How we update

From the publisher to this site

1. Our server checks each connected agency source on its own schedule.
2. A response is validated before it replaces that source’s stored snapshot.
3. The website reads those stored records; opening or filtering the map does not trigger a new request to the agency.
4. An open map or incident page normally checks our server for updates every 30 seconds.

The 30-second page refresh is not a promise that agency information arrives every 30 seconds. Publisher schedules, our collection interval and outages can delay what appears here. The sources page shows the configured interval and last successful collection for each source.

Different timestamps mean different things

An agency update time is the timestamp supplied for a report, where we can interpret it reliably. A collection time is when our system fetched or validated source data. A recent collection does not mean the agency has recently updated every report.

Incident pages display times in the relevant state or territory’s local timezone. We do not invent an update time when the publisher supplies an ambiguous or unrecognised value. Change histories show when a change was first observed by us, rather than claiming the exact time the agency issued it.

Selection and map locations

We search and filter the eligible records before selecting up to 250 for the map and list. Explicit warning levels take priority, followed by report type and agency update time. Planned works and routine records appear behind higher-priority reports. Zooming or panning loads matching reports for the visible area; a text search searches beyond that area.

Nearby symbols can be grouped into numbered clusters. The number counts the selected records in that cluster, not every report in the region or distinct emergencies. Reports without usable geometry stay in the list.

The time filter uses agency update times. Current Emergency Warning and Watch and Act records remain eligible regardless of their update age, unless you are browsing retained records. They must still match your other filters and the display limit.

Points and areas come from publishers and can be approximate. We do not calculate fire spread, evacuation boundaries or safety based on proximity.

Failures, missing records and expiry

If the latest source check fails or its response cannot be validated, we preserve the last successful data. A source failure icon beside the publishing source explains the situation and identifies the last successful collection. A source may also become delayed even without a failed attempt; check its status on the sources page.

A record missing from a successful complete snapshot is retained in storage and excluded from the normal current view. Publisher cancellations and expired warnings are also excluded from that view. Existing detail pages may remain accessible with a retained-information notice. None of these states, by itself, confirms that an incident has ended or an area is safe.

Warning levels and corrections

Emergency Warning, Watch and Act and Advice labels are based on the publisher’s supplied classification. Unknown or unrelated categories are not promoted to an emergency warning. Display priority does not create a new warning level. Red markers represent the publisher’s Emergency Warning classification.

If you find a problem in how we display information, please send a correction request with the page URL and an official reference.

Historical records and public archive periods

An archive label means the record is no longer presented as current. It does not establish that an incident has ended or an area is safe. Disappearance from a successful complete source snapshot, a publisher cancellation and publisher expiry are labelled separately. A failed source check does not, by itself, move a record into the archive.

• Reports that reached Emergency Warning or Watch and Act are kept as historical records, including their observed warning history. They can remain discoverable in search when the available source facts are sufficient.
• Other significant incidents remain accessible as history. They can stay in search for the first 90 days of their archive period; longer-term search visibility requires sufficient retained publisher narrative or instructions.
• Routine incidents, minor traffic reports, controlled burns and planned roadworks remain publicly accessible for 90 days after entering the archive, with their archived details excluded from search.
• Routine ratings, bans and short updates remain publicly accessible for 30 days after entering the archive, with their archived details excluded from search.

After a routine record’s public period expires, its URL returns a “no longer published” response (HTTP 410) with a link to current regional reports. We do not automatically redirect it to the home page. Retiring a public page does not delete the stored report or captured revisions.

The archive clock uses the first observed archive transition or an earlier publisher expiry time. Repeated missing-source snapshots do not restart it. If a record reappears as current, its current status is restored while its historical maximum warning level is preserved.