the home of online investigations

You can support the work of Bellingcat by donating through the following link:

Geolocating the MH17 Buk Convoy in Russia

September 29, 2014

By Veli-Pekka Kivimäki

Over the last few weeks Bellingcat has been examining footage of a Buk missile launcher filmed inside Russia that matches the Buk missile launcher linked to the down of MH17. The movements of the Buk launcher unit linked to the downing of MH17 were previously examined on Bellingcat, and the reporting relied on several videos of a convoy moving through Belgorod oblast in Russia. Further analysis of the videos yields information which can be used to accurately locate where the videos were filmed.

The first video examined was one posted on Instagram, with the location tagged as Neznamovo (Незнамово), south of Stary Oskol and uploaded on June 23rd. First observation upon watching the video is that at the beginning of it the vehicles of the convoy are on a road slightly elevated compared to the vehicles where the video was filmed from, sloping downwards as the clip progresses and finally merging with the main road.


Towards the end of the video, an important clue is revealed: the sign with the letters ‘ОГК’ and ‘АГЗС’, the latter being a Russian shorthand for a gas station. Searching for these terms quickly yields results to the proximity Stary Oskol and Neznamovo, and a Wikimapia link gives us a location where the area features (roads, road markings, buildings) are a strong match to the video. The elevated stretch of the road is an on-ramp to road P-188 heading south. Street View being available for this location, the site can be confirmed with confidence. [Coordinates: 51.233946, 37.940584]


The second video referenced was posted to VK on the morning of June 24th, said to show the convoy in Alexeyevka (Алексеевка), also referencing ‘Magnit’, a major Russian supermarket chain. This narrows down the search area, and section of the road where the trucks are turning give a good reference point to look for in satellite images. A Yandex map search for “Алексеевка Магнит” for gives list of seven locations, one of which is next to a section of road which resembles the area where the trucks can be seen turning.


A closer look confirms this location as a strong candidate, as buildings opposite to the reported Magnit location seem like they could match what was shown on the video. Checking Google Maps reveals Street View is available at this location, making verification easy: the location is identical to what can be seen on the video. [Coordinates: 50.624196, 38.649911]


A third video, reportedly from Stary Oskol, seems to have been filmed at dusk or dawn. Viewing the metadata for the video reveals the upload time to be 04:16 ‘Zulu time’, another term used for Coordinated Universal Time, a.k.a. UTC. During the summer months Ukraine is on daylight savings time, meaning the time difference to UTC is +3 hours, so the local upload time of the video has been 07:16 in the morning of June 24th. The video alone provides a few clues about the shooting location, but it’s useful to first look at the other videos to narrow down the search area.

Video number four shows the convoy of vehicles during the daytime, reportedly at Krasnensky district of Belgorod oblast. The scenery in the video appears to show countryside with scarce buildings, but it does offer some details for analysis. First, there are fairly long stretches of straight road, with a curve at around 1 minute into the video. Near the curve, a pair of bus stops is visible, as well as an intersection and a tall structure to the right. Two minutes in, two large structures are visible to the right of the road.


As the Krasnensky district presents a fairly large search area, the road shape gives a good starting point for narrowing the possibilities down. Looking at the main roadways of the area, a stretch of road near Raskhovets (Расховец) appears to be a possible match. Overview satellite imagery also shows two large structures along the road, and zooming in on the satellite imagery of this area reveals two promising details: a pair of bus stops near an intersection, as well as a tall structure casting a long shadow. As there’s Street View images available of the location, these findings can be verified, and the location definitely matches the video. [Coordinates: 50.902533, 38.458406]


As the third video’s upload time was morning on June 24th, it likely was shot sometime between the Neznamovo and Raskhovets videos. With the knowledge of the shooting location of the previous videos, it’s possible to make estimations of the routes the convoy may have taken between Neznamovo and Raskhovets.


One such suggested route goes through the village of Gorodishche (Городище) in the Starooskolsky district. In the video, some small buildings are visible on both sides of the road, mixed together with trees. Gorodishche matched this description, warranting a further look.


