Pricing Search Start free trial
// EVENT · TRAFFIC DISRUPTION NO DISRUPTION · brief closed 7 Sep 2026 — the −65% flag was Perth’s weekly Saturday trough: thirteen measured days followed it and eleven landed within about 4% of their own same-weekday norm, the closing working week ran 608/698/648/648 Mon–Thu, and no corroborating signal ever appeared in any layer Data as of 7 Sep 2026, 15:52 UTC

Perth (PER) Traffic ‘Drop’ — a −65% Anomaly That Was the Airport’s Normal Saturday, August 2026

← All airspace events

Perth International (PER) tripped our traffic-anomaly detection on Saturday 22 August at −65% against a ~506 seven-day rolling average — a flag fired on a preliminary count of 178, since settled at 190 observed movements, 62% below that average. The cross-checks said this was not a disruption, and sixteen days of follow-up confirmed it, so we are closing this brief. 22 August was a Saturday, and Perth sheds roughly 64% of its weekday traffic every Saturday by rhythm. Against its Saturday baseline (~216 movements across the four prior Saturdays: 209, 233, 199, 223), the day measured 12% below — a normal weekend low, inside the band our detector scores normal, on a balanced ledger of 91 arrivals against 99 departures rather than the one-sided departure collapse that marks a receiver-coverage failure. Thirteen measured days followed the flag, and eleven landed within about 4% of their own same-weekday norms. The closing working week is the clearest evidence: 608 movements on Monday 31 August (~604 norm), 698 on Tuesday 1 September (~682), 648 on Wednesday 2 September (~664) and 648 on Thursday 3 September (~670) — four consecutive ordinary working days, each two-sided and within about 3.5% of its own norm. The last Saturday of the window, 5 September, came in at 179 (88 departures / 91 arrivals) — 17% below the ~216 baseline, −15% against the 211-movement Saturday baseline our own detector uses, and scored normal by it: the third Saturday in a row inside Perth’s trough. The one caveat we carried last time has reversed: Wednesday 2 September, which we published at 179 and explicitly refused to read as an airport event, has re-settled at 648 (383 departures / 265 arrivals), dead on its Wednesday norm — the sixth day on this page to move, and the sixth time the caution was the right call. 4 and 6 September are network-wide collection gaps on our side, not measurements of Perth, and are excluded. No closure signal ever registered for the airport or its FIR (YPIY), the weather station never broke its reporting cadence (latest observation 7 Sep, 15:00 UTC: CAVOK, wind from the east-south-east at 13 kt, 17 °C, no thunderstorm or gust flags), GNSS interference sat at its measurement floor throughout, and no volcanic-ash exposure was ever in the frame. We publish this as the anatomy of a day-of-week seasonality false alarm, honestly labeled — because the same raw −65% at a major airport is exactly the kind of number that gets misread as a crisis.
// Flying through Perth? The short version
  • Perth airport was operating normally throughout — the −65% ‘drop’ on 22 August is what a Saturday looks like at PER when it is compared against a weekday-heavy weekly average. This brief is closed. The thirteen measured days since confirmed it, ending with a clean working week: 608 observed movements on Monday 31 August, 698 on Tuesday 1 September, 648 on Wednesday 2 September and 648 on Thursday 3 September, each within about 3.5% of its own same-weekday norm.
  • Perth runs one of the deepest weekly traffic rhythms we measure in Australia: ~600 observed movements on a typical Monday–Friday, ~216 on a Saturday — roughly a 64% weekend trough, week after week. All three Saturdays we measured sit inside it: 190 on the flagged 22 August, 211 on 29 August and 179 on 5 September, the last of them scored normal by our own detector.
  • The flagged count was a balanced 91 arrivals and 99 departures — real, evenly split traffic, not the one-sided coverage loss we flag elsewhere — and no closure, weather or interference signal ever corroborated anything unusual. Our record on this page kept correcting itself in both directions, and the last correction reversed our own caution: 2 September, which we published at 179 and drew no conclusion from, has re-settled at 648. 4 and 6 September are collection gaps on our side and are excluded, not counted as quiet days. For a specific flight, check your airline as always.

