Tor, Iran, and Ukraine International Airlines Flight 752

ckfinite
ckfinite
Jan 11 · 16 min read
A Russian Tor-M1E operator.

We don’t know exactly what happened near Tehran, Iran, during the early morning of January 8th, 2020. However, based on the available open-source evidence, we can construct a plausible story as to what happened that morning.

16 miles away from Imam Khomenini International Airport, two Islamic Revolutionary Guards Aerospace Forces (IRGC ASF) crew were awaiting any sign of attack. Just hours prior, Iran had launched a barrage of 15 ballistic missiles at US-Iraqi bases, and the air defenders were most likely on high alert, waiting for a potential US response. They were what separated the pride of the nation from potential destruction, after all.

To perform their critical task, the crew had been entrusted with one of the most advanced surface to air missile (SAM) systems operated by Iran: the Tor-M1 complex. Originally designed by the Soviet Union to protect against precision guided munitions, the eight vertically-launched missiles and two radars integrated into the vehicle could put a serious dent in any air attack — if the crew was attentive enough to put them to use.

Reaction time is an essential part of any air defense engagement. The IRGC ASF soldiers would know that the US operates the most sophisticated penetrating strike capability in the world, with a focus on both very low observable aircraft (commonly called stealth) and cruise missiles. A single B-2 bomber, for example, could drop 18x 2,000lb bombs onto its target, far more than a single Tor could hope to defeat with 8 missiles. The crew would have known that, if they detected a threat, they would have to engage the threat as soon as possible before it could release its ordinance.

Despite the political turbulence that night, for most of the evening nothing untoward had happened. After likely going on watch some time late the previous day, the crew would have been looking at the strikes on the US-Iraqi facilities with baited breath. Soon, they might be the ones in the firing line. 14 minutes and 24 seconds after 6 in the morning, they picked up a contact. We will refer to this time of first detection as T+0 in this scenario.

Their Tor missile system had two ways of acquiring targets to engage. The primary means is by a rotating acquisition radar located on the rear of the turret, equipped with a rudimentary identify friend-foe (IFF) receiver. Secondarily, the system can be given targets by an external Integrated Air Defense System (IADS) connection. On this occasion, their radar set would have picked up an aircraft suddenly climbing over a hill, overlooking their position.

The crew would now have seconds to react. The target could release its bombs —potentially far too many for their Tor system alone to engage on its own — at any second. There would be no time to contact the IADS to verify its identity; they might be dead before the IADS could respond. This kind of last-ditch short-range engagement is what Tor was made to do, and, in this scenario, they would do it.

12 to 15 seconds after its first detection, the crew decided to take the shot. 7 to 10 seconds after that, a missile was fired, elapsing approximately 12 seconds to reach its target. The crew waited anxiously to see if they had saved the day: would the target evade? Would it drop its payload, turning their victory into a pyrrhic one? Would the target be able to jam the missile? The crew had no way of knowing, and time seemed to slow down as their missile closed on the target.

The missile scored a direct hit. However, to their collective horror, the target kept coming — it would need a second missile. Firing a second time, the missle streaked towards the target, also impacting. This time, though, the aircraft turned away, heading back towards the east and beginning a uncontrollable descent. Their radar would track it as it fell faster and faster, continuing to disintegrate, until it fell below their radar horizon and crashed in a great fireball. The crew (and their Tor) had saved the day: no American bombs would attack their missile development complex. But had they?


This story is supposition. We don’t know what actually happened that day, and while Iran has admitted that the aircraft was accidentally shot down, we don’t know anything like this much detail. For example, there’s no hard evidence that there was a SAM system defending the missile development facility at all, and we cannot know what would have been going through the heads of the operators of such a system. However, we know a lot more about one particular airliner’s journey that day.

At 02:42 UTC, 6:12 local and just 2 minutes prior, Ukraine International Airlines Flight 752 (PS752) lifted off from runway 29R at Imam Khomenini International Airport for an expected 3 and three quarters hour long flight to Kiev. Aboard were 176 passengers and and 9 crew.

A UIA Boeing 737–800, similar to the one that crashed. Image courtesy of Anna Zvereva.

Six minutes later, PS752 impacted the ground in a vast fireball. What happened? While no official report has came out at the time of writing, we can draw inferences from a wide range of openly-available sources.

This article draws heavily on other sources, most notably Bellingcat and FlightRadar24. Their work will be cited multiple times in this article, and was essential to its compilation. The first two sections of this article simply summarize the known facts of the situation, while the remainder discuss a potential scenario that describes the loss of the aircraft.

