Bushfire case study · New South Wales, Australia · 18 March 2018
The Tathra fire illustrates two problems in understanding a disaster: damage totals change as evidence improves, and a community’s understanding of its exposure can differ from the way fire actually spreads. This analysis separates the dated loss assessments from later research into residents’ responses.
Evidence reviewed: 27 October 2026. Documentary analysis of NSW Rural Fire Service releases, selected coronial findings and the published summary of commissioned community research. No property-level loss database or interview dataset was independently reanalysed.
What happened at Tathra?
The Reedy Swamp fire affected Tathra on 18 March 2018. The 2021 coronial findings place ignition at approximately 12:15–12:20 pm: a falling tree struck an electrical conductor, which fell and arced against the ground, igniting vegetation. The findings describe extensive spotting ahead of the main fire and fire activity extending into the coastal town. [3]
Why do sources report 69, 65 or 56 houses destroyed?
| Record | Destroyed houses | What the figure means |
|---|---|---|
| NSW RFS, 19 March 2018 | 69 | Initial assessment, with approximately 90% of the fireground assessed; revisions explicitly anticipated. [1] |
| NSW RFS, 21 March 2018 | 65 | Completed building impact assessment; closer inspection, including housing units and restricted-access areas, reduced the initial total. [2] |
| Later investigation described in the 2021 findings | 56 | Strike Force Butia calculation drawing on insurance, council and other documentation. [3, paragraph 26] |
The initial release also recorded 39 damaged houses and 30 destroyed caravans or cabins. [1] The completed RFS assessment reported 48 damaged houses and 35 destroyed caravans or cabins. Its category of 810 houses “saved or untouched” combines two outcomes; it cannot be presented as 810 individually demonstrated firefighting saves. [2]
Tecnobosque analysis: a comparison table should select and label one assessment consistently. Combining the later destroyed-house count with the earlier damaged-house count creates a composite that no cited assessment necessarily produced. The difference between 65 and 56 is not evidence that nine houses were rebuilt, nor does it justify silently deleting the contemporary RFS record.
Embers challenged assumptions about the town boundary
The NSW RFS commissioned research by the Bushfire and Natural Hazards CRC, undertaken with the University of Wollongong and Macquarie University. Its published summary reports that many residents had not expected fire to penetrate beyond the bush on the town’s western edge. It identifies a need for better understanding of ember spread into developed areas. [4]
This is a useful distinction for interpreting a map: a line marking the edge of vegetation is not, by itself, evidence of where exposure ends. Conversely, this case does not supply a universal distance at which every building is safe or unsafe. That would require a different, site-specific analysis.
Warning delivery and readiness are separate questions
The community research identifies warning clarity, power or technology failures and late evacuation as important issues. It also found that many people regarded preparation as an activity undertaken when directly threatened. Some late leavers said they would leave earlier in a future event. Those statements describe intended behaviour, not observed performance in a subsequent fire. [4]
For evaluating a warning system, three questions should therefore remain separate: was a message issued, did the person receive and understand it, and could they act on it? Counting issued messages alone cannot resolve all three. This is an analytical framework, not a reconstruction of each household’s decisions.
No fatalities does not mean no harm
The coronial findings record no deaths or serious injuries, while noting four patients treated for smoke inhalation and at least one with minor burns. [3, paragraph 25] The absence of fatalities should not erase displacement, property loss or the difficulty of decisions made during the event.
Scope and related cases
This record explains evidence and interpretation; it does not assign responsibility for individual losses or reproduce the inquiry’s full findings. Historical accounts are not current evacuation instructions. For contrasting warning geography and transport constraints, read the Waroona–Yarloop analysis and the Dunalley analysis, or browse the wildfire case-study archive.
Explore wildfire case studies from Australia →
Compare changing records and reconstructed timelines
Wangary 2005 preserves differing institutional totals where the sources reviewed do not establish a full reconciliation. That differs from Tathra’s explicitly dated assessment sequence. Roleystone–Kelmscott 2011 adds a separate distinction between logged fire presence and arrival of the main front.
Cite this analysis
Tecnobosque. Tathra Fire 2018: Changing Damage Counts, Embers and Community Response. https://tecnobosque.com/tathra-fire-2018/. Accessed [your access date].
For an incident figure or research finding, also cite its numbered source below. This is a live page; retain a copy of the passage and the date you accessed it.
Sources and verification
- NSW RFS: Initial assessment of fire affected area — Tathra, 19 March 2018.
- NSW RFS: Tathra area bush fire update, 21 March 2018, one-page PDF.
- NSW Coroners Court: Inquiry into the fire at Reedy Swamp, Tarraganda, Bega/Tathra, findings 17 December 2021; opening findings and paragraphs 22, 25–26.
- Whittaker and colleagues: Understanding bushfire risk, warnings and responses, 6 March 2020; commissioned report summary hosted by Macquarie University.