Advisory timeline — how active is this episode

Perth’s week has a shape you can see with the naked eye. Monday through Thursday the network observes ~600–690 movements a day; Friday steps down to ~450; Sunday runs in the high-300s — and then Saturday is the trough at ~216, the quietest day of the week. The flagged day sits squarely in it: 190 on 22 August, and so do the two Saturdays that followed, 211 on 29 August and 179 on 5 September. The weekdays between them close the argument: Sunday 23 August measured 365, Monday 24 August 597, Tuesday 25 August 667, Thursday 27 August 668 and Friday 28 August 457 — each squarely on its own same-weekday norm (~381 Sundays, ~604 Mondays, ~682 Tuesdays, ~670 Thursdays, ~450 Fridays) — and the closing working week runs the same way: 608 on Monday 31 August, 698 on Tuesday 1 September, 648 on Wednesday 2 September and 648 on Thursday 3 September, every one of them within about 3.5% of its own norm. The tail of the window moved once more since our last update, and this time both moves ran toward normal: Wednesday 2 September — which we published here at 179 and explicitly declined to read as an airport event — has re-settled at 648 (383 departures / 265 arrivals), about 2% under the ~664 Wednesday norm, and Thursday 3 September has firmed from 585 to 648 (391 dep / 257 arr), about 3% under the ~670 Thursday norm. That is the sixth time on this page a fresh day has moved, and the sixth time the number settled closer to Perth’s ordinary rhythm than the provisional one suggested. Two bars are starred because they are not measurements of Perth at all: 4 September, when network-wide observed volume ran at roughly 50% of its 14-day median — below the 60% floor at which we stop treating a day as measured — leaving a stored count of 49 (31 dep / 18 arr); and 6 September, a near-total collection gap at roughly 8% of median, whose stored 70 is one-sided at 68 departures against 2 arrivals. Both are kept visible and excluded from every baseline on this page. One softer day we do keep — 453 on Wednesday 26 August — is a day we measured but do not read as a traffic event: arrivals came in close to their same-weekday norm while departures fell well short (208 against ~397), the lopsided shape of partial capture rather than a real drop. Same-weekday history is the honest baseline for an airport with a weekly pulse this deep:

190
22 Aug
365
23 Aug
597
24 Aug
667
25 Aug
453
26 Aug
668
27 Aug
457
28 Aug
211
29 Aug
381
30 Aug
608
31 Aug
698
1 Sep
648
2 Sep
648
3 Sep
49*
4 Sep
179
5 Sep
70*
6 Sep

Fig. 1 — Observed daily movements at Perth International (PER), 22 Aug – 6 Sep 2026 (4 and 6 Sep: network-wide collection gaps on our side — undercounts, excluded from every baseline) · * partial day · Data as of 7 Sep 2026, 15:52 UTC · Source: FlySafe monitoring (crowdsourced ADS-B receiver network).

Airports in the exposure zone

AirportSizeObserved signalFIR
Perth International (PER) large — · Sat 22 Aug: 190 observed vs ~216 Saturday baseline — balanced 91 arr / 99 dep · Sat 29 Aug: 211 (97 arr / 114 dep) · Sat 5 Sep: 179 (91 arr / 88 dep) — scored normal · closing working week 31 Aug–3 Sep: 608 / 698 / 648 / 648, each within ~3.5% of its own norm YMMM

Distance alone does not determine impact — wind direction and the flight levels the ash reaches decide which airports and routes are affected on a given day. FIR = Flight Information Region.

What our monitoring showed: a weekly saw-tooth, and Saturday is the trough

FlySafe tracks observed departures and arrivals at 493 airports from crowdsourced ADS-B receiver data. Observed counts are a floor, not a total — receiver coverage varies — but an airport measured against its own baseline is a reliable shape. Perth International's last 30 days:

Fig. 2 — Observed daily movements at PER, 7 Aug – 5 Sept (latest complete UTC day). Data as of 7 Sep 2026, 15:52 UTC · Source: FlySafe monitoring (ADS-B receiver network).

  • Monday–Friday across the 30-day window charted above: ~600 observed movements/day on average, peaking around 690 on a mid-week Tuesday–Thursday and stepping down to ~450 on Fridays.
  • Saturdays: the four that preceded the flagged day measured 209, 233, 199 and 223 — an average of ~216, a ~64% trough that repeats every week. Sundays sit higher, in the high-300s.
  • Saturday 22 Aug: 190 observed movements — 12% below the ~216 Saturday baseline, the lowest of the recent Saturdays yet still plainly a Saturday, far from the ~600 weekday plateau. Against the 223-movement median of Perth’s other recent Saturdays — the baseline our own detector uses — it read −15%, inside the band we score normal.
  • The ledger balances: the 190 is 91 arrivals against 99 departures — an even, two-sided count. That is the decisive difference from a coverage-loss anomaly: when receivers fail, departures vanish from the ledger first (as at Palermo and Monterrey). Here both sides were present and roughly equal — real, evenly distributed light-Saturday traffic, not a measurement hole.
  • What tripped the detector: the day was flagged at −65% against Perth’s ~506 seven-day rolling average — fired on a preliminary count of 178 that settled at 190, or 62% below that same average. Either way the comparison was the problem, not the airport: that rolling average blends five busy weekdays and a Sunday with the one quiet Saturday, so measuring a Saturday against it simply reproduces the ~64% trough that is baked into Perth’s week.
  • Three Saturdays, one trough. 29 August measured 211 movements — 97 arrivals against 114 departures on a day when network-wide collection ran at ~98% of its 14-day median, 2% below the ~216 baseline. 5 September, the closing Saturday, measured 179 — 91 arrivals against 88 departures, the most evenly balanced ledger of the three, on a day when network-wide collection ran at ~94% of median. That is 17% below the ~216 baseline and −15% against the 211-movement Saturday baseline our own airport detector uses — a reading it scores normal. Three Saturdays measured since the flag fired, all three inside Perth’s weekly trough.
  • The working week came back on schedule — the check that would have overturned this page, run thirteen times. Thirteen measured days follow the flag. Sunday 23 August settled at 365 observed movements (249 departures / 116 arrivals) against a ~381 Sunday norm; Monday 24 August at 597 (369 dep / 228 arr) against ~604; Tuesday 25 August at 667 (399 dep / 268 arr) against a ~682 Tuesday norm; Thursday 27 August at 668 (382 dep / 286 arr) against a ~670 Thursday norm — dead on it; Friday 28 August at 457 (239 dep / 218 arr) against a ~450 Friday norm; Saturday 29 August at 211 (114 dep / 97 arr) against the ~216 Saturday baseline; Sunday 30 August at 381 (256 dep / 125 arr) against the ~381 Sunday norm, exactly on it. The closing working week is the cleanest stretch of the window: Monday 31 August 608 (363 dep / 245 arr) against ~604, Tuesday 1 September 698 (426 dep / 272 arr) against ~682, Wednesday 2 September 648 (383 dep / 265 arr) against ~664, and Thursday 3 September 648 (391 dep / 257 arr) against ~670 — four consecutive working days, each within about 3.5% of its own same-weekday norm and each with both halves of the ledger present. Eleven of the thirteen measured days landed within about 4% of their own same-weekday norm. The other two are Wednesday 26 August (treated below) and Saturday 5 September, 17% under the Saturday baseline and scored normal by our own detector. A quiet Saturday followed by flat weekdays would have been the signature of disruption; we measured thirteen days and never found it.
  • The caveat we carried last time has reversed — in our favour, which is not the point. Our previous update stored Wednesday 2 September at 179 movements (130 dep / 49 arr), ~73% under the ~664 Wednesday norm and by far the largest gap in the window. We reported that number, left it in the chart and in the averages, and drew no conclusion from it — because it had been published at 586 the day before, because the shortfall was lopsided toward arrivals, and because network-wide collection that day ran at ~65% of median. It has since re-settled at 648 (383 dep / 265 arr), about 2% under the Wednesday norm with both halves of the ledger on their own norms (~397 dep, ~267 arr). Thursday 3 September moved the same way, from a published 585 to 648 (391 dep / 257 arr), about 3% under ~670. Both are now ordinary Perth working days. We are recording this the way we recorded the number that made us cautious.
  • One soft day is measured but not trusted as traffic: Wednesday 26 August. It settled at 453 movements against a ~664 average across the window’s other Wednesdays, on a day when network-wide collection ran at roughly 96% of its 14-day median — well above the 60% floor at which we stop treating a day as measured — so we keep it in the chart and in every average rather than quietly dropping an inconvenient number. But its composition argues against reading a traffic event into it: arrivals came in at 245, only ~8% under the ~267 Wednesday norm, while departures fell to 208, ~48% under the ~397 norm. Arrivals near-normal with departures short is the fingerprint of partial capture, not of an airport that stopped flying. Keeping it in drags our averages slightly low, which is the conservative direction.
  • Two collection gaps at close: 4 and 6 September. On 4 September network-wide observed volume ran at roughly 50% of its 14-day median — below our 60% floor — and Perth’s stored reading for it is 49 movements (31 dep / 18 arr); the count has already moved once, from the 69 we published last time. On 6 September collection collapsed to roughly 8% of median, and the stored 70 is one-sided at 67–68 departures against 2 arrivals — the shape of a data hole. Two independent layers say Perth flew on both days: its weather station filed 49 observations on 4 September and 50 on 6 September, on its normal cadence, and the aircraft-telemetry feed behind our GNSS layer observed 2,077 aircraft over Perth’s FIR on 4 September and 1,624 on 6 September. Both bars are starred in the chart above and excluded from every baseline on this page. They are not evidence about Perth, and on the record of this page they may re-settle.
  • Freshness note — the corrections, which ran in both directions six times. The 22 August count firmed up over the nightly refetches, from a preliminary 178 to a settled 190, and it stayed a normal Saturday. 28 August went out, then came back: first published as a settled Friday of 436, overwritten down to 31 by a refetch inside a degraded window and published as “not measured”, then re-settled at 457 (239 dep / 218 arr). 29 August followed the identical arc: published at 196, overwritten to 15, withdrawn as “not measured” — and settled at 211 (114 dep / 97 arr). 30 August, published at 316 and discounted as one-sided, re-settled at 381 (256 dep / 125 arr) — caveat withdrawn. 31 August ran the arc twice: published as a gap at 115, then at 499 and called ~17% light, finally at 608 and dead on its Monday norm. 1 September ran it three times: 621, then 285 inside a thin window — where we reported the number but drew no conclusion from it — and now 698, its highest Tuesday in the window. 3 September went from a starred 133 to 585 to 648. And 2 September moved down and then back: published at 586, overwritten to 179 after a refetch inside a ~65%-of-median window, now settled at 648. The standing lesson, demonstrated repeatedly in both directions: our newest days are provisional, a refetch can take a day back and give it back, and this page corrected itself either way — which is why we published the number that was in our data on the day rather than the one that suited the story. It applies to the closing numbers too: 5 September is two days old at close and is provisional in exactly the same way.