The path of PS752 can be reconstructed from several sources. The initial climb was captured by ADS-B transponder data sent by the aircraft to inform other aircraft and air traffic control of its position and captured for posterity by FlightRadar24. In it, the aircraft taxies to position, takes off, and begins its climb out on its scheduled departure.

Just two minutes after taking off, the aircraft experiences a malfunction of some kind, disabling its transponder. It would never send an ADS-B message again.

On the ground around this time, a loud noise was heard in the sky. Already alert for a US retaliation, [Mr.] Nariman Ghraeb pointed their camera skyward, catching what was most likely PS752 and a bright object hitting it just seconds later.

PS752 getting hit. Courtesy of Bellingcat and Nariman Ghraeb.

Bellingcat has identified the location of this video and estimated the position of the impact, based on the time delay of arrival of the noise from the explosion. This gives us one last position for the aircraft, just seconds after the last ADS-B position. At this time, the aircraft is apparently still flying.

Less than two minutes later, the aircraft is seen on fire and impacting the ground, in multiple videos.

Just 6 minutes after taking off, PS752 and its 185 occupants met their end, impacting the ground just 7 miles from where it took off. There were no survivors.

Many questions arise from this timeline. What spawned the noise that Mr. Ghraeb heard, and what was the flash that he recorded from the aircraft? Most accidents have a long and complex string of events that lead up to them, but this one seems particularly abrupt: in the middle of its climb out of Tehran airport, at least one object (and more likely two) appears hit PS752 and exploded.


PS752 climbed into turbulent skies. Less than an hour previously, Iran had launched ballistic missiles at US-Iraqi airfields, and the countries’ military was watching for any potential escalations. As part of this effort, the Islamic Revolutionary Guards Aerospace Forces (IRGC ASF) would have been ready to defend against air attack. Against this background — and the excellent record of international aviation — it seems unlikely that the crash of PS752 was unconnected to the tensions.

These suspicions were soon given new substantiation. While no precise location is known (due to the steep camera angle), several new images of an apparent missile component were soon posted online.

A segment of missile. Via Bellingcat.

If we assume that this was taken in Tehran, near the time and place of the incident, then this is highly suggestive. Depicted here is the nose steering section and fuze of a 9M330 family surface to air missile, designed and built in Russia and part of the Tor complex.

A Tor missile system. Note that this is baseline Tor, not Tor-M1.

The Tor air defense complex was designed in the 1980s for the Soviet Army, intended to counter precision guided missiles with its 8 interceptors. The system is a fully integrated package: the rotating acquisition radar (seen on the back of the turret) identifies targets for engagement, while the fixed engagement radar (on the front) tracks targets and missiles, allowing commands to be sent to guide the missiles to the target. This makes Tor a so-called “command guided” surface to air missile (SAM) system.

As a country constantly under the threat of cruise missile attack, Iran acquired Tor missile systems for the IRGC ASF, and deploys them to guard key strategic sites. On this day of national threat, the Tor system was likely employed to defend important military locations, including those associated with missile development.

Tor is not the only missile suitable for the task, however. While no pictures have been found to indicate its presence, numerous other systems in the IRGC ASF exist that could also be used to defend a facility such as this one. For example the Croatale and its Iranian copy the Ya Zahra are very suited to short ranged air defense of this nature.

Ya Zahra air defense missile system. By M-ATF, via

The apparent missile, seen in the video, traveled left to right across the view, suggesting that the missile came from the northwest.

Just to the northwest of the impact location is a suspiciously regular site in the middle of the desert. This facility, whose name is currently unknown, is widely suspected to be home to part of Iran’s ballistic missile program, and could have played a role in developing and manufacturing the very ballistic missiles fired at the joint bases in Iraq. It would have been high on the list of sites for the IRGC ASF to protect in this time of crisis.


If we posit, then, that there was an air defense system sitting somewhere on this missile development base, what would they have seen? Using the ADS-B data, we can simulate the path of the aircraft from the perspective of a notional missile complex defending the facility.

For the sake of this analysis, we have selected a revetment to simulate the missile battery’s position; however, the facility is small enough that the perspective with respect to the trajectory of PS752 is relatively constant. From this position, looking towards Imam Khomenini International Airport, the IRGC ASF Tor system would have had a view like the following, where the trajectory (reported by ADS-B) of PS752 is shown in green.