Fortunately, this route is covered by Street View, so a virtual drive-through of the village is possible. One unique feature clearly visible at the beginning of the video is an arch on the building next to where the video was shot, and scrolling through the Street View images, on the east side of the village a building can be found with arches similar to the video. A closer look at the site confirms this to very likely be the shooting location of the video: the fence and buildings on the opposite side of the road match the video, the intersection briefly visible can be seen, and from the frames where the camera is pointing towards the convoy, a nearby church’s dome is visible, with the buildings and treeline also matching the video. [Coordinates: 51.137286, 38.064599]


While not referenced in the published article, Magnitsky had located another video of the same convoy, again from Stary Oskol. The video is fairly short and shaky with lots of panning, but the units are clearly identifiable. With the footage, it was possible to create a rudimentary panorama to get a better sense of the overall area where the video was filmed.


From the footage, a few key details stood out: on the very left of the scene, rails and cables are visible. This seemed to indicate the footage might have been filmed near a light rail or tram station. English Wikipedia confirmed that there was one operating in the city, and Russian Wikipedia provided more details on this. It became apparent the Stary Oskol light rail operated on a single rail line, so checking the assumption would be fairly simple: find the end of the line, and follow it down to see if there are any matching locations.

Following the rail line south, a location can be found on Street View where the brick wall and the building on the right appear to match the video. Further orienting the camera confirms the features match the site exactly, with the video seemingly filmed from the light rail station. [Coordinates: 51.311605, 37.897013]



Exploring the city streets on Street View also gave an indication that the Instagram image of the convoy might be from along the same route. In fact, the picture can be located to approx. 1km of the previous video, based on the red apartment building and pavement features. The picture appears to have been taken from the Eldorado store parking lot. [Coordinates: 51.323770, 37.882489]


Further searches for additional material from the same dates has uncovered another video from Kursk, reportedly from June 23rd, the same date the convoy had been filmed passing through Stary Oskol. The dashcam video is of poor quality and many details are not visible, but composition of the convoy gives a strong indication this may indeed be the same one. In fact, compared to the convoy seen near Raskhovets, the first 16 vehicles seem to be in the same order.


To locate this video, there are a few clues: reported location (Kursk), treelines on both sides of the road, a large blue sign on the side of the road, a crosswalk and a small reddish object on road’s wide shoulder on the right. From overhead satellite imagery, it can be seen that the road A144 on the east side of Kursk is lined by trees, and zooming in on this area further reveals crosswalks on this road. One section of this road has a wide shoulder near an intersection, and having a look with Street View again reveals matching details: a wide shoulder as in the video with a dark orange colored bus stop and a crosswalk next to it, and looking in the opposite direction a big blue sign on the side of the road. [Coordinates: 51.722592, 36.336530]


Putting all these data points together we can plot these points on a map, which would fit on a possible route from Kursk to Alexeevka. Compared to the shortest routing suggested by Google, only one manual adjustment to the route has been made in Stary Oskol, which adds 1km to the route, but avoids smaller roads and turns. The green markers show the locations at which the BUK unit had been filmed as part of the convoy, and the yellow markers show where other vehicles of the same convoy were filmed.

Veli-Pekka Kivimäki

Veli-Pekka is a doctoral student at Finnish National Defence University, researching social media and open source intelligence. He has a long background in the technology industry, more recently focusing on defense research.

Join the Bellingcat Mailing List:

Enter your email address to receive a weekly digest of Bellingcat posts, links to open source research articles, and more.


  1. Rob

    Thank you Veli-Pekka and team at Bellingcat, for this awesome work in tracking down the MH-17 BUK system. I’m a great admirer of your work, and hope that by crowd sourcing we can shed some light on this disaster.

    One question I have :
    You made a pretty compelling case that the convoy (most notably with the important BUKs) went through Gorodishche [Coordinates: 51.137286, 38.064599].