Observed counts are a floor, not a total — but an airport measured against its own same-weekday history is a reliable shape, and Perth’s Saturday is one of the deepest weekly troughs we measure in Australia. The last day that is both measured and past our collection floor is Saturday 5 September at 179 (91 arr / 88 dep), inside the same weekly trough that produced the original flag and scored normal by our own airport detector. This brief is closed; the live view keeps updating: PER airport brief.

What it means for flights

What fired, and why we published it anyway. Our detection layer flagged PER when the 22 August observed count came in at −65% against a ~506 seven-day rolling average — a traffic anomaly at Western Australia’s principal airport, fired on a preliminary count of 178 that settled at 190 movements. Automated multi-signal analysis then cross-checked the anomaly against every other layer we track: closure signals, weather-reporting continuity, volcanic-ash exposure, GNSS interference, and the composition and same-weekday history of the count itself. Nothing corroborated a disruption — and the same-weekday history explained the entire signal. We published the result anyway, for the same reason we published the Memphis Sunday false alarm and the Monterrey coverage collapse: a raw −65% at a well-known airport is exactly the kind of number that becomes a phantom headline, and showing why it is not one is part of publishing observed counts honestly.

One of the deepest weekly rhythms we measure in Australia. Perth’s observed traffic runs a pronounced weekly cycle: ~600 observed movements on an average Monday–Friday, against ~216 on an average Saturday — roughly a 64% weekend trough, repeating week after week (the four Saturdays preceding the flag measured 209, 233, 199 and 223; the two after it, 29 August and 5 September, measured 211 and 179). Fridays step down to ~450 ahead of it; Sundays recover into the high-300s; Saturday is the quietest day of Perth’s week. A shape this deep is consistent with Perth’s role as the gateway to Western Australia’s resources sector — weekday-heavy fly-in/fly-out charter and business travel, much lighter on the weekend, with goldfields and mine-site routes (Kalgoorlie–Boulder among its busiest connections) — though we label that mechanism as interpretation, not measurement. What is measured is the rhythm itself: a detector comparing any Saturday against an all-days average will ‘find’ a collapse, which is a statement about the baseline, not the airport.