From this angle, we can start to see their perspective. They likely had recently been moved to this location to defend it from an urgent American threat; a threat that would tend to mask behind terrain using stealth aircraft and suddenly pop over it. In this scenario, they would be presented with a target that must be engaged as soon as possible. From their angle, they could not see PS752’s takeoff — all they could see was an aircraft that could be a threat to them and the facility they were protecting.

Now, let us consider the timeline. Comparing our reference position to the ADS-B track, we can determine that the PS752 became visible to the Tor battery between 2:44:13 UTC and 2:44:24 UTC, with sightline vectors shown in blue and green on the below map.

The blue line, for 2:44:13 UTC, intersects with a hill. By 2:44:24 UTC, the Tor can clearly see PS752 along the green line, suggesting that this was probably the first time the operators saw the aircraft, coming towards them at hundreds of kilometers per hour at a range of less than 13km. We will refer to times with respect to this 2:44:24 time of likely detection; in this system, the last ADS-B message received by FlightRadar24 was at T+34s, when the aircraft was 8.46km away from our notional launch site.

If we suppose that the ADS-B transponder was disabled immediately after this message was sent, what would that engagement look like?

To determine how long the missile would take to arrive, we can refer to this helpful chart, from the technical information on the system:

This graph depicts the missile’s velocity versus time. To determine position versus time, we can simply graphically integrate the area under the curve. PS754 was not maneuvering, so we take the upper line. Each square in the chart represents a distance of 400 meters, depicted as an area of 4284px. Thus, to cover the 8.42km to the airliner (reduced as the airliner is approaching the missile site), we subtend this area:

Thus, it would take around 12 seconds for the first missile to reach the target from the notional launch site. This means that the missile would have had to have been fired at T+22 seconds; just below the green line from the crew’s point of view and when the target was around 10km away.

It takes a Tor system some time to respond once the decision has been made to engage a threat; according to the aforementioned specifications, this is at minimum around 7 seconds. However, this time was derived using the very best operators, against a target whose time of arrival they likely knew, working with perfectly functional equipment. In real operation under fighting conditions where none of this is true, this time is most likely closer to 10 or as much as 15 seconds. Using 10 seconds as our baseline, the decision to engage would have had to have been made at T+12 seconds after first detection.

We now can see into the minds of the Tor’s crew, at least to some extent. They saw an unknown — or at least unidentified — aircraft appear over a ridgeline without warning, coming straight towards their position. If it was an American bomber, they would have only seconds to react: they would have to get their missiles off before it would have a chance to release its bombs. In this scenario, it appears that they took around 10 seconds to make this choice; 10 seconds later, the first missile was away, and 12 seconds after that, PS752 had a large number of holes in its bottom.

PS752 getting hit. Courtesy of Bellingcat and Nariman Ghraeb.

That explains what the first noise that Mr. Ghraeb heard was. However, what was the missile that he captured? His video is clearly several seconds after the first intercept — if we assume that the missile was what stopped ADS-B transmissions — so what did he see? His video depicts PS752 getting hit approximately 3.6km away from the last transponder position. This distance would have taken 25 seconds for the aircraft to travel, assuming it did not slow down. What happened?

The second missile would have taken about 8 seconds to arrive from the notional launch site, suggesting that it was fired around 17 seconds after the first missile impacted. To explain this, we need to describe typical SAM engagement doctrine.

The principle objective of an air defense system is not to destroy its target. Rather, its purpose is to defend the air. As a result, their goal is to divert an incoming threat, be it by destroying them or forcing them to turn away. They will continue engaging a target until it has either been destroyed or is no longer a threat. To do so, they may need to fire more than one missile.

A single surface to air missile can have a relatively low probability of actually destroying its target. For example, against an F-15, the stated probability of kill (pK) is around 45% with a single Tor missile under optimal conditions (source: the above manual). This is unacceptably low under many scenarios, so a second missile is needed, raising the pK to an aggregate 69%. However, the timing between the two missiles can vary.

We will contrast two engagement doctrines: shoot-shoot-look and shoot-look-shoot. Under shoot-shoot-look doctrine, the system will fire two missiles (shoot-shoot) at a predetermined interval and guide both to the same target. Finally, the crew will observe the target (look) to determine if it has been destroyed or diverted. Under shoot-look-shoot, the operators will fire one missile, then determine if the target has been diverted or destroyed, and finally re-engage if the target has not been neutralized.

Two-missile impact geometry. Purple is first missile, turqoise is second.

