Pricing Search Start free trial
// EVENT · TRAFFIC DISRUPTION CLOSED 14 OCT · NO DISRUPTION — THE −64% FLAG WAS A NORMAL SUNDAY AT A CARGO SUPERHUB (99 VS ~103) · EIGHT WEEKS ON: TUE–FRI 429–480, SUNDAYS 183–193 · 19 MEASURED SUNDAYS, ALL IN 34–213 · NO CLOSURE EVENT FOR THE MEMPHIS FIR · NO CLOSURE NOTICE IN OUR SET (REFRESHED TO 4 OCT) · GNSS AT FLOOR Data as of 14 Oct 2026, 01:25 UTC · closed · MEM series settled through 9 Oct; 10 Oct a clipped day plotted as a recount; 11 Oct provisional; 12 Oct partial, not plotted; 13 Oct not collected; NOTAM set refreshed to 4 Oct

Memphis (MEM) Traffic 'Drop' — a −64% Anomaly That Is the FedEx Superhub's Normal Sunday, August 2026

← All airspace events

Observed traffic at Memphis International (MEM) came in at 99 movements on 16 August — 64% below its ~276 all-days daily average — and tripped our traffic-anomaly detection. The cross-checks say this is not a disruption: 16 August was a Sunday, and Memphis — home of a global express-cargo superhub — sheds roughly 71% of its weekday traffic every weekend by design. Against its Sunday-only baseline (~103 movements — the mean of the five Sundays in the 30 days to 16 August; the nineteen Sundays since early June that landed on a normal collection day, the nineteenth provisional, all sit in a 34–213 range), the day measured −4%: a normal weekend. No airspace-closure signal has registered for the Memphis FIR (KZME) in our layers since early June, the airport's weather station kept its normal hourly cadence in benign VFR conditions, GNSS interference sat at its measurement floor, and traffic at Atlanta, Chicago O'Hare and Dallas–Fort Worth was normal the same day — ruling out a regional receiver-coverage loss. We publish this as the anatomy of a seasonality false alarm, honestly labeled — because the same raw number could just as easily have become a phantom headline. This page closed on 14 October: eight weeks of same-weekday history since the flag — every Tuesday–Friday of the four most recent weeks inside or above MEM's normal range, every Sunday inside its Sunday range — and no corroborating signal in any layer.
// Flying through Memphis? The short version
  • Closed 14 October, no disruption — the −64% 'drop' on 16 August is what every Sunday looks like at MEM when compared against a weekday-heavy average, and eight weeks of same-weekday history since say the same.
  • Memphis has the deepest weekly traffic rhythm we measure at any large US airport: ~351 observed movements on a typical Tuesday–Friday, ~103 on a Sunday — the express-cargo network largely stands down on weekends.
  • No disruption signal corroborated anything unusual while this page was open: no closure event for the Memphis FIR, no closure notice in our NOTAM set (last refreshed 4 October), the weather station reporting on its normal cadence, GNSS interference at its floor — for a specific flight, check your airline as always.

Advisory timeline — how active is this episode

This chart deliberately shows only Sundays — because that is the comparison our detector should have made. Sixteen August was a Sunday, and at Memphis a Sunday is not a smaller version of a Tuesday: it is a different airport. Across the nineteen Sundays since early June, every one of them measured on a normal collection day (the nineteenth provisional), our receiver network observed between 34 and 213 movements at MEM — never anywhere near the 268–430 range of a normal Tuesday–Friday. Sunday 16 August, at 99 observed movements, sits almost exactly on the Sunday baseline (~103 — the mean of the five Sundays in the 30 days to 16 August). The Sundays after it have each been re-read since publication, and each now stands as a measured, ordinary Sunday. 23 August — first published here as a starred 19-movement collection gap — settled at 128. 30 August was read three times — published as a measured 98, withdrawn by us as a collection gap, then restored — and has held unchanged through every refresh since. That day's global observed volume reads ~93% of its 14-day median — comfortably above the 60% mark our own rule sets — with Atlanta (2,043 movements), Chicago O'Hare (2,197) and Dallas–Fort Worth (1,861) all at normal levels the same day. The withdrawal was our error: 30 August is a genuine collection day and an ordinary Sunday, 98 movements, −5% against the ~103 Sunday baseline. 6 September — published as a measured 103, withdrawn on 9 September when that day read ~26% of median, then restored — has since held at a normal collection day (~90% of its 14-day median), with Atlanta 1,897, Chicago O'Hare 2,306 and Dallas–Fort Worth 1,735, at MEM measuring 108 observed movements: an ordinary Sunday, +5% against the ~103 Sunday baseline. 13 September is restored: published here on 15 September as a measured 172 and withheld on 16 September when the day read ~59% of its 14-day median, it has since backfilled to a normal collection day (~91% of its 14-day median), with Atlanta 1,952, Chicago O'Hare 2,445 and Dallas–Fort Worth 1,869, on which MEM measures 174 observed movements: +69% against the ~103 Sunday baseline and, until 20 September, the busiest Sunday on this page — still far below the weekday floor. 20 September is settled at 213 observed movements, firmed from the 193 we published on 23 September and unchanged on today's re-read — with Atlanta 2,141, Chicago O'Hare 2,520 and Dallas–Fort Worth 1,816 all normal that day: +107% against the ~103 Sunday baseline, the busiest Sunday on this page, which widens the measured Sunday range to 34–213 — still far below the weekday floor. It has left the window our own feed re-reads on the nightly refetch, and the three passenger hubs no longer read below their own five-Sunday spans (Atlanta above 1,897–2,043, Chicago O'Hare above 2,197–2,452, Dallas–Fort Worth inside 1,735–1,971) — which is why the 193 moved up. 27 September is restored, and our own reading of it as a hole is withdrawn. On 29 September we published that Sunday as not collected — 19,389 movements worldwide, ~8% of the 244,092 14-day median, with the daily rollup and the underlying detection record agreeing on the figure — and said it might backfill. It has: the detection record now holds 233,963 movements worldwide for the day (~96% of the median), with Atlanta 2,014 inside its five-Sunday span of 1,897–2,141, Dallas–Fort Worth 1,884 just above its 1,735–1,869, Louisville 339 inside 288–350 and Indianapolis 476 above 351–464, and MEM at 189 observed movements (124 departures, 65 arrivals): +83% against the ~103 Sunday baseline, the second-busiest Sunday on this page at the time, after 20 September, and inside the 34–213 range, which is unchanged. The hole was in our ingestion, not in the day. The two records agreeing on a low count told us only that the rollup was not clipping; it did not tell us the movements were absent, and we no longer offer that equality as the sign of a real hole. The bar is now settled. The day has left the window our feed re-reads: its last nightly pass, on 1 October, re-counted it in full, so the daily rollup and the detection record now agree on 189 at MEM and 233,963 worldwide, and the clipped 177 we noted on 30 September is gone. Chicago O'Hare, at 2,101, still reads below its own five-Sunday span of 2,197–2,520 and did not move on that last pass, so that is the day, not our collection. Sunday 4 October is settled at 183 observed movements (124 departures, 59 arrivals), the rollup and the detection record agreeing after its last nightly pass on 8 October, on a normal collection day — 226,013 movements worldwide, ~90% of the 14-day median, which on today’s run is 250,150 — with Atlanta 1,981 inside its five-Sunday span of 1,897–2,141, Chicago O'Hare 2,392 inside 2,101–2,520, Dallas–Fort Worth 1,882 inside 1,735–1,884, Louisville 340 inside 288–350 and Indianapolis 437 inside 351–476: +78% against the ~103 Sunday baseline, the fourth-busiest Sunday on this page and inside the 34–213 range, which is unchanged. It was also a day of low cloud at MEM — 11 of the station's 27 reports in IFR conditions — and the Sunday count is unremarkable regardless. Sunday 11 October is provisional at 193 observed movements (128 departures, 65 arrivals), the rollup and the detection record agreeing, on a collection day that reads ~81% of the median (201,817 worldwide — above the 60% mark our rule sets, so no flag) with all five cross-hubs inside their five-Sunday spans (Atlanta 1,972, Chicago O'Hare 2,350, Dallas–Fort Worth 1,874, Louisville 317, Indianapolis 366): +87% against the ~103 Sunday baseline, provisionally the second-busiest Sunday on this page, ahead of 27 September's 189, and inside the 34–213 range. Its last nightly pass is due on 15 October, after this page closed, so 193 is a floor; the settled figure will stand on the live MEM airport brief, not here. Nineteen measured Sundays now, the nineteenth provisional. Same-weekday history is the honest baseline for an airport with a weekly pulse this deep:

52
7 Jun
111
14 Jun
59
21 Jun
34
28 Jun
85
5 Jul
99
12 Jul
153
19 Jul
89
26 Jul
116
2 Aug
58
9 Aug
99
16 Aug
128
23 Aug
98
30 Aug
108
6 Sep
174
13 Sep
213
20 Sep
189
27 Sep
183
4 Oct
193
11 Oct

Fig. 1 — Observed daily movements at Memphis International (MEM) on each of the last 19 measured Sundays, 7 Jun – 11 Oct 2026 (11 Oct provisional) · Data as of 14 Oct 2026, 01:25 UTC · closed · MEM series settled through 9 Oct; 10 Oct a clipped day plotted as a recount; 11 Oct provisional; 12 Oct partial, not plotted; 13 Oct not collected; NOTAM set refreshed to 4 Oct · Source: FlySafe monitoring (crowdsourced ADS-B receiver network).

