Worked examples
What official sources recorded near a place, together with every disaster declaration that covered its local government area. Five real locations, answered from the live data.
POST /api/v1/hazard-history/point, documented at the hazard history referenceEach row is one query. Declarations are counted once per AGRN, so an event three authorities recorded counts once rather than three times.
| Location | Declared disasters | Records nearby | Period recorded | Nearest gauge |
|---|---|---|---|---|
| Molesworth Street, Lismore Lismore, NSW | 27 since 2009-03-30 | 19 | 1988-04-07 to 2026-10-06 | Lismore (Wilsons River), 0.4 km |
| Great Western Highway, Blackheath Blue Mountains, NSW | 18 since 2011-07-05 | 278 | 1977-07-01 (first day of a season; no day stated) to 2026-10-03 | Penrith (Nepean River), 40.6 km |
| Oxlade Drive, New Farm Brisbane, QLD | 21 since 2008-11-16 | 2,508 | 1974-01-26 to 2026-10-08 | Brisbane City (Brisbane River), 1.3 km |
| Mundaring Weir Road, Mundaring Mundaring, WA | 3 since 2014-01-12 | 603 | 1937-07-01 (first day of a season; no day stated) to 2026-06-01 | John Forrest National Park (Jane Brook), 10.1 km |
| Anzac Highway, Glenelg East Holdfast Bay, SA | none | 462 | 2026-04-22 to 2026-09-20 | Frederick Street Drain at Glenelg, 1.0 km |
The case that proves the two lanes are both needed. The February 2022 flood put this street under water, and the observed lane holds no record of it at this precision. It is here through the declared lane, as AGRN 1012.
Mapped fire extents rather than points. The Grose Valley fire and the Black Summer boundaries both cover this location, and 16 of the 247 archive records carry no date.
What a large count really means. 2,452 of these 2,508 records sit at a postcode centroid, so they are near this postcode, not near this street. A count without its precision would badly mislead. The 16 mapped flood extents are different: they are the Brisbane River floods of 1974, January 2011 and February 2022 as the agencies drew them.
Depth. The record here runs back to the 1937 fire season, and fire management shows up as clearly as fire: 253 of the 603 records are planned burning, not emergencies.
The honest negative, and an honest gap. No disaster declaration has ever covered this local government area. 400 older road-crash records nearby count the crashes at one spot over several years and carry no date, so none is invented for them. The dated records are the 62 the live feeds have logged since April 2026.
Three things decide whether a count means what it appears to mean, and every response carries all three.
Some records are left out on purpose. Medical calls are left out for privacy. Alarm calls are left out because an alarm is not a hazard: the agency either found no fire or did not record the outcome, and a real fire is recorded as a fire. Road and traffic feeds are left out because they record roadworks, closures and traffic, not hazards.
Mapped areas stay in. A fire’s footprint or an official warning area is a record of what reached a place, often a better one than the incident’s own point, so one fire can be counted once for each source that recorded it.
Declarations are matched to the local government area containing the point, not to the address. A declaration with no published end date has not been shown to have ended, and that is not the same as being in force.