The Iranian crew was likely using a shoot-look-shoot doctrine. They waited for the first missile to impact and then observed the target post impact. After being hit (end of the green line, missile along magenta line), PS752 continued on its course towards the missile development facility as indicated by its position in Mr. Ghraeb’s video: thus it was still potentially a threat and should be engaged a second time. After observing for 15 to 17 seconds (as the aircraft flies along the red line in the graphic), the crew decided to fire again. This second shot would take 7 to 8 seconds to arrive (dark turquoise line), and is shown in the video.

Likely, the only reason why the aircraft was not engaged again is because it begun its turn to the right. While we do not have a detailed track of the aircraft after the ADS-B data stops, it ended up substantially to the northeast of the intercept position. Thus, the aircraft no longer presented a threat and the engagement would have stopped.


A segment of missile. Via Bellingcat.

A good question is why did the missile’s nose and steering section survive? The Tor missile warhead is located behind the steering section and the missile is proximity fuzed. When it detonates, it will destroy the body of the missile and send fragmentation into its surroundings in a disc-shaped pattern, but will not necessarily destroy the nose section.

A diagram of the Tor missile

Earlier, the Crotale system (or Iranian copies thereof) was mentioned as a possible alternative to Tor. However, Crotale is substantially slower than Tor, especially at range. Using the video of the likely second missile hit, we can discriminate between the two systems. In particular, Crotale uses a single-pulse rocket motor, and would not be producing exhaust (and thus light) while flying post-burnout. In contrast, Tor uses a sustainer motor pulse, producing a small thrust for 8 seconds after the main pulse burns out, causing it to give out light.

Velocity vs. time charts for Tor and Crotale. Tor charts from the above document, Crotale from AirPowerAustralia.

By visually integrating the area under the velocity vs. time curves for both systems, we can determine that at the estimated range of the intercept (5km-5.6km) Tor would be on its sustainer pulse and Crotale would have burnt out entirely. Consequently, Crotale would not produce the light we see from the missile in the video. Additionally, by stabilizing the video and measuring the angular rate of the missile as it tracks across the camera’s field of view for the 9 frames immediately prior to impact we can estimate a velocity of between 700m/s and 1,400m/s (with error being caused by issues in the stabilization). While this error is substantial, Crotale would be travelling far slower than even the slowest possible speed implied by the video.

Other flights since local midnight as seen from the position of the SAM system. From FlightRadar24.

Finally, why was only PS752 the only flight engaged? As seen from FlightRadar24 data, at least two other flights flew along trajectories close to its path, most notably Azerbaijan Airlines flight J29006 (orange), 4 hours prior, and Austrian Airlines flight OS872 (purple) 2 hours before. It cannot be because the aircraft flew closer to the military installation as Iran claims, as multiple aircraft had flown closer prior to the incident flight. While the fundamental question of why this aircraft is unanswerable with what we know, a number of hypothesis could apply:

  • The crew could have switched at 6:00, with the engagement happening just 15 minutes later.
  • The IADS connection could have gone down, leaving them without civilian flight information (and worrying them that their command was under attack).
  • The crew could have become fixated on this particular target, similar to what happened in IA655.

While we can’t know what happened for sure inside of the control cabin, something happened that caused the engagement of PS752.


Based on this reasoning, we can posit a timeline of events:

  • 2:42:34 UTC (T-110s): PS752 lifted off from Runway 29R, bound for Kiev.
  • T-0s: The aircraft is first detected by a Tor complex defending an Iranian missile development site.
  • T+12s: The crew of the Tor decide to engage the target and commence firing procedure. PS752 continues its departure.
  • T+22s: The first missile is launched.
  • T+34s: The first missile proximity fuzes, sending shrapnel into PS752 and alerting Mr. Ghraeb.
  • T+51s: The crew determines that the threat is still oncoming and fires again.
  • T+58s: Mr. Ghraeb starts filming.
  • T+61s: The second missile proximity fuzes near the aircraft.
  • T+70–80s: PS752 begins a turn to the right and begins descending uncontrollably. It later catches fire.
  • ~T+120s: The aircraft hits the ground and explodes.

Based on the above-mentioned assumptions and information available to us, the probability that a Tor air defense complex shot down PS752 appears to be high. The Tor system is consistent with all known direct evidence, a potential location is known and presents a realistic timeline, and has been implicated by several nations in public statements. Moreover, the situation appears to provide a cogent explanation as to why the operators would decide to fire a missile.

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade