GNSS Interference East of Warsaw — 93% of Aircraft Flagging Degraded Navigation 70–140 km from Chopin, and a Steep Gradient Down to 47% over the City, September 2026

Why this brief exists — and why it is not the story the trigger told. Written 3 October. Our detection layer raises a GNSS candidate when a cell near a large airport crosses 60% of aircraft flagged on the latest day and the same cell was quiet a week earlier. The 52°N 22°E cell crossed the first bar on 21 September at 92.0%, and has stayed across it since — 91.4% on 22 September, 93.0% on 23 September, 97.7% on 24 September, 90.2% on 25 September, 94.3% on 26 September, 92.6% on 27 September, 91.0% on 28 September, 93.6% on 29 September, 94.2% on 30 September, 94.1% on 1 October, 93.1% on 2 October. It passed the second only on a technicality: the comparison day, 14 September, holds no data for any cell in the feed. Read against the actual series, the second condition fails outright — the cell has never posted a day below 87.6% since observation began on 10 August. We are publishing anyway, because the measured picture is worth having on record: a saturated, stable interference cell one hundred kilometres from a capital-city hub, with a gradient that a dispatcher can read. But we label it as what it is. There is no escalation here to report, and a reader who arrives expecting news of a sudden jamming surge over Warsaw should leave with the opposite understanding.

A gradient that runs east–west along the Warsaw–Brest corridor. Written 3 October. Interference has a footprint, and along the 52°N row it slopes rather than stops. On 2 October the flag rate ran 100% at 24°E (all 27 aircraft logged — a very thin sample), 92.9% at 23°E over Biała Podlaska and Brest (79 of 85, also thin), 93.1% at 22°E over the Siedlce corridor, 50.3% at 21°E over Mińsk Mazowiecki and Warsaw's eastern approaches, and 46.9% over Warsaw itself at 20°E (158 of 337 aircraft). The seven-day aggregates over 26 September–2 October say the same thing with less noise: 97.0%, 86.3%, 93.3%, 49.5%, 43.7%. The break between 22°E and 21°E is the sharpest step in the row — 43 points on 2 October and 44 on the week, in one degree of longitude, about 70 km. The two western cells are drifting up, and the honest reading is now a drift rather than noise — but not a step. Over 26 September–2 October 52°N 21°E ran 48.6%, 46.9%, 51.3%, 53.0%, 47.7%, 48.5%, 50.3%, and Warsaw's own cell 41.3%, 42.8%, 42.6%, 44.5%, 48.4%, 39.3%, 46.9%. Warsaw's cell has now posted the two highest days of its series eight days apart — 48.4% on 30 September (167 of 345) and 46.9% on 2 October — and its rolling week, 43.7%, is the highest it has recorded since observation began on 10 August, about a point above the 42.8% of 24–30 September and five points above its 38.3% September mean. The bar we set for calling this a change over Warsaw was a full week at 50% or more, and 43.7% does not meet it; neither cell has approached the 70–90% the eastern strip holds, and consecutive rolling weeks share six of seven days, so a record week that beats the last one by a point is a slow slope, not an inflection. South of Warsaw, the cell we put back on the watch list has stayed there. 51°N 21°E read 62.2% on 30 September (138 of 222), 50.0% on 1 October and 46.8% on 2 October (102 of 218) — four of the cell's eight highest days since 10 August fall in 27 September–1 October — and its seven-day aggregate of 51.8% over 26 September–2 October is again its highest rolling week on record, two points above the 49.9% we reported last time and eleven above its 41.0% September mean. The calendar week 28 September–4 October stands at 54.1% with five of its seven days in, which would be the cell's first week clear of 50%; we will call it either way once the week closes. On 218–250 aircraft a day a single reading still means little, but two record weeks and a calendar week at 54% are a sustained shift in that cell. Its western neighbour 51°N 20°E over Radom read 33.4% on 1 October and 36.4% on 2 October (100 of 275), against a 40.3% week lifted by the 47–49% run of 28–30 September; among the well-sampled cells it is still the largest month-on-month mover, from 25.1% across 10–31 August to 34.3% across September, and its two October days average 34.8%. North of the corridor the belt is denser still: the 53°N 22°E cell over Łomża and southern Podlasie read 97.4% (225 of 231) on 2 October and 98.0% across the week, and 53°N 23°E toward Białystok 97.6% across the week. South, toward Lublin, it thins: 81.1% at 51°N 22°E (Lublin's own cell) on 2 October, 75.7% across the week. In August-to-September comparison the eastern strip firmed slightly — 52°N 23°E from 79.9% to 85.8%, 51°N 22°E from 69.0% to 75.0%, 52°N 24°E from 84.9% to 95.8% on thin samples — while the Siedlce cell itself moved from 92.7% to 93.0%, which is no movement at all on a cell already near saturation. Warsaw's own cell drifted from 34.5% to 38.3%, and 52°N 21°E from 43.5% to 46.5%: a three-to-four-point creep toward the city, real in sign, still too small to call a trend on seven weeks of data.

FIR-average hides the strip — why the cell-level number matters. Written 3 October. By the FIR-wide measure, the whole Warsaw FIR (EPWW) sits at 3.9% over the last 7 days against a 3.7% 90-day average, reading 4.4% on 2 October on about 18,900 tracked aircraft that day (FIR-level, high confidence). The week has oscillated inside one band — 4.3% on 28 September, 3.4% on 29 September, 4.1% on 30 September, 4.0% on 1 October, 4.4% on 2 October, 3.4–4.4% across the seven days. The last of those is the week's high and the FIR's highest day since the 4.8% of 17 September, but it is a tenth of a point above the 4.3% of 28 September and still below the 4.5–5.6% the FIR posted on 10–12 September, so it is the top of the band rather than a break out of it. It is in any case a rounding error next to the eastern cells, as is the FIR's own September peak of 5.6% on 11 September, because the average pools a saturated strip along the Belarusian border with quiet central and western Poland, where the bulk of the FIR's overflights run. Zoom to the 52°N 22°E cell and the degraded share is about twenty-one times the FIR figure. Both numbers are true; only the cell-level one tells a dispatcher what the corridor east of Warsaw looks like today. The neighbouring FIRs at the same 7-day scale: Minsk (UMMV) ~23.6% against a ~23.2% 90-day average — on roughly 232 tracked aircraft a day, medium confidence, so indicative — and Vilnius (EYVL) ~19.9% against ~21.0%. Lviv (UKLV), to the south-east, is oscillating hard on a thin sample: ~6.6% over the week against a ~3.4% 90-day average, with 10.3% on 29 September, 5.2% on 30 September, 1.5% on 1 October and 10.8% on 2 October, all on about 130 aircraft a day at low-to-medium confidence. On that sample the swings are not separable from sampling, and we quote the cell-level numbers instead. The Kyiv FIR (UKBV) holds too few tracked aircraft in our feed — it has rows on only three of the last seven days, at zero confidence — to quote at all. One FIR-wide figure hides the strip; the strip alone overstates the FIR. We label which scale each number belongs to for exactly that reason.

Interference degrades navigation; it does not ground flights. Written 3 October. The figure counts aircraft broadcasting downgraded navigation integrity in their ADS-B position-quality fields — the consistent fingerprint of GNSS interference. It does not mean those aircraft are lost or off course. Airliners cross-check satellite positioning against inertial reference systems and ground-based aids — ILS, VOR/DME — and fall back to them when GNSS degrades, with published contingency procedures for GNSS loss. The real cost is precision and workload: satellite-based (RNP/RNAV) approaches can become unavailable, approach minima rise, and in poor visibility that can mean holding or a diversion to an alternate. That is exactly why Chopin's departure board shows nothing: the traffic section above reads an ordinary early-autumn profile — a September mean of about 626 observed movements a day, 665 on the busiest Sunday — while more than nine in ten aircraft over the corridor to the east flagged degraded GNSS. For Warsaw's own arrivals the relevant figure is the city cell: 46.9% on 2 October, nearly one aircraft in two, against 43.7% over the past week. For traffic routing east toward Belarus and Ukraine's closed airspace boundary, it is the 93.1%.

Honest arithmetic on a mid-sized cell and a seven-week series. Written 3 October. The Siedlce cell saw 170–212 tracked aircraft a day in September in our feed — large enough that a weekly aggregate on roughly 1,350 aircraft is a solid number, small enough that a single day can swing four or five points on sampling alone. That is why the 97.8% of 15 September (182 of 186) and the 97.7% of 24 September (168 of 172) are reported as peak days and not as the level, why the 90.2% of 25 September (184 of 204) and the 91.0% of 28 September (171 of 188) were not reported as declines, and why the 94.2% of 30 September (178 of 189), the 94.1% of 1 October (192 of 204) and the 93.1% of 2 October (190 of 204) are not reported as a recovery and a dip; we describe the state on the weekly and monthly aggregates rather than on any one day. A day with no data at all, like 14 September, is left out rather than filled in — and, as explained above, that same blank is what tripped the trigger. The series begins on 10 August, so it is 53 observed days deep and we make no claim about how long the cell has looked like this beyond that; the wider southern-Baltic and Belarusian-border belt has carried elevated interference for well over a year (background: Baltic GPS-jamming escalation and how GPS jamming works), and its Gdańsk end is covered in a separate active brief (Gdańsk, September 2026). The same discipline governs the traffic section, and it has cost us there repeatedly: Saturday 19 September was pulled from the chart on 22 September as a blank and turned out to be a day still being counted, since settled at 541; Tuesday 22 September was clipped to 286 by the rolling cut-off and plotted from a recount of 596, which the finished aggregate later confirmed to the individual flight; and five days in a row — 23, 24, 25, 26 and 27 September — were published as settled and each later gained between four and nine movements. Monday 28 September, published the same way at 630, did not move; Wednesday 30 September, published last time at 613 on the same aggregate-equals-recount test, has since gained nine and now reads 622, which makes six such days in two weeks. Two bars stay low for reasons that have nothing to do with Chopin: Tuesday 29 September, plotted at a measured 504 with a two-hour worldwide hole in our feed that four overnight passes have not refilled, and Tuesday 8 September, plotted since the first publication at 580 with a one-hour hole of the same kind. Each correction is stated in the traffic section rather than quietly applied, and we no longer present agreement between aggregate and recount as proof that a day is final — only that it is complete for now. The chart stops at 1 October, the latest fully collected day on that reading; 2 October has barely begun. Every figure here is an aggregate of crowdsourced receiver telemetry: a floor, not a total.

What would change the picture. Written 3 October. Three measured things are watched continuously by our automated multi-signal analysis. First, the western edge: the 52°N 21°E cell and Warsaw's own 52°N 20°E cell are the tell-tales — a step in those cells toward the 70–90% the eastern strip shows would put Chopin's approach paths inside the zone rather than on its margin. We said a week, not a day, at 50% or more over Warsaw would be a change. It has not happened, but the week is now the highest on record: 52°N 20°E reads 43.7% over 26 September–2 October against a 38.3% September mean, and 52°N 21°E 49.5% against 46.5%. Day by day, Warsaw's cell has posted the two highest readings of its series eight days apart, 48.4% on 30 September and 46.9% on 2 October, with 39.3% in between — which is why we hold to the weekly test rather than the daily one. The 50% week over Warsaw remains the marker. Second, south of Warsaw: 51°N 21°E, back on this list since 1 October, has set a second consecutive record rolling week — 51.8% over 26 September–2 October against 49.9% the week before and a 41.0% September mean — and its calendar week 28 September–4 October stands at 54.1% with five days in. We said a second week near 50% would make it a trend rather than a run of peak days; on the rolling measure that condition is met, with the caveat that consecutive rolling weeks overlap by six days, so the calendar week closing above 50% is the clean confirmation and we will report it next. Radom's 51°N 20°E read 33.4% and 36.4% on 1 and 2 October, inside its September range. Third, the FIR-level EPWW ratio breaking decisively above the 3.4–4.4% band it has held over the past week rather than oscillating inside it — the 4.4% of 2 October is the top of that band, a tenth of a point above the 4.3% of 28 September, not a break. A shift from jamming (signal loss) toward spoofing (a coherent false position), or interference reaching approach paths at WAW or WMI in low-visibility conditions, would also change the picture. So would regulatory action — NOTAMs suspending satellite-based approach procedures at Polish airports — which would be the clearest operational marker, and we flag its absence honestly: our NOTAM feed still carries no EPWW items at all, so its silence here is a gap in our coverage, not evidence that nothing has been issued. No traffic-based airspace closure has registered for the Warsaw FIR in our detection layer over the window. This page updates from the same monitoring cycle as the WAW brief and the live regional GPS-interference map; if the eastern cell drops below the band it has held, or Warsaw's cell climbs into it, the page will say so.