    But in doing so, it will approach Alexeyevka from the North, and it would be almost impossible for the convoy to go through the Magnit store intersection in Alexeyevka, which you would only get through if you were to approach Alexeyevka from the south…

    Is this convoy doing some sort of “parading” ?

  2. Rob

    I’m sorry. I should have been more clear :
    In the Magnit store Alexeyevka intersection video, both BUK 232 and suspect BUK 3’2 come from the right (the south) which suggests they took highway P185 to Alexeyevka. But if they did, they should not have come through Gorodishche. Isn’t that odd ?

    • Veli-Pekka Kivimäki

      The Alexeyevka scene is quite interesting, don’t have a solid explanation what happened. Seems the convoy had either split up and was regrouping, or some kind of reordering was going on. It’s also the last known video currently of this convoy together.

      • Rob

        A re-ordering makes sense ; Russian military may often do that to avoid satellite tracking of individual vehicles.

        The Alexeyevka intersection video is still very interesting though.

        It seems that most vehicles (including TELAR 232 and suspect TELAR 3’2) are moving towards the train station (along with TELAR 221, 231, loader 323), while loader 223, TELAR 211 and 212 are moving back into the direction that the convoy came from…

        • Alex Livesy

          suspect TELAR 3’2 , numbers 312 and 302 are Ukrainian owned, I haven’t seen any other numbers beginning with 3 and ending in 2 on either side.

  3. Rob


    The fact that the Alexeyevka scene appears to be the last scene where these BUKs are observed bothers me.

    You and the team at Bellingcat found some 6 videos taken of this convoy between Kursk and Alexeyevka. And then this whole convoy seems to disappear…

    Where was this BUK between June 24 and July 17 ?
    Why can we not find any videos that carry any one of these BUKs after Alexeyevka ?

    Something really weird is going on.

    If so many people are video-taping BUK movements, why do we not see any other appearance of BUK 3’2 until it shows up in Donetsk on the morning of July 17, on a picture of ParisMatch, who themselves incorrectly state the place and the time when that picture was taken (suggesting that they did not take that picture themselves, but somebody provided it to them).

    Something really weird is going on.

    And did anyone notice that the number on this BUK found in Donetsk and linked to the 53rd BUK brigade in Kursk is (possibly) 312 ?

    The same number of the Ukrainian BUK, a picture of which was used to implicate Ukraine ?

    Specifically, the Ukraine Interior Ministry released a picture of a BUK with number 312 which was supposed to show the missile system being trucked into Ukraine, while in fact it was very quickly established that this was a Ukrainian BUK in Ukrainian territory.
    And then used in Russian media to implicate Ukraine.

    How did these people that control the BUK venturing into Ukraine know that its number was 312 ?

    Something really weird is going on.

    And then there is the video, also released by the Ukraine government, of the BUK allegedly being trucked out of Ukraine.
    That video was quickly used by Russian media to implicate Ukraine as well ; that it was taken in an area under Ukraine control.

    But, after some serious crowd sourcing by bloggers, the video was geo-located in Luhansk, under Russian “rebel” control.

    And, yes, as the first one ever, you linked the BUK in Donetsk to exactly that BUK on that video in Luhansk (THANK YOU!), because of that spot on the right-side of BUK 3’2.

    That is VERY interesting, since that video (apparently first used to implicate Ukraine, but turned out to be Russian controlled) has MANY issues.

    For starters, that video, allegedly taken at sunrise on the 18th of July, (after the MH17 crash) was released July 18, but has an mp4 download creation date of 9 am UTC, July 19 which (and with the 1 day MPEG standard mix-up on YouTube) suggests a creation date of July 20. Which is obviously wrong, so the one who took that video did not want us to know when exactly it was taken.

    And July 18th in the morning, Lugansk had solid overcast, but the light on the video suggests it was clear sky right at sunrise…

    Something really weird is going on…

  4. Rob

    To add to the mystery, BUK 3’2 becomes public again only during its journey from Donetsk to Snizhne. Three photographs and two videos, plus a couple of “eye witness” accounts. 7 accounts between 9 am July 17 and 1:30 July 17.

    Nothing before (until your video from July 24), nothing after.

    Last account is 1:30 in Snizhne, with a video showing the BUK driving to its launch position.

    They arrive there, on a field close to Pervomaiske, just south of Snizhne, maybe at 1:45 or so.

    So do they power up the missile system and start staring at their radar, watching for Ukraine planes ?
    No. They don’t.
    They just wait. Maybe they drink some water, and have their lunch.
    Then, they wait some more.
    Maybe they tell jokes, or more likely, they watch out for Ukrainian fighters that they KNOW are aware of their presence in Pervomaisk.

    Still they wait for more than TWO HOURS !

    We KNOW they wait that long, since if they would have switched their radar on earlier, they would have seen MANY other “high birdies” passing overhead :
    flighttracker24 shows 4 international flights flew almost over their head during the last hour before MH-17 alone !

    In fact, they must have switched their radar on AFTER 3:50, or else they would have detected (and shot down) KC904, not MH-17.

    So after more than TWO HOURS of waiting, they finally switch on their radar a few minutes before MH17 comes over, a blip appears, and they press the button…. and MH17 goes down.

    Something smells REALLY WRONG with this whole picture.

    • Seb

      I remember reading that they had “insider informations” passed to them that a transport plane would be coming along that route. Maybe they stayed offline till that specific timeframe to avoid detection, and then just aimed for the first plane that seemed like a transport aircraft.

      • Rob


        There is quite clear evidence that the crew in the BUK in Snizhne were waiting for a call from an outsider rather than switch on their own radar :

        Igor Bezler received a call from “Naimanets”, 2 minutes before MH17 went down :

        Now, first of all, you may ask yourself why they were relied on an outside call rather than simple switch on their radar and scan the sky.

        After all, if they WOULD have switched their radar on, they would have found about a dozen “high bird” flights (international airliners) passing overhead during their 2 1/2 hour wait.

        Second, note that the call by “Naimanets” is made 2 minutes before MH17 goes down. The conversation lasts about 30 sec, the flight-time of the missile is about 30 sec, so overall, they would have had only one minute to :
        – Bezler “pass the information up the chain”
        – The crew to get on their feet after 2 hours of waiting
        – Switch on their scanning radar
        – Switch to target radar
        – lock-in their target radar
        – press the button.

        This very short timeframe suggests that Bezler was
        1) Either at the launch site of in very close (radio) contact with the BUK crew, or
        2) Bezler was out of the loop altogether, since the command structure went straight from whoever commanded this BUK crew to press the button was perfectly aware where they are shooting at.

  5. Rob

    There is an article in Der Spiegel,

    which mentions :

    “The BND has intelligence indicating that pro-Russian separatists captured a BUK air defense missile system at a Ukrainian military base and fired a missile on July 17 that exploded in direct proximity to the Malaysian aircraft, which had been carrying 298 people.”

    This appears to imply that the BUK that shot down MH17 came from a Ukrainian airbase. But note that that is not EXPLICITLY stated.

    If the BUK DID come from a Ukrainian airbase (possibly BUK 112 on the picture released by the rebels on June 29) then this information creates more questions than it answers. For example, is the BUK on the ParisMatch picture in Donetsk the same as this ‘captured’ Ukrainian BUK, and if so, what was it doing in Starvy Oskol on June 23 ? And if it is NOT the same BUK as the German intelligence report mentions, then there were TWO BUKs in Donesk area on July 17, which raises a whole set of new questions.

    On the other hand, if the German report did NOT state that the BUK captured from a Ukrainian base was the one that shot down MH17, then the question is why Der Spiegel put the two statements in one sentence.

    Does anyone have a copy of that German intelligence report, or a copy of the slides form the Oct. 8 presentation given to members of the parliamentary control committee ?

  6. Joe Morton

    Higgins can’t have it all his own way — it will be interesting what backflips he and his ‘colleagues’ at Interpreter Mag come up with to spin this since Higgins has insisted that the BUK came from Russia.

    The BND report is awkward because it’s the first time a NATO government has admitted the Ukrainians lied and tried to use photos of their own BUKs to depict the alleged operational separatist BUK which Kiev said couldn’t possibly have been captured from its own stocks. How could the Ukrainians either be so corrupt as to sell the rebels a BUK Eliot or so incompetent as to not sabotage one in retreating? Unless, of course, the rumors among the Novorossiya blogs like Col. Cassad that Khodakovsky and Kurginyan are Ahkmetov plants infiltrated into the NAF are true and the whole rebel BUK handover was an immense set up by the Ukrainian/US (CIA) side to create the ultimate false flag, and Russia failed to prevent it. Think of it as a version of the ‘third theory’ surrounding the East Ghouta chemical attack that Higgins has always insisted could have only been a deliberate, ordered assault by Assad’s troops — what if Prince Bandar simply offered one of Assad’s most trusted chemical weapons officers the mother of all bribes to gas Syrians and create a pretext for US intervention? Could this explain why the NSA/US never released the audio of the Syrian army radio traffic after East Ghouta, because it was all of ‘WTF just happened’ variety? Which is not unlike the supposed damning phone calls leaked by the same SBU that published all the now BND-admitted fake BUK photos…the separatists are saying ‘was it us?’ not ‘it was definitely us’.

    Meanwhile, the Russians can say to the Germans: ok BND, release the satellite photos you’re basing this on, we’re ready — and throw in that AWACS radar track from your NATO plane hundreds of miles to the west in Poland that was probably out of range to detect a Ukrainian SU-25, unlike our radars.

    Which is exactly what they’ve been saying to the US since the July 21 Russian MoD presentation — hence the standoff over MH17 — the US/NATO simply won’t release the raw intel that would prove this beyond all reasonable doubt, and that is likely because they need to protect their Ukrainian clients from the gross negligence charge if it turns out the Ukies were ‘baiting’ a BUK they deliberately allowed to fall into the hands of Ahkmetov’s paid mercs allied to Novorossiya.

    Higgins entire flailing effort has been about obfuscating the lack of NSA/DIA released hard evidence showing the Kupola radar operating from Torez, rather than from Ukrainian BUK sites that Russia says it detected. Now he’s got some splainin’ to do as to why if we can accept the BND’s word for it that they know the rebels or a group of rebels shot down MH17 why we can’t accept their word for it that the BUK came from Russia.

    Maybe Higgins and Interpreter Mag will find an Orwellian solution to salvage their credibility, and simply memory hole Higgins conclusions that the BUK had to come from Russia? If you all were clever my advice would be to simply say the Germans can be trusted when they say MH17 was shot down by separatists, but not on the part that it was a captured Ukie BUK because they’re trying to save face for Putin or spare German industries in Russia. Right Eliot?

    • bellingcatadmin

      I stand by Bellingcat’s work on MH17, and we still believe evidence points towards the Buk missile launcher seen in Donetsk and photographed by Paris Match as coming from Russia.

  7. Joe Morton

    “Now he’s got some splainin’ to do as to why if we can accept the BND’s word for it that they know the rebels or a group of rebels shot down MH17 why we can’t accept their word for it that the BUK DID NOT COME from Russia.” [sic]

    • bellingcatadmin

      If you read our report you’d know we didn’t ID 312 as the Buk seen on July 17th, in fact our side skirt comparisons show that 312 doesn’t match the one seen on July 17th.

    • Rob

      Globalresearch has been pushing a steady stream of mis-information and straw-man arguments since the MH17 was downed, and they present ZERO evidence material.

      I read this globalresearch article, but there is NOTHING in there that contradicts Bellingcat’s findings that the BUK that was photographed in Donetsk on the morning of the 17th originated from a Russian army base in Kursk.


Leave a Reply

  • (will not be published)

You can support the work of Bellingcat by donating through the following link: