Matthew Halaba

OSINT report · 22 November 2026

Kaliningrad air-raid alert, 21 November 2026: what we actually know

A drone strike on the Sparta II, sirens across the oblast and “explosions in Baltiysk”. Verifying the photos and videos, version 2.

Window: 20 Nov 2026 22:00 to 22 Nov 2026 08:30 CET. All times are Polish time (CET, UTC+1). Kaliningrad is on UTC+2. Telegram and X timestamps are UTC and have been converted.

Confidence levels: Fact at least two independent layers; Claimed a single source or an interested party; Assessment my own inference from the data; Speculation; No data.

Open sources only, with no logins and no accounts created. I do not publish faces or addresses of private individuals. Frames from third-party material are short quotations, each credited.

Prepared by: Matthew Halaba ·

About this report

Title
Kaliningrad air-raid alert, 21 November 2026: what we actually know
Window
20 Nov 2026, 22:00 to 22 Nov 2026, 08:30 (Polish time, CET)
Date
Version
2 (after frame verification, with corrections)
Prepared by
Matthew Halaba
Method
Open sources only, no logins, no accounts created. No face recognition, no data on private individuals

Download the report (PDF) 9 pages · 0.7 MB · English

Pobierz raport (PDF) 9 strony · 0.7 MB · polski

Polska wersja raportu

All reports

Summary

The only confirmed attack that day was a strike by two drones on the ro-ro vessel Sparta II, bound for Kaliningrad, in neutral waters of the Gulf of Finland in the morning Fact.

Sirens sounded across the oblast at around 18:00–18:15 CET, and Khrabrovo airport accepted flights only with clearance for about 40 minutes Fact. At 18:36 CET Governor Besprozvannykh wrote of a “suspicious object from the direction of Lithuania” over the Neman district and said there was “no threat” Fact (as worded). Lithuania denies it, and its air force detected no drones Fact.

Later that evening Ukrainian channels posted videos captioned “Baltiysk”: tracer fire and air defence at work. None of the ones I checked shows an impact or a fire caused by one Assessment.

The most viral “photo of the Baltiysk explosion” is certainly not from 21 November Fact: the same scene has circulated since 8 November, when Iranian and pro-Iranian channels captioned it as smoke over Qeshm Island. That it is Qeshm rests only on their word Claimed. A video described as Baltiysk actually shows the new High Bridge under construction in central Kaliningrad Fact.

The Russian MoD mentions the oblast neither in its daytime report for 21 November nor in its overnight report for 22 November. FIRMS shows no hotspot at Baltiysk, and no Sentinel satellite has passed over the port since the event Fact.

Conclusion: air defence was active at Baltiysk, at least gun systems Assessment. An attack with real effects cannot be confirmed No data. From the first minute, the Russian line has blamed Lithuania.

Corrections after verification

The first version (early on 22 November) contained conclusions that were wrong or overstated. I correct them here:

WasNowWhy
The “Baltiysk” photo is near-identical to the @HormuzReport post (pHash 6/64)Same fixed scene and viewpoint, but a different frame: the smoke column sits elsewhere. Scene FactA SIFT homography aligns the building, bushes, ridge and power lines, while the smoke does not. Smolii vs 8 Nov: 31 inliers, NCC 0.777, edge NCC 0.744. Controls: 0.23 and −0.05
Source: @HormuzReport, 8 Nov 21:40 CETEarliest found: Al-Mehwar News, 8 Nov 20:42 CET, watermarked. HormuzReport is a secondary copy. Original author: No dataAl-Mehwar: 642 inliers, NCC 0.927. @xxoperatpyz at 21:39 CET carries the same watermark. HormuzReport was created on 20 May 2026
Location: Qeshm FactQeshm Claimed, with doubtsOnly Iranian and pro-Iranian outlets report strikes on Qeshm on 8 Nov. FIRMS for 5–11 Nov shows no event anomaly. Lush greenery and a wooded ridge do not fit an arid island
Smolii first (19:03 CET)Earliest found: Sternenko, 19:02:03 CET, the identical file (pHash 0). INSIDER UA posted before 19:14, but the post is gone No datat.me/ssternenko/66372; INSIDER UA screenshot in voenacher/98970
Kaliningrad bridge AssessmentFactConcrete towers, steel lattice lift towers, the blue-and-white fence and site containers match reference photos from RG and Novy Kaliningrad
Sparta II in the footageFact, with extra evidenceThe name СПАРТА II is legible on the hull

Verifying the “Baltiysk explosion” photo

What I checked. I compared the versions posted by Smolii and Incident Kaliningrad on 21 November with the copies from 8 November (Al-Mehwar, @xxoperatpyz, @HormuzReport), using SIFT feature matching after contrast equalisation (CLAHE), a RANSAC homography, and normalised cross-correlation on both the image and its edges. Unrelated frames from the same evening served as controls.

Result. The fixed elements (building, bushes, the ridge on the horizon, the wires) line up after the homography. The smoke column does not, so this is a different frame of the same scene, probably from the same burst of shots Fact. The controls give NCC close to zero. Smolii’s version is the very same file as Sternenko’s (pHash 0, 800×708).

First publication. The earliest copy I found is Al-Mehwar News, 8 November at 20:42 CET, captioned “Columns of smoke rise from Qeshm Island after the American aggression”, with its watermark. Then came @xxoperatpyz (21:39 CET) and @HormuzReport (21:40 CET, 107k views). I could not identify the original photographer No data. I found no copy of the Smolii frame from before 21 November (Yandex returns only look-alikes from Lebanon, Gaza and Donetsk).

Is it Qeshm? Strikes on Qeshm on 8 November are reported by Al-Mehwar (from 20:17 CET), IRNA via Anadolu, and Iran International Claimed. The FIRMS archive (VIIRS 375 m, NASA Worldview) for 5–11 November shows no event-related anomaly: the points around 55.5–55.9 E appear every day, so they are industrial sites No data. Lush grass, broadleaf shrubs and a wooded ridge are unusual for arid Qeshm. They do not fit flat Baltiysk in November either Assessment.

How it spread on 21 November. Sternenko at 19:02 CET (“alert in the oblast”, with this photo), Smolii at 19:03, and an INSIDER UA screenshot reposted by Voenacher at 19:14. At 19:36 Sternenko reported a “hit in Baltiysk”, then added that the first photo was fake (NV puts the retraction at about 19:44 CET). The Russian side (Incident Kaliningrad at 20:04, the governor’s spokeswoman Bashkirova, RIA, TASS) used the debunk to dismiss the whole episode as fake.

A: Al-Mehwar copy from 8 Nov (short quotation, source t.me/alomhoar/115796). B and C: the Smolii and Incident Kaliningrad versions from 21 Nov overlaid on A after homography (cyan and red). The fixed scene matches; the smoke does not. Fact: the photo is not from 21 November.
Fig. 1. A: Al-Mehwar copy from 8 Nov (short quotation, source t.me/alomhoar/115796). B and C: the Smolii and Incident Kaliningrad versions from 21 Nov overlaid on A after homography (cyan and red). The fixed scene matches; the smoke does not. Fact: the photo is not from 21 November.
Copies side by side: Smolii 21 Nov, two @HormuzReport frames from 8 Nov, Incident Kaliningrad 21 Nov. Short quotation for verification purposes; sources as captioned.
Fig. 2. Copies side by side: Smolii 21 Nov, two @HormuzReport frames from 8 Nov, Incident Kaliningrad 21 Nov. Short quotation for verification purposes; sources as captioned.

Timeline (CET)

Time CETEventSourceLevel
8 Nov 20:42Al-Mehwar News posts the smoke photo “over Qeshm”, the earliest copy foundalomhoar/115796Fact
21 Nov morningTwo drones hit the Sparta II (Russian flag, US-sanctioned) in the Gulf of Finland, 3 miles from the Russian borderDrozdenko via RIA; besprozvannih/6750; ReutersFact
21 Nov 16:16Oblast government: crew evacuation “was not needed”. Finland and Reuters: fire, drift, 16 people evacuatedgovernment_kgd/8688; ReutersAssessment
21 Nov ~18:00–18:15Sirens in Kaliningrad and several towns. Text messages warning of possible mobile internet limitsKlops, Kaliningrad.ru, BazaFact
21 Nov 18:31–18:56Rosaviatsia: Khrabrovo by clearance only, then back to normalmash/79557; vestiruFact
21 Nov 18:36Governor: “suspicious object from the direction of Lithuania”, “no threat”besprozvannih/6752Fact
21 Nov 18:52–20:13A series of “Baltiysk” videos: tracers, air defence, shelterssupernova_plus, exilenova_plusClaimed
21 Nov 19:02Sternenko posts the 8 Nov photo as the oblast; at 19:36 a “hit in Baltiysk”, then an update: fakessternenko/66372, /66376Fact
21 Nov 19:03Smolii: “explosions in Baltiysk”, with the same photosmolii_ukraine/182111Fact
21 Nov 20:04Incident Kaliningrad: the photo is Qeshm. Bashkirova debunks it on her channel (RIA)incidentkld/39067; RIAFact
22 Nov 03:32Lithuanian FM Budrys: Lithuania never allowed use of its airspace. No drones detectedRIA; Daily BalticFact
22 Nov 06:36Russian MoD overnight: 170 drones, no mention of KaliningradRIAFact
22 Nov 06:55RVvoenkor: a truncated repost of the 21 Nov text, not a new alertRVvoenkor/127724Fact
22 Nov morningMeduza: sirens “for the first time since the start of the war”. DW reported an alert in the oblast on 25 MayMeduza; DWAssessment
22 Nov morningYle via Euromaidan Press: Sparta II crew evacuated, fire still burning, tow to Russian watersEuromaidan PressClaimed
22 Nov 08:01TASS: Finland allows for Kyiv’s involvement in the Sparta II attackincidentkld/39073Claimed

Other frames

Video with sirens: the High Bridge construction site in Kaliningrad. Other channels captioned it as Baltiysk. First posted by Supernova+ 63316 at 18:28 CET as “Königsberg”. Fact
Fig. 3. Video with sirens: the High Bridge construction site in Kaliningrad. Other channels captioned it as Baltiysk. First posted by Supernova+ 63316 at 18:28 CET as “Königsberg”. Fact
Tracer fire over low-rise buildings (Exilenova+ 35001). Consistent with air-defence guns. Location not confirmed. Assessment
Fig. 4. Tracer fire over low-rise buildings (Exilenova+ 35001). Consistent with air-defence guns. Location not confirmed. Assessment
Air defence over a housing estate (Supernova+ 63325, later Astra), attributed to Baltiysk. Claimed
Fig. 5. Air defence over a housing estate (Supernova+ 63325, later Astra), attributed to Baltiysk. Claimed
Dnipro Osint graphic with a pin on Baltiysk port. A graphic, not a geolocation of any video. Claimed
Fig. 6. Dnipro Osint graphic with a pin on Baltiysk port. A graphic, not a geolocation of any video. Claimed
Sparta II after the attack, name visible on the hull (Supernova+ 63338). Fact
Fig. 7. Sparta II after the attack, name visible on the hull (Supernova+ 63338). Fact

Versions and narratives

  • “An object from the direction of Lithuania”. Lithuania is blamed from the outset, echoed by TASS, Vesti, SolovievLive, Readovka, RV and Baza Claimed. Vilnius denies it Fact. RIA undercuts the denial with the word “allegedly”.
  • “The Baltic states lend their airspace”. Swodki on 22 Nov at 05:06 CET calls for a “response” and mentions a “nuclear response” Speculation.
  • The whole episode is fake. The Russian side debunks the 8 Nov photo and stays silent on the air-defence videos. Assessment: selective debunking.
  • The Ukrainian side. “First attack on the oblast” (UNN, Moscow Times, NV) and “the authorities called it a drill”, which the governor’s statement does not say. Assessment: premature and misattributed.
  • Sparta II. Oblast government: no evacuation needed. Finland and Yle: evacuation and fire. A contradiction Fact.
  • Poland. No Russian channel links the alert directly to Poland. Vesti runs an anti-Lithuanian series and items on “Polish claims” Assessment.

Technical layers

LayerResultLevel
FIRMS, oblast, 21–22 NovNo detections at Baltiysk. The only point in the oblast is a permanent industrial site at 54.755 N 20.332 EFact
FIRMS, Qeshm 5–11 NovNo anomaly tied to 8–9 NovNo data
Sentinel-1Last pass 21 Nov 17:27 CET, before the alert. None since (as of 22 Nov 08:35 CET)Fact
Sentinel-2Only 19–20 Nov, before the event, 41–85% cloudNo data
Image alignment (SIFT, NCC)Fixed scene matches, smoke shiftedFact
Reverse image searchYandex: look-alikes only. TinEye blocked; Bing and Lens need JavaScriptNo data
MetadataTelegram strips EXIF. mp4 creation_time: 18:10–19:52 CETFact
DuplicatesAbout 20 unique videos across 81 media postsFact

Gaps

  • No statements from the Baltic Fleet, Peskov or Zakharova (as of about 08:30 CET on 22 Nov).
  • No images of damage and no satellite imagery after the event.
  • VK and MAX not checked (login required). I could not find Bashkirova’s post in the public preview; I know it from RIA.
  • The photographer and location of the original 8 Nov photo remain unknown.
  • The INSIDER UA post with the photo has been removed, so its time is unknown.

Sources