A ledger that balances — why this was not the Palermo or Monterrey failure mode. The decisive cross-check is the composition of the count. Palermo’s and Monterrey’s anomalies were one-sided: arrivals normal, departures gone — the fingerprint of receiver-coverage loss, because a departing aircraft is heard first on the runway and in initial climb, the shortest-range signal there is, so departures vanish from the ledger first when close-in receivers fail. Perth’s 190 was nothing of the kind: 91 arrivals against 99 departures, an even two-sided day — and so were the Saturdays that followed it, 29 August at 97 arrivals against 114 departures and 5 September at 91 against 88. The recovery weekdays are just as even: Friday 28 August closed at 239 departures against 218 arrivals on a ~450 Friday norm, Sunday 30 August at 256 against 125 on a ~381 norm, Monday 31 August at 363 against 245, Tuesday 1 September at 426 against 272, Wednesday 2 September at 383 against 265 and Thursday 3 September at 391 against 257 — every one of them with both halves on or near their own same-weekday norms (~397 dep / ~267 arr on Wednesdays, ~390 / ~280 on Thursdays). Both sides of the ledger present and roughly equal is what real traffic looks like and what a coverage hole does not. The same test is what tells us to discount Wednesday 26 August, where arrivals held on their norm while departures fell ~48% short, and it is also what marked the 2 September reading we carried last time as suspect: its 179 was an arrival-side collapse (49 against a ~267 norm) inside our thinnest collection window of the fortnight, and it has since re-settled at 648 with both halves normal. 4 and 6 September are withheld entirely — stored at 31 departures against 18 arrivals and 68 against 2, inside network-wide collection gaps.

The independent checks we ran, and what they said at close. No airspace-closure signal ever registered for Perth or its FIR (YPIY) in our detection layers — the closure table holds no event for either, open or recent. The airport’s weather station reported throughout the entire window: 1,526 observations over the last 30 days, with IFR or worse 4% of the time, and the latest observation (7 Sep, 15:00 UTC) reporting CAVOK — no cloud below 5,000 ft, visibility 10 km or better — with wind from the east-south-east at 13 knots, 17 °C, and no thunderstorm or gust flags. That cadence is also its own control on both of our collection gaps: on 4 September the station filed 49 observations and on 6 September it filed 50, on its normal schedule, so the hole was on our side and not at the airport. The GNSS-interference ratio for the area sat at its measurement floor across the closing 7-day window (0.0004 average, 0.0007 peak, level low), measured on 29 of 29 grid cells and last updated 6 September — and the aircraft-telemetry feed behind that layer observed 15,468 aircraft over Perth’s FIR across those seven days, including 2,077 on 4 September and 1,624 on 6 September, the two days our airport-movement collection went thin. No volcanic-ash exposure was ever in the frame (ash: none). Every layer is consistent with one explanation: a normal quiet Saturday at Perth, followed by ordinary working weeks.

What the data showed — the closing numbers. We opened this page with one falsifier: if the working week resumed without traffic returning to Perth’s weekday norm, the −65% would stop being seasonality and start being an event. That test has now run thirteen times and never fired. Thirteen measured days follow the flag, and eleven of them landed within about 4% of their own same-weekday norm (365 vs ~381, 597 vs ~604, 667 vs ~682, 668 vs ~670, 457 vs ~450, 211 vs ~216, 381 vs ~381, 608 vs ~604, 698 vs ~682, 648 vs ~664, 648 vs ~670). The closing number we would put on this event is the working week of 31 August – 3 September: 608, 698, 648 and 648 observed movements against norms of ~604, ~682, ~664 and ~670 — four consecutive two-sided working days, the widest of them 3.3% under its own norm, at an airport our detector had flagged eleven days earlier at −65%. The two days that fell outside the 4% band are not counter-evidence: Wednesday 26 August (453, arrivals on norm and departures ~48% short — the shape of partial capture) and Saturday 5 September, which at 179 movements (88 departures / 91 arrivals) is 17% below the ~216 Saturday baseline, −15% against the 211-movement baseline our own airport detector uses, and scored normal by it — the third consecutive Saturday inside the same weekly trough that produced the original flag. Two collection gaps on our side, 4 and 6 September, are excluded from every one of these figures and starred in the chart. No corroborating signal appeared in any layer at any point in the sixteen days.

What we cannot say, and what would reopen this page. Observed counts are a floor, not a total — our receiver network does not hear every movement, and we make no claim about Perth’s true schedule-level totals, on the flagged day or on any day since. This page is not a verdict about any individual flight; the operating picture on a given day belongs to the airport, the airlines and the Australian authorities. Our freshness limit stands at close as it stood throughout: the newest two or three days of this feed are provisional in both directions, which applies to 5 September as much as it applied to every day this page corrected. And it corrected six times, publicly: 28 August went 436 → 31 → 457; 29 August 196 → 15 → 211; 30 August 316 → 381, caveat withdrawn; 31 August 115 → 499 → 608; 1 September 621 → 285 → 698; 3 September 133 → 585 → 648. The last of them reversed the caution we carried into this update: 2 September, published at 586, overwritten to 179 inside a ~65%-of-median window where we reported the number and refused to draw a conclusion from it, has settled at 648 (383 dep / 265 arr) — dead on its Wednesday norm. Two things would reopen this page: a genuine closure, weather deterioration, ash exposure or GNSS-interference signal appearing for Perth or its FIR in any layer, retrospectively or fresh — none has, in sixteen days; or Perth’s weekdays falling away from their own same-weekday norms, which is the fastest thing to check and the only reading that would turn this from seasonality into an event. Automated multi-signal analysis keeps re-checking every layer every three hours and the PER airport brief keeps updating from the same cycle; this page stops here unless one of those two things appears. The standing lessons it leaves behind: at airports with a weekly rhythm this deep, same-weekday baselines are the standard we hold our own alerts to — and the newest days in any window are provisional in both directions, so a settled-looking number can be taken back and a withdrawn one can come back.

Passenger questions, answered from the data

Was Perth airport closed or disrupted?

No signal in the layers we track indicated a disruption at any point: no closure registered for Perth or its FIR, the weather station reporting on its normal cadence with no thunderstorm or gust flags, GNSS interference at its measurement floor, and no volcanic-ash exposure. The −65% figure came from comparing a Saturday against a weekday-heavy weekly average at an airport whose observed traffic drops about 64% every Saturday by rhythm — and the count itself was a balanced 91 arrivals against 99 departures, the shape of real light traffic, not a coverage failure. The thirteen measured days since ran on Perth’s own same-weekday history, eleven of them within about 4% of their own norm, ending with a working week of 608 movements on Monday 31 August, 698 on Tuesday 1 September, 648 on Wednesday 2 September and 648 on Thursday 3 September, and a Saturday, 5 September, at 179 — scored normal by our own detector. This brief is closed. For the status of a specific flight, check your airline or the airport’s own channels.

Why is Perth so much quieter on Saturdays?

Perth runs a pronounced weekly traffic rhythm — in our measurements ~600 observed movements on a typical Monday–Friday versus ~216 on a Saturday, roughly a 64% weekend trough that repeats every week, with Saturday the quietest day. The shape is consistent with Perth’s role as the gateway to Western Australia’s resources sector, where weekday fly-in/fly-out charter and business travel dominate and weekends run much lighter — though we label that mechanism as interpretation. What is measured is the rhythm: four consecutive prior Saturdays came in at 209, 233, 199 and 223 movements, the flagged Saturday at 190, and the two after it at 211 (29 August) and 179 (5 September).

Did flights at Perth really drop 65%?

The flag fired at −65% on a preliminary count of 178 against Perth’s ~506 seven-day rolling average; the observed count for Saturday 22 August settled at 190, which is 62% below that average — and 12% below Perth’s Saturday average of ~216 (209, 233, 199, 223 on the four prior Saturdays). The drop was real arithmetic against the wrong baseline: at an airport with a ~64% built-in weekend trough, same-weekday comparison is what distinguishes a normal Saturday from a genuine disruption. The composition confirmed it was normal traffic, not a data hole — 91 arrivals against 99 departures, an even two-sided ledger, rather than the one-sided departure collapse that marks a receiver-coverage loss — and the two Saturdays that followed, 29 August at 211 and 5 September at 179, measured the same balanced shape inside the same trough.

Why is this brief closed?

Because the signal that opened it resolved and the check that could have overturned it ran thirteen times without firing. Perth’s FIR carried no closure signal at any point, GNSS interference stayed at its measurement floor, the weather station never broke its reporting cadence, and no ash exposure was in the frame — while thirteen measured days followed the flag, eleven of them within about 4% of their own same-weekday norm, closing with a working week of 608, 698, 648 and 648 movements from Monday 31 August to Thursday 3 September. We also stopped carrying an open caveat: the 2 September reading of 179 that we published and refused to interpret has re-settled at 648, dead on its Wednesday norm. Two things would reopen the page — a corroborating signal appearing in any layer, or Perth’s weekdays falling away from their own norms. The PER airport brief keeps updating either way.

Is it safe to fly to Perth right now?

We publish measured signals, not safety verdicts — operating decisions belong to airlines, airports and the Australian authorities, who work from their own complete data. Nothing in the layers we track — closure signals, weather reporting, ash exposure, GNSS interference — indicated anything unusual at Perth across the sixteen days of this brief; what our detector flagged was, on the evidence, the airport’s normal weekly Saturday rhythm, and the thirteen measured days since ran on their own same-weekday norms. For any specific flight, check your airline on the day of travel.

For media — the full kit

Everything on this page — charts, CSV, quotes — is published under CC BY 4.0: free to reuse with attribution to FlySafe.zone. Charts come in dark (social/web) and light (print/light layouts) versions; attribution and license are embedded in every file.

// Copy-paste quote

“Observed traffic at Perth International fell to 190 movements on Saturday 22 August — a −65% anomaly against the airport’s weekly rolling average — but against its Saturday baseline of about 216 the day was a normal weekend low, down 12%: Perth sheds roughly 64% of its weekday traffic every Saturday by rhythm, the count was a balanced 91 arrivals against 99 departures, the shape of real light traffic rather than a coverage loss, and the thirteen measured days that followed ran on their own same-weekday norms — eleven of them within about 4%, closing with a working week of 608, 698, 648 and 648 movements from Monday 31 August to Thursday 3 September and a Saturday, 5 September, at 179 — according to flight-monitoring service FlySafe.zone, which closed the brief on 7 September with no corroborating closure, weather, ash or GNSS-interference signal found in any layer.”

Chart: observed traffic at PER
PNG · 1600px · attribution baked in
Observed daily movements at PER — light version preview
Chart: advisory cadence
PNG · 1600px · attribution baked in
Observed daily movements at Perth International (PER), 22 Aug – 6 Sep 2026 (4 and 6 Sep: network-wide collection gaps on our side — undercounts, excluded from every baseline) — light version preview
↓ Data behind the charts — CSV CC BY 4.0 · dates + daily values, ready for your own chart
Boilerplate — about the source (for your article footer)

FlySafe.zone monitors airspace risk and observed airport activity worldwide: 334 flight-information regions scored daily, GPS-interference signals, volcanic-ash exposure and daily observed movements at 493 airports from the crowdsourced ADS-B receiver network. Methodology: flysafe.zone/methodology.

Need a different cut of the data, a comment, or numbers for another airport? Leave an email — we answer fast during active events.

No newsletters — we only reply to this request. Press page
// Cite this
"FlySafe.zone — Perth (PER) Traffic ‘Drop’ — a −65% Anomaly That Was the Airport’s Normal Saturday, August 2026", https://flysafe.zone/events/per-traffic-drop-2026-08/, data as of 7 Sep 2026, 15:52 UTC — free to cite with a link. Observed-traffic figures are FlySafe’s own ADS-B-derived counts (floor, not total). Observed movement counts from the crowdsourced ADS-B receiver network (a floor, not a total); weather-reporting continuity from published METARs; GNSS-interference ratio from aircraft telemetry. Method: /methodology/ · Press: /press/.

Related coverage