Airports in the exposure zone

AirportSizeObserved signalFIR
Memphis International (MEM) large — · Sun 16 Aug: 99 observed vs ~103 Sunday baseline KZME

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 shows: a saw-tooth by design — the weekly breath of a cargo superhub

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. Memphis International's last 30 days:

Fig. 2 — Observed daily movements at MEM, 21 Jul – 11 Oct (latest complete UTC day). Data as of 14 Oct 2026, 01:25 UTC · closed · MEM series settled through 9 Oct; 10 Oct a clipped day plotted as a recount; 11 Oct provisional; 12 Oct partial, not plotted; 13 Oct not collected; NOTAM set refreshed to 4 Oct · Source: FlySafe monitoring (ADS-B receiver network).

  • Tuesday–Friday, last four full weeks: ~351 observed movements/day on average, peaking at 430 (28 Jul).
  • Sundays in the 30 days to 16 Aug: ~103 on average — a ~71% trough, every single week. Saturdays sit in between — averaging ~192 over the four full weeks to 16 Aug, with 15 Aug settling at 297 and the next Saturday, 22 Aug — first published here as a 19-movement thin-day read — settling at 335 in the nightly refetch, its busiest Saturday of the 30-day window.
  • Sunday 16 Aug: 99 observed movements — −4% against the Sunday baseline, inside the 34–213 range of the nineteen Sundays since early June, every one of them measured on a normal collection day (the nineteenth provisional).
  • The rebound has now been measured across a full working week and through the weekends that followed: Monday 17 Aug came in at 102 — inside MEM's thin-Monday norm (77–157 over the window, the sort cycle spins up Monday night) — then Tuesday 18 Aug at 334, Wednesday 19 Aug at 307, Thursday 20 Aug at 318 and Friday 21 Aug at 353, all back inside the normal Tuesday–Friday range (268–430 across the four full weeks to 16 Aug). Saturday 22 Aug, first published here as a 19-movement thin day, finalised at 335 — its busiest Saturday of the 30-day window — and Sunday 23 Aug, which we then published as a network-wide collection gap, finalised at 128: an ordinary Sunday, +24% against the ~103 Sunday baseline. Monday 24 Aug measured 157 (+22% against its Monday median of 129 — the four Mondays in the four full weeks to 16 Aug, the same pinned window as every other baseline here), then Tuesday 25 Aug 277, Wednesday 26 Aug 179, Thursday 27 Aug 425, Friday 28 Aug 286 and Saturday 29 Aug 217 — an ordinary Saturday, +13% against the ~192 Saturday average, on a day whose global volume closed at ~96% of median. Sunday 30 Aug holds at 98 through every refresh since we restored it (global volume ~93% of median; ATL 2,043, ORD 2,197, DFW 1,861), −5% against the ~103 Sunday baseline. Monday 31 Aug also holds at 78 on a normal collection day (~98% of median) — the low edge of MEM's thin-Monday norm (77–157 across the window), −40% against its Monday median of 129, with no corroborating signal in any layer. Tuesday 1 Sep stands at 376 on a clean day (~100% of its 14-day median; ATL 1,930, ORD 2,618, DFW 1,864) and Wednesday 2 Sep at 347 (~101% of median; ATL 1,903, ORD 2,478, DFW 1,819) — both ordinary weekdays inside the 268–430 range, both unchanged through today's refresh after the reversals we log on the record below. Thursday 3 Sep is a normal collection day (~100% of median, neighbouring hubs normal) and holds at 259 observed movements, unchanged again. That is the second soft weekday: 259 sits just under the low edge of the 268–430 Tue–Fri range and −26% against the ~351 weekday average, with no corroborating signal in any layer — we label it as such rather than rounding it into the range. The first one is unchanged: Wednesday 26 Aug at 179 sits well below the normal Tue–Fri range on a day whose global volume closed at ~94% of median — an ordinary collection day — with no corroborating signal in any layer, and MEM measured 425 the next day. A third, far deeper one sits the day after Labor Day (Tuesday 8 Sep, below), and two of the eleven Tue–Fri days measured after it (9 and 11 Sep) are soft as well: five soft weekdays, each labelled for exactly what it is. Friday 4 Sep holds at 331 on a clean day (~107% of median; ATL 2,081, ORD 2,523, DFW 2,068) — an ordinary Friday inside the weekday range. Saturday 5 Sep — published as a measured 272, withdrawn as a collection gap, then restored — holds at 276 observed movements on a genuine collection day (~94% of its 14-day median; ATL 1,839, ORD 2,538, DFW 1,887): an ordinary Saturday, +44% against the ~192 Saturday average of the pinned window and inside the 150–335 span of every Saturday measured since 21 July. Sunday 6 Sep — published as a measured 103, withdrawn on 9 September when that day read ~26% of median, then restored — holds at a normal collection day (~90% of its 14-day median; ATL 1,897, ORD 2,306, DFW 1,735) measuring 108 observed movements — an ordinary Sunday, +5% against the ~103 Sunday baseline. Monday 7 Sep holds as a normal collection day (~100% of its 14-day median; ATL 1,964, ORD 2,521, DFW 1,955) and reads 104 observed movements, inside MEM's thin-Monday norm (77–157), −19% against its Monday median of 129. Tuesday 8 Sep stands at 32 observed movements on a normal collection day (~96% of its 14-day median; ATL 1,747, ORD 2,252, DFW 1,673): −91% against the ~351 weekday average, the deepest weekday reading on this page. What we can add is also measured: the overnight sort-bank hours (02–05 UTC) carry 2 movements that day against 105 on 9 Sep; MEM's weather station kept its hourly cadence (24 reports, all VFR); no closure event registered for the Memphis FIR; and the two other cargo hubs on this page sat low the same day — Louisville (SDF) at 316, −46% against its 1–4 Sep average, Indianapolis (IND) at 441, −25%. Monday 7 Sep was the US Labor Day holiday, and a hub week that starts one night late would look exactly like this — that reading is interpretation, not measurement. Wednesday 9 Sep reads 236 observed movements on a normal collection day (~102% of its 14-day median; ATL 1,891, ORD 2,438, DFW 1,789): below the 268–430 Tue–Fri range, −33% against the ~351 weekday average, the fourth soft weekday. Its overnight sort-bank hours (02–05 UTC) carry 105 movements, with arrivals (113) and departures (123) balanced. Thursday 10 Sep is restored: withdrawn on 12 September as a ~15% collection gap, it has backfilled to a normal collection day (ATL 1,943, ORD 2,452, DFW 1,839) on which MEM measures 397 observed movements — an ordinary Thursday inside the 268–430 range, +13% against the ~351 weekday average, and the first of the four post-Labor-Day weekdays so far inside MEM's normal range (10, 23, 25 and 29 Sep). Friday 11 Sep reads 250 observed movements on a normal collection day (ATL 1,692, ORD 2,502, DFW 2,142) — up from the 243 we published on 14 September — −29% against the ~351 weekday average and the fifth soft weekday. It is arrival-light (50 arrivals against 200 departures), and the total holds through today's refresh. MEM's weather station recorded a thunderstorm late that day (from 21:40 UTC, visibility down to ½ mile in heavy rain, a 34-knot peak gust at 21:33, back to MVFR by 22:18), after the day's measured departure banks, so we do not offer it as the explanation. Saturday 12 Sep is restated: published on 15 September as a soft 79 on a partly-heard day, it has settled to a clean collection day (~95% of its 14-day median; ATL 1,973, ORD 2,442, DFW 1,917) on which MEM measures 203 observed movements — an ordinary Saturday, +6% against the ~192 Saturday average and inside the 150–335 span of every Saturday measured since 21 July. The 79 was collection, not traffic, and the cargo-layer softness we described beside it is withdrawn with it. Sunday 13 Sep is restored: published on 15 September as a measured 172 and withdrawn on 16 September at ~59% of median, it has backfilled to a normal collection day (~91% of its 14-day median; ATL 1,952, ORD 2,445, DFW 1,869) on which MEM measures 174 observed movements (103 departures, 71 arrivals) — +69% against the ~103 Sunday baseline, the busiest Sunday on this page until 20 September; the measured Sunday range then stood at 34–174 across fifteen Sundays. Monday 14 Sep is restored: published on 16 September as a measured 170 and withdrawn on 17 September at ~47% of median, it has backfilled to a normal collection day (~102% of its 14-day median) (ATL 2,100, ORD 2,468, DFW 2,024) on which MEM measures 180 observed movements (129 departures, 51 arrivals): +40% against its Monday median of 129 and above the 77–157 thin-Monday norm — the busiest Monday on this page until 28 September. A busy Monday is not a disruption signal, and nothing in any layer points to one. Wednesday 16 Sep is restored and Thursday 17 Sep restated: a later re-read reversed how we placed both on 19 September, and both have held since. Tuesday 15 Sep holds at 535 observed movements (293 departures, 242 arrivals) on a normal collection day (ATL 2,057, ORD 2,318, DFW 1,874): +52% against the ~351 weekday average and above the 268–430 Tue–Fri range. Its overnight sort-bank hours (02–05 UTC) carry 175 movements, against 105 on 9 Sep and 2 on 8 Sep. Wednesday 16 Sep — published on 18 September as a measured 536, then withdrawn a day later as a ~32% collection gap with MEM reading 57 — has backfilled to a normal collection day (~102% of its 14-day median) (ATL 1,930, ORD 2,427, DFW 1,797) on which MEM measures 539 observed movements (292 departures, 247 arrivals): +54% against the ~351 weekday average, above the 268–430 range and the busiest day on this page, with 161 movements in the overnight sort-bank hours. The withdrawal was our error: 16 September is a genuine collection day, and the 57 was collection, not traffic. Thursday 17 Sep — published on 19 September as a measured 510 — now reads 518 observed movements (297 departures, 221 arrivals) on a normal collection day (~104% of its 14-day median; ATL 2,070, ORD 2,066, DFW 1,935): +48% against the ~351 weekday average and above the range, with 157 movements in the sort-bank hours. Friday 18 Sep is restored: published here on 20 September as a measured 445, withdrawn on 22 September when the day read ~15% of its 14-day median with MEM at 17, it has backfilled to a normal collection day (~106% of its 14-day median) (ATL 2,139, ORD 2,613, DFW 2,020) on which MEM measures 451 observed movements (284 departures, 167 arrivals): +28% against the ~351 weekday average and above the 268–430 range, with 120 movements in the overnight sort-bank hours. The withdrawal was our error on the same pattern as 16 September: 18 September is a genuine collection day, and the 17 was collection, not traffic. Across 15–18 Sep Louisville (SDF) read 615, 618, 628 and 578 and Indianapolis (IND) 581, 526, 556 and 480, against 1–4 Sep averages of 586 for both — SDF within 10% throughout, IND within 10% on 15 and 17 Sep and below it on 16 and 18. A high count is not a disruption signal and nothing in any layer points to one. Saturday 19 Sep has firmed from the 305 we published on 22 September to 320 observed movements (203 departures, 117 arrivals) on a collection day that now reads ~92% of its 14-day median (ATL 1,815, ORD 1,784, DFW 1,897): an ordinary Saturday, +67% against the ~192 Saturday average and inside the 150–335 span of every Saturday measured since 21 July, with 76 movements in the overnight sort-bank hours. Chicago O'Hare still reads well below its own five-Saturday span (2,436–2,549) that day, so 320 remains a floor; the two cargo hubs beside MEM read normal for a Saturday — Louisville 458 (five prior Saturdays 438–464) and Indianapolis 392 (323–416). Sunday 20 Sep has firmed from the 193 we published on 23 September to 213 observed movements (135 departures, 78 arrivals) on a collection day that now reads ~89% of its 14-day median (ATL 2,141, ORD 2,520, DFW 1,816; Louisville 329 against a five-Sunday span of 286–350, Indianapolis 464 against 351–454): +107% against the ~103 Sunday baseline, the busiest Sunday on this page — and still under two-thirds of the ~351 weekday average. Monday 21 Sep has firmed from the 171 we published on 23 September to 175 observed movements (121 departures, 54 arrivals) on a collection day that reads ~94% of its 14-day median (ATL 2,078, ORD 2,507, DFW 2,019; Louisville 200 against a five-Monday span of 189–235, Indianapolis 400 against 429–503): +36% against its Monday median of 129 and above the 77–157 thin-Monday norm, at the time the second-busiest Monday on this page after 14 September. Tuesday 22 Sep has firmed from the 446 we published on 24 September to 455 observed movements (312 departures, 143 arrivals), with ATL 2,136, ORD 2,511 and DFW 1,972 normal the same day (Louisville 618 against a five-Tuesday span of 316–621, Indianapolis 504 against 441–675): +30% against the ~351 weekday average and above the 268–430 range, with 80 movements in the overnight sort-bank hours — well under the 157–175 those hours carried on 15–17 September. The collection-health flag we refuted on 22 September has cleared, exactly as we said it would. The nightly refetch refilled that day: the daily rollup and the detection record now both hold 244,092 movements worldwide — they were 0.02% apart when we first refuted the flag — and 455 at MEM in both. The flag was our own trailing-day artefact, not a hole in the day, and the 455 we published against it stands unchanged. Wednesday 23 Sep has firmed from the 423 we published on 25 September to 430 observed movements (290 departures, 140 arrivals), with ATL 2,000, ORD 2,441 and DFW 1,883 normal the same day (Louisville 611 inside a five-Wednesday span of 591–626, Indianapolis 496 just under its own 526–589): +23% against the ~351 weekday average, sitting exactly on the top edge of the 268–430 Tue–Fri range, with 94 movements in the overnight sort-bank hours. The flag we refuted on 24 September has cleared, and so has the one we refuted on 25 September. On 24 September the rollup holds 255,236 movements worldwide (~105% of the 244,092 14-day median) and 431 at MEM, matching the detection record to the movement: the ~15% reading we published against was our artefact, and the 431 stands. On 25 September the nightly refetch refilled the day we refuted on 29 September — when the rollup showed 37,196 worldwide, ~15% of the median, and clipped MEM to 10 — and the rollup now holds 264,575 movements worldwide and 398 at MEM, matching the detection record exactly; the 398 we published against our own flag stands. No collection-health flag stands anywhere in the pulse window today. Thursday 24 Sep is settled at 431 observed movements (289 departures, 142 arrivals) — unchanged on today's re-read and out of the window our feed re-reads — with ATL 2,072, ORD 2,411 and DFW 1,960 each inside its own five-Thursday span that day (Louisville 643 inside 574–647, Indianapolis 492 inside 487–591): +23% against the ~351 weekday average and — at 431 — a single movement above the top edge of the 268–430 Tue–Fri range, with 84 movements in the overnight sort-bank hours. Friday 25 Sep is settled at 398 observed movements (279 departures, 119 arrivals) — unchanged on today's re-read, with the rollup and the detection record now agreeing on it to the movement: ATL 2,158 just above its five-Friday span of 1,692–2,139, O'Hare 2,437 inside 2,404–2,613, Louisville 612 above its own 575–595 and Indianapolis 568 inside 480–572: +13% against the ~351 weekday average and inside the 268–430 Tue–Fri range — the third post-Labor-Day weekday to land in it, after 10 and 23 September — with 78 movements in the overnight sort-bank hours. Dallas–Fort Worth (2,003 against 2,020–2,142) reads a little under its own five-Friday span that day and did not move on the refetch, so that is the day, not our collection. Saturday 26 Sep has firmed from the 301 we published on 29 September to 307 observed movements (205 departures, 102 arrivals) on a collection day that now reads ~98% of the 244,092 14-day median (239,418 worldwide, the rollup and the detection record agreeing to the movement; ATL 1,909 inside its five-Saturday span of 1,815–2,026, O'Hare 2,407 inside 1,784–2,549, Louisville 492 above its own 438–461, Indianapolis 460 above 323–416): an ordinary Saturday, +60% against the ~192 Saturday average of the pinned window and inside the 150–335 span of every Saturday measured since 21 July, with 57 movements in the overnight sort-bank hours. Dallas–Fort Worth (1,856 against a five-Saturday span of 1,860–1,917) still reads four movements under its own and did not move on the nightly refetch, so that is the day, not our collection. 20 through 30 September are settled: all eleven have left the window our feed re-reads on the nightly refetch — 29 September was re-read for the last time on the 3 October pass, 30 September on today's 4 October pass — and the detection record moved on none of the first nine; 29 September firmed on its last pass and 30 September held at the 511 we published as a floor, as set out below. Sunday 27 Sep is restored, and the 'no bar, not our artefact' we published for it on 29 September is withdrawn. The day then held 19,389 movements worldwide in the rollup and the detection record alike, and it has since backfilled: the detection record now holds 233,963 worldwide (~96% of the 244,092 14-day median; ATL 2,014 inside its five-Sunday span of 1,897–2,141, DFW 1,884 just above 1,735–1,869, Louisville 339 inside 288–350, Indianapolis 476 above 351–464) and 189 at MEM (124 departures, 65 arrivals): +83% against the ~103 Sunday baseline, the second-busiest Sunday on this page at the time, after 20 September, and inside the 34–213 range, with 11 movements in the overnight sort-bank hours — a Sunday's share. The hole was in our ingestion, not in the day; the two records agreeing on a low count told us only that the rollup was not clipping, and we no longer offer that equality as the sign of a real hole. The rollup clipped the day to 177 at MEM on our 30 September refresh; its last nightly pass, on 1 October, re-counted it in full, and the two records now agree on 189 at MEM and 233,963 worldwide, so 189 is settled. Chicago O'Hare, at 2,101, still reads below its own five-Sunday span of 2,197–2,520 and did not move on that last pass, so that is the day, not our collection. Monday 28 Sep has firmed from the 193 we published on 30 September to 199 observed movements (140 departures, 59 arrivals) on a normal collection day (~104% of the 14-day median, which on today's run is 239,418: 248,730 worldwide in the detection record; ATL 2,167 just above its five-Monday span of 1,964–2,150, O'Hare 2,733 above 2,143–2,521, DFW 2,003 inside 1,955–2,024, Louisville 247 above 189–235, Indianapolis 442 inside 400–503): +54% against its Monday median of 129, above the 77–157 thin-Monday norm and the busiest Monday on this page, ahead of 14 September's 180, with 22 movements in the overnight sort-bank hours. A busy Monday is not a disruption signal, and nothing in any layer points to one. The rollup clipped the day to 158 at MEM on our 1 October refresh; its last nightly pass, on 2 October, re-counted it in full, and the two records now agree on 199 at MEM and 248,730 worldwide, so 199 is settled. Tuesday 29 Sep has backfilled, as we said it might, and is now settled. On 30 September we published it with no bar — 19,615 movements worldwide, ~8% of the median, MEM at 9 — and on 1 October as a backfilled 394 that we called a floor. Its last nightly pass, on 3 October, re-counted it in full: the rollup and the detection record now agree on 222,287 worldwide (~93% of the median, which on today's run is again 239,418) and 405 at MEM (274 departures, 131 arrivals): +15% against the ~351 weekday average and inside the 268–430 Tue–Fri range — the fourth post-Labor-Day weekday to land in it, after 10, 23 and 25 September — with 97 movements in the overnight sort-bank hours. Atlanta (1,739), Chicago O'Hare (2,073) and Dallas–Fort Worth (1,592) each rose by about thirty movements on that pass and still read below their own five-Tuesday spans (1,747–2,136, 2,252–2,618, 1,673–1,972), while Louisville (539, inside 316–621) and Indianapolis (525, inside 441–604) read normal; with the day now out of the re-read window, that is the day, not our collection. Wednesday 30 Sep has backfilled, as we said it might, and is now settled. On 1 October we published it with no bar — 20,172 movements worldwide, ~8% of the median, MEM at 6 — and on 3 October as a backfilled 511 that we called a floor, with the rollup clipping it to 458 that day. Its last nightly pass, on today's 4 October run, re-counted it in full: the rollup and the detection record now agree on 254,563 movements worldwide (~106% of the median, which on today's run is again 239,418) and 511 at MEM (300 departures, 211 arrivals), so the 511 we published against the clip stands: +46% against the ~351 weekday average and above the 268–430 Tue–Fri range, with 158 movements in the overnight sort-bank hours. Atlanta (2,015) reads just above its five-Wednesday span of 1,891–2,000 and Indianapolis (592) just above 496–589, while Chicago O'Hare (2,384, inside 2,339–2,478), Dallas–Fort Worth (1,769, inside 1,677–1,882) and Louisville (622, inside 591–626) read inside theirs, and none of the five moved on that last pass. Thursday 1 Oct has firmed from the 498 we published on 3 October to 506 observed movements (298 departures, 208 arrivals): +44% against the ~351 weekday average and above the 268–430 range, with 164 movements in the overnight sort-bank hours. We read 245,335 movements worldwide for the day on 3 October and a clipped 185,522 on 4 October; its last nightly pass, on 5 October, left the rollup and the detection record agreeing on 250,150 (~100% of the 14-day median, which on today’s run is itself 250,150), so the day is settled. Atlanta (2,047, inside 1,916–2,077), Chicago O'Hare (2,360, inside 2,066–2,486), Louisville (621, inside 574–647) and Indianapolis (578, inside 492–591) read normal against their five-Thursday spans; Dallas–Fort Worth, at 1,379 against 1,839–2,007, still reads well below its own and did not move on that last pass, so that is the day, not our collection; we report the Dallas reading as measured and draw no conclusion from it. Friday 2 Oct has backfilled, as we said it might, and is now settled. On 3 October we published it with no bar — 20,176 movements worldwide, ~8% of the median, MEM at 6 — and on 4 October as a backfilled 478 we called a floor. Its last nightly pass, on 6 October, left the rollup and the detection record agreeing on 252,986 movements worldwide (~101% of the 250,150 median) and 487 at MEM (298 departures, 189 arrivals): +39% against the ~351 weekday average and above the 268–430 Tue–Fri range, with 128 movements in the overnight sort-bank hours. The cross-hubs settled at Atlanta 2,191, just above its five-Friday span of 1,692–2,158, Chicago O'Hare 2,557 inside 2,437–2,613, Dallas–Fort Worth 2,130 inside 2,003–2,142, Louisville 613 a movement above 578–612 and Indianapolis 511 inside 480–572. Saturday 3 Oct has backfilled, as we said it might, and is now settled. On 4 October we published it with no bar — 17,976 movements worldwide, ~8% of the median, MEM at 4. Its last nightly pass, on 7 October, left the two records agreeing on 228,908 movements worldwide (~92% of the median) and 310 at MEM (211 departures, 99 arrivals): an ordinary Saturday, +61% against the ~192 Saturday average and inside the 150–335 span of every Saturday measured since 21 July, with 75 movements in the overnight sort-bank hours. Atlanta 1,967 (1,815–2,019), Chicago O'Hare 2,439 (1,784–2,538), Dallas–Fort Worth 1,908 (1,856–1,917) and Indianapolis 453 (323–460) read inside their five-Saturday spans, Louisville 498 just above its own 438–492. MEM's weather station spent much of that day in low cloud — 16 of its 33 reports in IFR conditions, 15 more in marginal VFR — the kind of corroborating signal this page said would matter, and the count stayed inside its Saturday span. Sunday 4 Oct is settled at 183 observed movements (124 departures, 59 arrivals) on a normal collection day (~90% of the median: 226,013 worldwide, the two records agreeing after its last nightly pass on 8 October; Atlanta 1,981 inside its five-Sunday span of 1,897–2,141, Chicago O'Hare 2,392 inside 2,101–2,520, Dallas–Fort Worth 1,882 inside 1,735–1,884, Louisville 340 inside 288–350, Indianapolis 437 inside 351–476): +78% against the ~103 Sunday baseline, the fourth-busiest Sunday on this page and inside the 34–213 range, with 16 movements in the overnight sort-bank hours — a Sunday's share. Low cloud again: 11 of the station's 27 reports in IFR conditions, and the Sunday count unremarkable. Monday 5 Oct is settled at 191 observed movements (141 departures, 50 arrivals) on a normal collection day (~97% of the median: 242,490 worldwide; Atlanta 2,215 above its five-Monday span of 1,964–2,167, Chicago O'Hare 2,454 inside 2,143–2,733, Dallas–Fort Worth 2,096 above 1,955–2,024, Louisville 229 inside 189–247, Indianapolis 492 above 400–489): +48% against its Monday median of 129, above the 77–157 thin-Monday norm and the second-busiest Monday on this page after 28 September, with 34 movements in the sort-bank hours. The station's reports were back in VFR all day and stayed there through 13 October. Tuesday 6 Oct is settled at 455 observed movements (299 departures, 156 arrivals) on a normal collection day (~102% of the median: 255,419 worldwide; Atlanta 2,122 inside its five-Tuesday span of 1,739–2,136, Chicago O'Hare 2,494 inside 2,073–2,618, Dallas–Fort Worth 2,102 above 1,592–1,972, Louisville 628 above 316–618, Indianapolis 634 above 441–604): +30% against the ~351 weekday average and above the 268–430 Tue–Fri range, with 110 movements in the sort-bank hours. Wednesday 7 Oct is settled at 480 (297 departures, 183 arrivals) on a normal collection day (~104% of the median: 260,632 worldwide; Atlanta 2,010 inside 1,891–2,015, Chicago O'Hare 2,427 inside 2,384–2,478, Dallas–Fort Worth 2,076 above 1,769–1,882, Louisville 619 inside 591–626, Indianapolis 573 inside 496–592): +37% against the ~351 weekday average and above the range, with 123 movements in the sort-bank hours. Thursday 8 Oct is settled at 468 (317 departures, 151 arrivals) on the busiest collection day of the fortnight (~106% of the median: 265,788 worldwide; Atlanta 2,108 above its five-Thursday span of 1,916–2,072, Chicago O'Hare 2,484 above 2,066–2,452, Dallas–Fort Worth 2,086 above 1,379–1,960, Louisville 636 inside 574–647, Indianapolis 596 above 492–591): +33% against the ~351 weekday average and above the range, with 103 movements in the sort-bank hours. Friday 9 Oct is settled at 429 (282 departures, 147 arrivals), its last nightly pass on 13 October leaving the two records agreeing on it and on 261,407 worldwide (~105% of the median; Atlanta 2,114 inside its five-Friday span of 1,692–2,191, Chicago O'Hare 2,437 on the low edge of 2,437–2,613, Dallas–Fort Worth 2,080 inside 2,003–2,142, Louisville 588 inside 578–613, Indianapolis 608 above 480–568): +22% against the ~351 weekday average and inside the 268–430 range, a single movement under its top edge — the fifth post-Labor-Day weekday to land in it, after 10, 23, 25 and 29 September — with 90 movements in the sort-bank hours. Saturday 10 Oct is plotted from the detection record, as a floor. Our rollup holds 28,657 movements worldwide for it — ~11% of the median, which fires our collection flag — with MEM at 11, because the 13 October 21:25 UTC refresh re-counted the day from that hour onward and nothing earlier; the detection record holds 221,968 worldwide (~89% of the median) and 343 at MEM (207 departures, 136 arrivals), with Atlanta 1,957 inside its five-Saturday span of 1,815–1,973, Chicago O'Hare 2,508 inside 1,784–2,538, Dallas–Fort Worth 2,015 above 1,856–1,917, Louisville 493 inside 438–498 and Indianapolis 443 inside 323–460. The flag is our own trailing-day artefact on the pattern we have logged since 22 September, not a hole in the day. At 343 the day reads +79% against the ~192 Saturday average and above the 150–335 span of every Saturday measured since 21 July — the busiest Saturday on this page, ahead of 22 August's 335 — with 90 movements in the sort-bank hours. A high count is not a disruption signal. Its last nightly pass is due at 02:33 UTC on 14 October, after this page closed, so 343 is a floor and the settled figure will stand on the live MEM airport brief. Sunday 11 Oct is provisional at 193 observed movements (128 departures, 65 arrivals), the rollup and the detection record agreeing, on a collection day that reads ~81% of the median (201,817 worldwide — above the 60% mark our rule sets, so no flag; Atlanta 1,972, Chicago O'Hare 2,350, Dallas–Fort Worth 1,874, Louisville 317 and Indianapolis 366 each inside its own five-Sunday span): +87% against the ~103 Sunday baseline, provisionally the second-busiest Sunday on this page and inside the 34–213 range, with 15 movements in the sort-bank hours — a Sunday's share. Its last nightly pass is due on 15 October, so 193 is a floor. Monday 12 Oct carries no bar, and 13 Oct has not been collected. Our feed holds 18,172 movements worldwide for 12 October (~7% of the median; MEM at 6) and nothing at all after 23:59 UTC that day — the reading 27, 29 and 30 September and 2 and 3 October each gave before backfilling to a normal day. We close this page without them: they belong to the live airport brief now. The open question we stated on 14 September closes with the page: three of the twenty post-Labor-Day Tue–Fri days measured (8, 9 and 11 Sep, all in the first week after the holiday) sit below MEM's normal range, with 10, 23, 25 and 29 Sep and 9 Oct inside it and 15, 16, 17, 18, 22, 24 and 30 Sep and 1, 2, 6, 7 and 8 Oct above it — none below it in the four working weeks since 11 September. No corroborating signal sits beside those three: no closure event for the Memphis FIR, the weather station keeps its reporting cadence, and Louisville (SDF) at 626 and 595 and Indianapolis (IND) at 529 and 566 — each within 10% of its 1–4 Sep average (586 for both) — read normal alongside a soft MEM on 9 and 11 Sep. So the weekday softness is specific to MEM, not network-wide. Observed counts are a floor, and the freshest days on this page have moved by hundreds of movements on re-read (10 Sep went from a published 393 to a withdrawn 13 to a restored 397); 12 Sep went from 79 to 203, 13 Sep from a measured 172 to a withdrawn gap to a restored 174, 14 Sep from a measured 170 to a withdrawn gap to a restored 180, 15 Sep from a measured 523 to a withdrawn gap to a restored 535, 16 Sep from a measured 536 to a withdrawn gap to a restored 539, 17 Sep from 510 to 518, 18 Sep from a measured 445 to a withdrawn gap to a restored 451, 19 Sep from 305 to 320, 20 Sep from 193 to 213, 21 Sep from 171 to 175, 22 Sep from 446 to 455, 23 Sep from 423 to 430, 24 Sep from 420 to 431, 25 Sep from 389 to 398, 26 Sep from 301 to 307, 27 Sep from a withdrawn hole to a restored 189, 28 Sep from 193 to 199, 29 Sep from a no-bar 9 to a backfilled 394 to a settled 405, 30 Sep from a no-bar 6 to a backfilled 511 to a settled 511, 1 Oct from 498 to 506, 2 Oct from a no-bar 6 to a backfilled 478 to a settled 487, and 3 Oct from a no-bar 4 to a settled 310; the 236, 250, 180, 535, 539, 518, 451, 320, 213, 175, 455, 430, 431, 398, 307, 189, 199, 405, 511 and 506 we published as measured have each held through their last nightly pass, and 487 and 310 stand where theirs left them. Every day inside the pulse window up to 11 October carries a measured bar; the one collection-health flag standing in it today, on 10 October, is the trailing-day artefact described above, and 12 and 13 October are not plotted. Every weekend day that has settled across the weekends since the flag matches the pattern of a quiet-weekend cargo hub.
  • What tripped the detector: 99 compared against the ~276 all-days average reads as −64%. Compared against other Sundays, the day is unremarkable.
  • Composition on 16 Aug: 66 arrivals vs 33 departures — an inbound-heavy Sunday. That skew is consistent with aircraft positioning in ahead of the working week, and we label that reading as interpretation, not measurement.

Observed counts are a floor, not a total — but an airport measured against its own same-weekday history is a reliable shape, and Memphis's shape is the deepest weekly saw-tooth we measure at any large US airport. Live view: MEM airport brief.

What it means for flights

What fired, and why we publish it anyway. Our detection layer flagged MEM when the 16 August observed count came in at 99 movements against an all-days average of ~276 — a −64% traffic anomaly at a major US hub. Automated multi-signal analysis then cross-checked the anomaly against every other layer we track: closure events, weather-reporting continuity, volcanic-ash exposure, GNSS interference, and same-day traffic at neighboring hubs. Nothing corroborated — and the same-weekday history explained the whole signal. We publish the result anyway, for the same reason we published the Monterrey coverage-loss anomaly: a raw −64% at a famous 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.

The weekly breath of a cargo superhub. Memphis International is the home base of a global express-cargo operation, and express networks run on an overnight sort cycle tied to the business week: banks of aircraft in and out every weeknight, and a much thinner schedule once the working week ends. In our data that shows up as the deepest weekly saw-tooth we measure at any large US airport — ~351 observed movements on an average Tuesday–Friday vs ~103 on an average Sunday (the five Sundays in the 30 days to 16 August), a ~71% trough repeating every week. Nineteen Sundays back to early June, every one of them measured on a normal collection day (the nineteenth provisional), all fall in the 34–213 range. A detector comparing any of those Sundays against an all-days average will 'find' a collapse — which is a statement about the baseline, not the airport.

The cross-hub corroboration: cargo share predicts the weekend trough. The same weekly signature, measured the same way over the same four weeks and each airport against its own weekday level: Louisville (SDF), another express-cargo hub city, drops −47% from its Tuesday–Friday average to its Sunday average; Indianapolis (IND), a major cargo station with substantial passenger traffic, drops −29%; Atlanta (ATL), the archetypal passenger hub, drops just −7%. The deeper an airport's cargo share, the deeper its weekend trough. Memphis, at −71%, is simply the extreme end of a spectrum — and that spectrum is why per-airport, per-weekday baselines are the only honest way to read observed counts. Re-measured at close-out over the four most recent weeks — Tuesday–Friday 15 September to 9 October against the Sundays of 20 September to 11 October, the last provisional — the order holds: Memphis −58% (~469 on a weekday, ~195 on a Sunday), Louisville −46%, Indianapolis −21%, Atlanta −1%. Every trough is shallower than the pinned window measured, and observed counts are a floor from a network whose coverage is not constant, so we read the order, not the exact depth.

The independent checks we ran. No airspace-closure event has registered for the Memphis FIR (KZME) in our detection layers from early June to this close-out on 14 October (a brief 1 June event has since been invalidated in our own review). MEM's weather station kept reporting on its normal hourly cadence throughout, adding special reports during a brief thunderstorm late on 11 September and through three days of low cloud on 2–4 October: 802 observations over the 30 days to this close-out, with IFR conditions 4% of the time; the latest METAR (14 Oct, 00:54 UTC) shows 10-mile visibility with a few clouds at 5,000 ft and a 6-knot southerly wind — VFR, as every report since 5 October. The GNSS-interference ratio for the Memphis FIR read 0.0001–0.0018 on every day from 1 to 12 October, at its measurement floor. Our NOTAM set holds no closure notice for the aerodrome; that set was last refreshed on 4 October, so it speaks for the period to that date and not after. And unlike the Monterrey case, this is not receiver-coverage loss: Atlanta, Chicago O'Hare and Dallas–Fort Worth all showed normal observed traffic on 16 August, and MEM's own count snapped back into its weekday pattern the following Monday-to-Friday cycle in prior weeks every time. Every layer is consistent with one explanation: a quiet Sunday at a cargo airport.

What we cannot say, and what changes this page. Observed counts are a floor, not a total — our receiver network does not hear every movement, and we make no claim about MEM's true schedule-level totals. This page also is not a verdict about any individual flight; the operating picture on a given day belongs to the airport, the airlines and the FAA. Two things would have changed this story: the working week resuming without traffic returning to the ~351 weekday norm — a genuinely anomalous weekday is a different and much more serious signal than a normal Sunday — or a corroborating signal (closure event, weather deterioration, interference) appearing in any layer. In the weeks after the flag neither happened (the post-Labor Day weekdays are logged below): Tuesday 18 August came back at 334 observed movements, Wednesday 19 August at 307, Thursday 20 August at 318 and Friday 21 August at 353 — all inside the normal Tuesday–Friday range (268–430 over the four full weeks to 16 August) — and no corroborating signal has appeared in any layer. (Corrections to what we published earlier, on the record. Saturday 22 August — initially a 19-movement thin-day read — settled at 335, its busiest Saturday of the 30-day window; Sunday 23 August, which we then published as a network-wide collection gap on our side, finalised at 128 — an ordinary Sunday, +24% against the ~103 Sunday baseline — and 25 August, which we flagged as a partial-collection day, finalised at 277. Wednesday 26 August, published here as 162 on a feed we then measured under median, finalised at 179 on a day whose global volume closed at ~94% of median: an ordinary collection day, so the feed explanation for that Wednesday is withdrawn. At 179 it is below the normal weekday range with no corroborating signal in any layer, and MEM measured 425 the next day — we leave it labelled as a soft weekday we cannot explain. Friday 28 August, published first as a collection gap and then as a thin-but-measured 148, finalised at 286 — an ordinary Friday inside the 268–430 weekday range. Sunday 30 August holds at 98 (global volume ~93% of median) — an ordinary Sunday, −5% against the ~103 Sunday baseline — and Monday 31 August holds at 78 on a normal collection day (~98% of median), the low edge of MEM's thin-Monday norm (77–157 across the window), −40% against its Monday median of 129, with no corroborating signal in any layer. Tuesday 1 September took four readings to settle — published as a measured 371, withdrawn as a collection gap, then restored — and stands at 376 on a clean day (~100% of its 14-day median, with Atlanta at 1,930 movements, Chicago O'Hare at 2,618 and Dallas–Fort Worth at 1,864, all normal): an ordinary Tuesday inside the 268–430 weekday range. Wednesday 2 September took three — published as a measured 344, withdrawn a day later when it read 4 movements on a global volume of ~15% of median, then restored at 347 on a normal collection day (~101% of median; ATL 1,903, ORD 2,478, DFW 1,819), an ordinary Wednesday inside the same range. Thursday 3 September is a normal collection day (~100% of median) and holds at 259 — the second soft weekday, just under the low edge of the 268–430 range and −26% against the ~351 weekday average, with nothing corroborating it in any layer; we label it rather than round it into the range. Friday 4 September — published first as a collection gap at ~9% of median, then as a half-heard day at ~68% with MEM reading 48 movements and no bar — settled at 331 on a clean day (~107% of median; ATL 2,081, ORD 2,523, DFW 2,068) and holds there: an ordinary Friday inside the 268–430 range, and the 48-movement reading was collection, not traffic. Saturday 5 September, wrong twice before it settled — published as a measured 272, withdrawn as a collection gap, then restored — holds at 276 observed movements on a genuine collection day (~94% of median; ATL 1,839, ORD 2,538, DFW 1,887): an ordinary Saturday, +44% against the ~192 Saturday average of the pinned window and inside the 150–335 span of every Saturday measured since 21 July. Sunday 6 September reverses our 9 September refresh. We published it as a measured 103, then withdrew it when that day read ~26% of median; it has now backfilled to a normal collection day (~90% of median; Atlanta 1,897, Chicago O'Hare 2,306, Dallas–Fort Worth 1,735) measuring 108 observed movements. We restore the bar — an ordinary Sunday, +5% against the ~103 Sunday baseline — and the count returned to fourteen measured Sundays, all then in the 34–153 range. Monday 7 September holds as a normal collection day (~100% of median; ATL 1,964, ORD 2,521, DFW 1,955) and reads 104 on every refresh since, up from the 101 published on 9 September — inside MEM's thin-Monday norm (77–157), −19% against its Monday median of 129. Tuesday 8 September holds at 32 observed movements on a normal collection day (~96% of median; ATL 1,747, ORD 2,252, DFW 1,673) — −91% against the ~351 weekday average, the deepest weekday reading on this page. That is the kind of weekday this section says would matter, so we checked it against every layer: the overnight sort-bank hours (02–05 UTC) carry 2 movements against 105 on 9 September, the weather station kept its hourly cadence in VFR, no closure event registered for the Memphis FIR, and Louisville (−46%) and Indianapolis (−25%) sat low the same day. It follows the US Labor Day holiday on 7 September; a hub week starting one night late is our interpretation, not a measurement. The weekdays measured after it have moved on re-read. Wednesday 9 September, published as a measured 233 and then withdrawn at ~14% of median, is now restored at 236 on a normal collection day (~102% of median; ATL 1,891, ORD 2,438, DFW 1,789), −33% against the weekday average. Thursday 10 September has reversed twice: published as a measured 393, withdrawn on 12 September at ~15% of median, then restored on a later refresh to a normal collection day (ATL 1,943, ORD 2,452, DFW 1,839) on which MEM measures 397 observed movements — an ordinary Thursday inside the 268–430 range, the first of the four post-Labor-Day weekdays inside MEM's normal weekday span so far (10, 23, 25 and 29 September). Friday 11 September holds at 250 on a normal collection day (ATL 1,692, ORD 2,502, DFW 2,142), up from the 243 we published on 14 September — −29% against the weekday average, with 50 arrivals against 200 departures; a thunderstorm crossed the field from 21:40 UTC, after the day's measured departure banks. Saturday 12 September has reversed twice: first withdrawn here as an ~8% collection gap, then published on 15 September as a soft 79 on a partly-heard day, it has now settled to a clean collection day (~95% of median; ATL 1,973, ORD 2,442, DFW 1,917) on which MEM measures 203 — an ordinary Saturday, +6% against the ~192 Saturday average and inside the 150–335 span of every Saturday measured since 21 July. The 79 was collection, not traffic, and the cargo-layer softness we reported beside it is withdrawn. Sunday 13 September has reversed twice: published on 15 September as a measured 172, withdrawn on 16 September when the day read ~59% of median, it has now backfilled to a normal collection day (~91% of median; ATL 1,952, ORD 2,445, DFW 1,869) on which MEM measures 174 — +69% against the ~103 Sunday baseline, the busiest Sunday on this page; the chart then carried fifteen measured Sundays, all in the 34–174 range. Monday 14 September has reversed twice: published on 16 September as a measured 170, withdrawn on 17 September when the day read ~47% of median, it has now backfilled to a normal collection day (~102% of median; ATL 2,100, ORD 2,468, DFW 2,024) on which MEM measures 180 — +40% against its Monday median of 129 and above the 77–157 thin-Monday norm, the busiest Monday on this page until 28 September. Tuesday 15 September holds at 535 on a normal collection day (ATL 2,057, ORD 2,318, DFW 1,874) — +52% against the weekday average and above the 268–430 range, with 175 movements in the overnight sort-bank hours. Wednesday 16 September reverses our 19 September refresh: published on 18 September as a measured 536, then withdrawn a day later when the day read ~32% of median with MEM at 57, it has now backfilled to a normal collection day (~102% of median; ATL 1,930, ORD 2,427, DFW 1,797) on which MEM measures 539 — +54% against the weekday average, above the range, and now the busiest day on this page, with 161 movements in the sort-bank hours. The withdrawal was our error: 16 September is a genuine collection day and the 57 was collection, not traffic. Thursday 17 September is restated upward: published on 19 September as a measured 510, it now reads 518 on a normal collection day (~104% of median; ATL 2,070, ORD 2,066, DFW 1,935) — +48% against the weekday average and above the range, with 157 movements in the sort-bank hours. Friday 18 September has reversed twice: published on 20 September as a measured 445, withdrawn on 22 September when the day read ~15% of median with MEM at 17, it has now backfilled to a normal collection day (~106% of median; ATL 2,139, ORD 2,613, DFW 2,020) on which MEM measures 451 — +28% against the weekday average and above the range, with 120 movements in the sort-bank hours. The withdrawal was our error: 18 September is a genuine collection day and the 17 was collection, not traffic. Saturday 19 September has firmed from a published 305 to 320 on a day that reads ~92% of median (ATL 1,815, ORD 1,784, DFW 1,897) — an ordinary Saturday, +67% against the ~192 Saturday average and inside the 150–335 span of every Saturday measured since 21 July. Sunday 20 September has firmed from a published 193 to 213 on a day that now reads ~89% of median (ATL 2,141, ORD 2,520, DFW 1,816) — +107% against the ~103 Sunday baseline, the busiest Sunday on this page, which then widened the chart to sixteen measured Sundays in a 34–213 range; it has left the re-read window, and its three passenger hubs no longer read below their own five-Sunday spans, so we publish it as settled. Monday 21 September has firmed from a published 171 to 175 on a day that reads ~94% of median (ATL 2,078, ORD 2,507, DFW 2,019) — +36% against its Monday median of 129 and above the 77–157 thin-Monday norm. Tuesday 22 September has firmed from a published 446 to 455 (ATL 2,136, ORD 2,511, DFW 1,972 all normal) — +30% against the ~351 weekday average and above the 268–430 range, with 80 movements in the overnight sort-bank hours, well under the 157–175 those hours carried on 15–17 September. The collection-health flag we refuted on that day has cleared: the nightly refetch refilled it, and the rollup and the detection record now both read 244,092 movements worldwide — 0.02% apart when we first refuted the flag — with 455 at MEM in both, so the 455 we published against our own flag stands. Wednesday 23 September has firmed from a published 423 to 430 (ATL 2,000, ORD 2,441, DFW 1,883; Louisville 611 inside a five-Wednesday span of 591–626, Indianapolis 496 just under its own 526–589) — +23% against the ~351 weekday average, exactly on the top edge of the 268–430 range, with 94 movements in the sort-bank hours. The flag we refuted on that day has cleared: the rollup now holds 251,617 movements worldwide for 23 September and 430 at MEM, matching the detection record to the movement, so the 430 we published against our own ~31% reading stands. Thursday 24 September is settled at 431 (ATL 2,072, ORD 2,411, DFW 1,960 each inside its own five-Thursday span; Louisville 643 inside 574–647, Indianapolis 492 inside 487–591) — +23% against the ~351 weekday average and, at 431, one movement above the top edge of the 268–430 range rather than inside it as we had it on the lower count, with 84 movements in the sort-bank hours. The flag we refuted on that day has cleared too: the rollup now holds 255,236 movements worldwide for 24 September (~105% of the 244,092 14-day median) and 431 at MEM, matching the detection record, so the 431 we published against our own ~15% reading stands. The flag we refuted on 25 September has cleared the same way: the nightly refetch refilled the day, and the rollup now holds 264,575 movements worldwide for it and 398 at MEM, matching the detection record exactly, so the 398 we published against our own ~15% reading stands. No collection-health flag stands in the pulse window today. Friday 25 September is settled at 398 — ATL 2,158 just above its five-Friday span of 1,692–2,139, O'Hare 2,437 inside 2,404–2,613, Louisville 612 above its own 575–595, Indianapolis 568 inside 480–572 — +13% against the ~351 weekday average and inside the 268–430 range, the third post-Labor-Day weekday to land in it, with 78 movements in the sort-bank hours; Dallas–Fort Worth (2,003 against 2,020–2,142) reads a little under its own five-Friday span and did not move on the refetch, so that is the day, not our collection. Saturday 26 September has firmed from a published 301 to 307 on a day that now reads ~98% of the 244,092 median (239,418 worldwide, the two records agreeing to the movement; ATL 1,909 inside its five-Saturday span of 1,815–2,026, O'Hare 2,407 inside 1,784–2,549, Louisville 492 above its own 438–461, Indianapolis 460 above 323–416) — an ordinary Saturday, +60% against the ~192 Saturday average and inside the 150–335 span of every Saturday measured since 21 July, with 57 movements in the sort-bank hours; Dallas–Fort Worth (1,856 against 1,860–1,917) still reads four movements under its own and did not move on the refetch. 20 through 29 September have all left the window our feed re-reads on the nightly refetch and the detection record moved on none of the first nine, so all ten are settled; 29 September firmed on its last pass. Sunday 27 September reverses our 29 September refresh, and the 'not our artefact' we wrote beside it is withdrawn: published as a hole at 19,389 movements worldwide with the rollup and the detection record agreeing, it has backfilled to a normal collection day (~96% of the median in the detection record, 233,963 worldwide; ATL 2,014, DFW 1,884, Louisville 339 and Indianapolis 476 each inside or just above its own five-Sunday span) on which MEM measures 189 — +83% against the ~103 Sunday baseline, the second-busiest Sunday on this page at the time and inside the 34–213 range, which took the chart to seventeen measured Sundays. The two records agreeing on a low count told us the rollup was not clipping; it did not tell us the movements were absent, and we no longer offer that equality as the sign of a real hole. The rollup clipped the day to 177 at MEM on our 30 September refresh; its last nightly pass, on 1 October, re-counted it in full and the two records now agree on 189, so the bar is settled — and Chicago O'Hare (2,101 against 2,197–2,520), still under its own span, did not move on that pass, so that is the day, not our collection. Monday 28 September has firmed from a published 193 to 199 on a normal collection day (~104% of the 14-day median, which on today's run is 239,418: 248,730 worldwide in the detection record; ATL 2,167, O'Hare 2,733, DFW 2,003, Louisville 247 and Indianapolis 442 each inside or above its own five-Monday span) — +54% against its Monday median of 129, above the 77–157 thin-Monday norm and the busiest Monday on this page, with 22 movements in the sort-bank hours; the rollup clipped it to 158 at MEM on our 1 October refresh, and its last nightly pass, on 2 October, re-counted it in full, so the two records now agree on 199 and the bar is settled. Tuesday 29 September has backfilled, as we said it might, and is now settled: published on 30 September with no bar at 19,615 movements worldwide and MEM at 9, then on 1 October as a backfilled 394 we called a floor, its last nightly pass on 3 October re-counted it in full to 222,287 worldwide (~93% of the median, both records agreeing) and 405 at MEM — +15% against the ~351 weekday average and inside the 268–430 range, with 97 movements in the sort-bank hours; Atlanta, Chicago O'Hare and Dallas–Fort Worth still read below their own five-Tuesday spans with the day out of the re-read window, so that is the day, not our collection. Wednesday 30 September has backfilled too, and is settled: published on 1 October with no bar at 20,172 movements worldwide and MEM at 6, its last nightly pass on 4 October left the rollup and the detection record agreeing on 254,563 worldwide (~106% of the median then in force) and 511 at MEM — +46% against the ~351 weekday average and above the 268–430 range, with 158 movements in the sort-bank hours. Thursday 1 October is settled at 506, firmed from the 498 we published on 3 October, on a normal collection day (250,150 worldwide, ~100% of today's median) — +44% against the weekday average and above the range, with 164 movements in the sort-bank hours; Atlanta, O'Hare, Louisville and Indianapolis read normal against their five-Thursday spans while Dallas–Fort Worth (1,379 against 1,839–2,007) stayed well below its own through its last pass, so that is the day, not our collection. Friday 2 October has backfilled and settled: published on 3 October with no bar at 20,176 movements worldwide and MEM at 6, then on 4 October as a backfilled 478, its last nightly pass on 6 October left the two records agreeing on 252,986 worldwide (~101% of the median) and 487 at MEM — +39% against the weekday average and above the range, with 128 movements in the sort-bank hours. Saturday 3 October has backfilled and settled the same way: published on 4 October with no bar at 17,976 worldwide and MEM at 4, it settled on 7 October at 228,908 worldwide (~92% of the median) and 310 at MEM — an ordinary Saturday, +61% against the ~192 Saturday average and inside the 150–335 span of every Saturday measured since 21 July, on a day the weather station spent largely in IFR conditions (16 of 33 reports). Sunday 4 October is settled at 183 (226,013 worldwide, ~90% of the median; all five cross-hubs inside their five-Sunday spans) — +78% against the ~103 Sunday baseline, the eighteenth measured Sunday and the fourth-busiest on this page, with 11 of 27 weather reports in IFR conditions that day. Monday 5 October is settled at 191 (~97% of the median worldwide) — +48% against its Monday median of 129, above the 77–157 thin-Monday norm. Tuesday 6 October is settled at 455 (~102%), Wednesday 7 October at 480 (~104%), Thursday 8 October at 468 (~106%) and Friday 9 October at 429 (~105%) — +30%, +37%, +33% and +22% against the ~351 weekday average, the first three above the 268–430 range and the fourth a single movement inside its top edge, with 110, 123, 103 and 90 movements in the sort-bank hours and every cross-hub inside or above its own five-weekday span on each of the four days. Saturday 10 October is plotted from the detection record as a floor — 343 at MEM (207 departures, 136 arrivals) and 221,968 worldwide, ~89% of the median, against a rollup our 13 October 21:25 UTC refresh clipped to 11 at MEM and 28,657 worldwide, which fires our collection flag; the five cross-hubs read inside or above their five-Saturday spans in the detection record, so the flag is our trailing-day artefact on the pattern logged since 22 September, and at 343 the day is the busiest Saturday on this page, +79% against the ~192 Saturday average; its last nightly pass is due on 14 October, after this page closed. Sunday 11 October is provisional at 193 (201,817 worldwide, ~81% of the median, no flag; all five cross-hubs inside their five-Sunday spans) — +87% against the ~103 Sunday baseline, the nineteenth measured Sunday and provisionally the second-busiest on this page; its last nightly pass is due on 15 October. Monday 12 October carries no bar (18,172 worldwide, ~7% of the median, MEM at 6) and 13 October has not been collected at all — the freshest movement in our feed at close-out is a landing at 23:59 UTC on 12 October — so neither is plotted; the live airport brief will carry them when they fill. High counts are not a disruption signal: across 15–18 September Louisville read 615, 618, 628 and 578 and Indianapolis 581, 526, 556 and 480 against 1–4 September averages of 586 for both — Louisville within 10% throughout, Indianapolis within 10% on 15 and 17 September and below it on 16 and 18. That leaves three of the twenty post-Labor-Day Tue–Fri days measured (8, 9 and 11 September, all in the first week after the holiday) below MEM's normal range, alongside 10, 23, 25 and 29 September and 9 October inside it and 15, 16, 17, 18, 22, 24 and 30 September and 1, 2, 6, 7 and 8 October above it; the condition this section names did not recur in the four working weeks since, so we flag the three soft days for what they were rather than explain them away. No corroborating signal sits beside them: no closure event for the Memphis FIR, Louisville (626 and 595) and Indianapolis (529 and 566) within 10% of their 1–4 September averages on 9 and 11 September, and normal traffic at Atlanta, Chicago O'Hare and Dallas–Fort Worth. More than fifteen days on this page have now been re-read at least twice — 8, 9, 10, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29 and 30 September among them, each moved since publication (10 September has gone 393 → withdrawn → 397, 12 September withdrawn → 79 → 203, 13 September 172 → withdrawn → 174, 14 September 170 → withdrawn → 180, 15 September 523 → withdrawn → 535, 16 September 536 → withdrawn → 539, 17 September 510 → 518, 18 September 445 → withdrawn → 451, 19 September 305 → 320, 20 September 193 → 213, 21 September 171 → 175, 22 September 446 → 455, 23 September 423 → 430, 24 September 420 → 431, 25 September 389 → 398, 26 September 301 → 307, 27 September withdrawn → 189, 28 September 193 → 199, 29 September no bar → 394 → 405, 30 September no bar → 511, 1 October 498 → 506, 2 October no bar → 478 → 487, 3 October no bar → 310); every day inside the pulse window up to 11 October carries a measured bar, and the one collection-health flag standing in it today, on 10 October, is the trailing-day artefact described above. That is what a settling feed looks like, and it is why every number here is published together with the collection health of the day it came from.) Automated multi-signal analysis re-checked all of them every three hours while this page was open, and the detection layer itself takes a lesson from this event: at airports with a measured weekly rhythm this deep, same-weekday baselines are the standard we hold our own alerts to.