What the data showed — close-out, 13 October. This brief opened on 22 September to put on record that an automated trigger had mistaken a saturated steady state for a new interference zone, and to carry the series forward with its gradient toward the capital. It closes on 13 October with no recovery to report: the interference did not end, it held — and we close it as a documented steady state, the way the Riga brief on the same belt was closed. The trigger cell. From 3 to 12 October the Siedlce-corridor cell read 92.0%, 92.9%, 93.1%, 94.1%, 92.7%, 96.9%, 92.1%, 97.9%, 94.6% and 90.4% (188 of 208) on samples of 188 to 212 aircraft a day. The calendar week of 28 September–4 October closed at 93.0% (1,284 of 1,381) and the week of 5–11 October at 94.5% (1,289 of 1,364), the second-highest calendar-week figure of the series after the six-day week of 14 September; 1–12 October aggregates to 93.6% (2,237 of 2,389) against 93.0% across September and 92.7% across 10–31 August. Sixty-three observed days, every one between 87.6% and 97.9%; the series high, 97.9% (191 of 195), arrived on 10 October, a tenth of a point above the 97.8% of 15 September, and the final plotted day's 90.4% is the softest reading since 25 September and inside the band. There is no inflection in it, in either direction. The tests this page set, answered. Over Warsaw: the marker was a full week at 50% or more over the city's own cell, 52°N 20°E. It was never reached. The cell closed the calendar week of 28 September at 42.6% (1,044 of 2,448) and the week of 5 October at 39.5% (844 of 2,135); its record rolling week stayed the 43.7% of 27 September–3 October (1,059 of 2,421), and 1–12 October aggregates to 40.0% (1,519 of 3,801) against 38.3% across September and 34.5% across August. The slope we described on 3 October continued at about two points a month and did not steepen; the cell's highest single day, 48.8% (159 of 326) on 8 October, was followed by 39.9%, 33.5% and 33.7%. The cell between, 52°N 21°E, did post one calendar week at 50% or more — 50.5% (1,246 of 2,469) over 28 September–4 October, its highest window of the series — and fell back to 48.0% (1,130 of 2,353) the week after; 1–12 October 49.0% (1,980 of 4,044) against 46.5% in September and 43.5% in August. Both cells stayed between 37 and 64 points below the corridor on every October day; the step between 22°E and 21°E read 40 points on 12 October and 46 on the final seven days. South of Warsaw: the 51°N 21°E calendar week of 28 September–4 October, which stood at 54.1% after five days, closed at 50.1% (881 of 1,758) — the cell's first calendar week at or above 50%, by a tenth of a point — and the week of 5 October came back to 46.0% (899 of 1,956). Across 1–12 October the cell aggregates to 45.3% (1,516 of 3,345) against 41.0% in September and 36.0% in August, so the shift we called sustained is in the monthly figure, nine points above August, while the 50% week did not repeat; its highest single day, 67.6% (138 of 204), came on 8 October. Radom's 51°N 20°E read 35.5% across 1–12 October against 34.3% in September — level. At FIR scale: the Warsaw FIR (EPWW) posted one day above the 3.4–4.4% band, 4.71% on 10 October, its highest since the 4.78% of 17 September, and read 3.85% and 3.68% on the two days after — a one-day excursion, not the decisive break the page named. Over 6–12 October it averaged 3.9% against a 3.7% 90-day average on 17,700–19,800 tracked aircraft a day (FIR-level, high confidence); 1–12 October 3.8%, September 3.9%, 10–31 August 3.3%. Regulatory action: our NOTAM feed still holds no items for the Warsaw FIR at all, so its silence remains a gap in our coverage and not evidence that nothing was issued; no traffic-based airspace closure registered for the Warsaw FIR in our detection layer over the whole window, and the entries it holds for the Minsk FIR are single-day, low-confidence candidates, none of them open. Nothing in our feed distinguishes jamming from spoofing, so on that question we make no claim either way. The shape of the field at close. Along the 52°N row on 12 October, the final plotted day: 95.7% at 24°E (22 of 23, a very thin sample), 85.1% at 23°E over Biała Podlaska and Brest (74 of 87), 90.4% at 22°E, 50.3% at 21°E, 39.7% over Warsaw at 20°E (115 of 290), 43.0% at 19°E. Over the final seven days, 6–12 October: 98.3%, 86.6%, 94.1%, 48.4%, 39.7%, 39.0%. North of the corridor the belt stayed denser than the corridor itself — 98.5% at 53°N 22°E across 1–12 October, 97.8% at 53°N 23°E — and south toward Lublin it firmed: 51°N 22°E, Lublin's own cell, aggregated to 78.4% (1,640 of 2,091) across 1–12 October against 75.0% in September and 69.0% in August, and its final seven days, 80.4%, are the highest rolling window of its series. Month on month, eleven of the twelve cells we watched read higher across 1–12 October than across September, by between half a point and four points; the one that fell is 52°N 19°E, west of Warsaw, 42.0% to 39.4%. Two days stand out as broad, one-day swings across the whole row rather than a westward step — 8 October (series-high days at 52°N 20°E, 51°N 21°E and 51°N 22°E, 96.9% in the corridor) and 10 October (the corridor's 97.9%, the FIR's 4.71%, 100% at 53°N 23°E) — and each reversed the following day. Neighbouring FIRs over 6–12 October: Minsk (UMMV) 23.9% against a 23.6% 90-day average on 201–263 aircraft a day (medium confidence); Vilnius (EYVL) 20.6% against 20.9%; Lviv (UKLV) 7.7% against 3.9%, swinging between 2.5% and 20.0% on 83–154 aircraft a day, which on that sample is not separable from sampling; the Kyiv FIR (UKBV) holds one to four aircraft on the days it has rows at all, and we do not quote it. Traffic. Warsaw Chopin averaged about 610 observed movements a day across 1–11 October against about 626 across September's full days and about 621 across 23–31 August, and the week of 5–11 October, at about 603, is the softest seven-day window of the plotted run — its Monday, Thursday and Friday the lowest of their weekday in the window, its Tuesday the lowest full Tuesday, its Wednesday the second-lowest. That is a measured softening of two to three per cent, and the traffic section above sets out why we report it and do not attribute it: our worldwide feed ran at 97–106% of its 14-day median on those weekdays, the hour-by-hour scan found no hole, and the corridor cell read 94.5% that week, within two points of the 92.4–92.7% it held in the weeks Chopin ran 629–633 a day. The navigation layer did not move when the movement count did, and the movement count did not move during the seven weeks the navigation layer sat at 88–98%; whatever the softer week reflects, this data cannot say, and we do not guess. Modlin ran 83–119 observed movements a day across 1–9 October, the settled days; Lublin 5–11. Why close now, and what would reopen it. This page had been frozen since 3 October while the series kept moving; leaving an active-monitoring label on a page ten days stale was itself a form of inaccuracy, and the measured picture has not changed in nine weeks of daily observation. The close-out is a judgement about the page, not a claim about the signal: the corridor east of Warsaw is as saturated on 12 October as it was on 10 August. The same monitoring continues on the live regional GPS-interference map and the WAW airport brief; a week at 50% or more over Warsaw's own cell, the eastern cell dropping below the 87.6% floor it has held, or a NOTAM suspending satellite-based approaches at a Polish airport would each be grounds for a new brief rather than a reopening of this one. Every figure here is an aggregate of crowdsourced receiver telemetry: a floor, not a total; 14 September is a gap in the per-cell feed, 29 September a partial day in the movement feed, and both are left as measured.