeSim to the 2011 Grantham floods, assessing its suitability for the Australian context.

The Grantham Floods Commission of Inquiry (2015) provided the critical data for this assessment, including a calibrated hydraulic model, reports and witness statements detailing the events leading up to the Township’s inundation. Key aspects such as structure stability thresholds and the survival rate of residents were critical in building the LifeSim model. The calibration recommended adopting the structure stability criteria from the 2014 University of NSW Water Research Laboratory Report (Smith et al. 2014), as it was based on Australian data and more accurately reflected the destruction observed in Grantham. A new fatality rate curve was prepared using the data for vehicle-related flood fatalities in Australia. This curve was proposed to be used while estimating the loss of life on roads. The model results showed a good match with the actual number of fatalities in Grantham.

This calibration reinforced the suitability of LifeSim for use in the Australian context and provided valuable insights which lead to improvements in the state of practice of LifeSim applications not only in Queensland but elsewhere. It also highlighted the importance of critically questioning default parameter values in LifeSim when alternative location specific information is available.