What the data showed — close-out, 14 October. The number. Sunday 16 August measured 99 observed movements against a Sunday-only baseline of ~103: −4%. The −64% our detector reported was the same day against an all-days average of ~276, at an airport whose Sundays ran ~71% below its Tuesday–Friday level in the pinned window. Eight weeks of same-weekday history since. The eight Sundays after the flag — 23 August to 11 October, the last provisional — measured 98 to 213, an average of ~161, every one inside the 34–213 range of the nineteen Sundays on this page. The four most recent working weeks (Tuesday–Friday, 15 September to 9 October) measured 398 to 539 on all sixteen days, an average of ~469 against the ~351 of the pinned window (+34%); the last of them, 6–9 October, read 455, 480, 468 and 429. The three soft weekdays this page flagged (8, 9 and 11 September: 32, 236 and 250 observed movements) all fell in the first week after the Labor Day holiday, and no Tuesday–Friday since has read below the 268–430 range. We report the higher weekday level as measured and do not read it as growth: observed counts are a floor from a receiver network whose coverage is not constant. The rhythm at close-out. Over those four weeks MEM's Sundays (~195) sit ~58% below its weekdays (~469); on the same measure Louisville sits −46%, Indianapolis −21% and Atlanta −1% — shallower than the pinned window measured, and in the same order: the deeper the cargo share, the deeper the weekend trough. The other layers. No closure event for the Memphis FIR from early June to this close-out. No closure notice for the aerodrome in our NOTAM set, which was last refreshed on 4 October and so speaks for the period to that date. The weather station reported on its normal cadence on every day — 802 observations in 30 days, 4% of them in IFR conditions — including three days of low cloud on 2–4 October, during which the Saturday and Sunday counts (310 and 183) stayed inside their same-weekday spans. GNSS interference for the FIR read 0.0001–0.0018 from 1 to 12 October, at its floor. The record. More than twenty days on this page were published, withdrawn or restated at least once as our feed settled; every reversal is logged above, and every figure from 15 September to 9 October now stands at the value its last nightly pass left. The last three days before close-out are labelled for what they are: 10 October a recount on a day our refresh clipped, 11 October provisional, 12 and 13 October not yet collected. What this adds up to. A traffic anomaly against the wrong baseline, cross-checked and found to be the airport's weekly rhythm, with every later week confirming it. We publish no verdict on operations at Memphis; the operating picture belongs to the airport, the airlines and the FAA. What would reopen this page. A Tuesday–Friday count at MEM below the 268–430 range on a normal collection day with a corroborating signal in any layer, or a closure event or closure notice for the Memphis FIR or the aerodrome. Neither appeared while this page was open.

Passenger questions, answered from the data

Is Memphis airport closed or disrupted?

This page closed on 14 October with no signal in the layers we track indicating a disruption: no closure event for the Memphis FIR, no closure notice in our NOTAM set (last refreshed 4 October), the airport's weather station reporting on its normal cadence, and GNSS interference at its measurement floor. The −64% figure comes from comparing a Sunday against an all-days average at an airport whose traffic drops ~71% every weekend by design. For the status of a specific flight, check your airline or the airport's own channels.

Why does Memphis lose two-thirds of its flights every weekend?

Memphis is a cargo superhub, and express-cargo networks run on an overnight sort cycle tied to the business week — banks of freighters arrive and depart every weeknight, then the schedule thins dramatically once the working week ends. In our measurements MEM averages ~351 observed movements on a Tuesday–Friday and ~103 on a Sunday, the deepest weekly rhythm we measure at any large US airport. Cargo-heavy Louisville shows a −47% weekend trough on the same measure; passenger-hub Atlanta only −7%.

Did flights at Memphis really drop 64%?

The observed count on Sunday 16 August was 64% below MEM's all-days average — and almost exactly on its Sunday-only average (99 vs ~103, −4%). Every one of the nineteen Sundays since early June that landed on a normal collection day measured between 34 and 213 movements (the nineteenth provisional); 16 August was the eleventh, the twelfth — 23 August — measured 128, the thirteenth — 30 August — 98, the fourteenth — 6 September — 108 (restored after we withheld it on 9 September when the day read ~26% of median), the fifteenth — 13 September — 174 (restored after we withheld it on 16 September when the day read ~59% of median), the sixteenth — 20 September — 213, firmed from the 193 we first published and settled since, the seventeenth — 27 September — 189, restored after we published it as a collection hole on 29 September and it backfilled, the eighteenth — 4 October — 183, and the nineteenth — 11 October — a provisional 193. The drop is real arithmetic against the wrong baseline: at an airport with a ~71% built-in weekend trough, only same-weekday comparison distinguishes a normal Sunday from a genuine disruption — and this was a normal Sunday.

Is it safe to fly to Memphis right now?

We publish measured signals, not safety verdicts — operating decisions belong to airlines, airports and the FAA, who work from their own complete data. Nothing in the layers we track — closure events, weather reporting, ash exposure, GNSS interference — indicates anything unusual at Memphis; what our detector flagged was, on the evidence, the airport's normal weekly rhythm.

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 Memphis International fell to 99 movements on Sunday 16 August — 64% below its all-days average — but against its Sunday-only baseline of about 103 the day was normal: the cargo superhub sheds roughly 71% of its weekday traffic every weekend by design, the deepest weekly rhythm measured at any large US airport, according to flight-monitoring service FlySafe.zone.”

Chart: observed traffic at MEM
PNG · 1600px · attribution baked in
Observed daily movements at MEM — light version preview
Chart: advisory cadence
PNG · 1600px · attribution baked in
Observed daily movements at Memphis International (MEM) on each of the last 19 measured Sundays, 7 Jun – 11 Oct 2026 (11 Oct provisional) — 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 — Memphis (MEM) Traffic 'Drop' — a −64% Anomaly That Is the FedEx Superhub's Normal Sunday, August 2026", https://flysafe.zone/events/mem-traffic-drop-2026-08/, data as of 14 Oct 2026, 01:25 UTC · closed · MEM series settled through 9 Oct; 10 Oct a clipped day plotted as a recount; 11 Oct provisional; 12 Oct partial, not plotted; 13 Oct not collected; NOTAM set refreshed to 4 Oct — 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 ratios from crowdsourced aircraft-position data; closure notices from published NOTAMs. Method: /methodology/ · Press: /press/.

Related coverage