15 March 2012

Air base missile attack

(latest updated September 17, 2012, except for the plead)

Please notice: this is a layman's post. If you read it please don't skip the notes at the bottom of the post. Also before you decide what to think of my writings: please read through my latest post (plead added 03/06/2014).

After viewing the video below of a missile attack on a fighter jet on the ground, I wanted to see if the attack could be mapped. Therefore I have looked into what can be found out about the attack, which is supposed to have taken place on March 7, 2012, in Syria.

video originally found uploaded on March 14. to a198123's YouTube channel. 

On his blog aviation journalist David Cenciotti has identified the air base, as the Abu Ad Duhour Airbase, 46 km East South-East of Idlib. He has also identified the weapon, that is being prepared in the beginning of the video. It's a 9K115-2 Metis-M with the NATO reporting name AT-13 Saxhorn-2. (I'm guessing Saxhorn is with reference to the shape of its sighting). The fighter jet being targeted is identified as a MiG-23MS. (minor edit on 18/09/2012)
over view of the air base. Site of attack marked with rectangle (left).
firing-position of the opposition fighters on the hill (left) - and the plane in front of the hanger (right).

From what I can see, the opposition fighters are preparing the attack somewhere among the buildings in the lower left corner of the image. They then run to the hill (marked), from where they fire the missile.

My reasons to think so are these: In the video there is a glimpse of a road down to the right of the hill. That fits. So does the angle towards the hangar. The direction of the fence between the hill and the hangar also looks right (but the distances can be hard to judge because of the zoom). 
An additional thought: As they arrive to the hill, and start to set up the missile, there is some gunfire. I also see at least one man running from the hangar and to the right. This gives me the idea that the gunfire might be some kind of diversion. That is however pure speculation. 

The distance of 850 meters for the shot was a bit of a surprise to me, but that's what I get from using the measuring tool in Wikimapia. 
Whether the jet was actually hit, or the hit was to the hangar behind it remains unclear - to me anyway.

Addition march 15, 2012:
I took a closer look at the route of the men who carried out the attack - and what I've found is indicated in the image below.

aerial of the area where the attack was carried out

The aerial photo is too old for me to be absolutely certain about the route. I have X-marked two buildings, that I cannot exclude completely, as being the half way building. But whether it's the building marked "B", or one of the others, does not change much in the general picture: a group of men meeting up and preparing an anti-tank missile attack in plain sight, and less than 200 meters from the missile launch position.
It all have a very un-military appearance. But if you want to pass for a civilian that would be a good tactic, and the guy testing the tripod seems to know what he's doing. On the other hand, the guy in black seems to have problems carrying the missile without having the strap of the kalashnikov slipping off his shoulder. I imagine something like that would have been solved if they were all trained.

Below: a series of frames with comments.

five frames from the video on a198123's YouTube channel.

update, more resent imagery:
(march 19, 2012, changed march 30, 2013)
The blog Open Source GEOINT has (or had) a post, which corroborates my findings about the route and firing position. The access to Open Source GEOINT has since been restricted, but a key element in the corroboration is about the low wall the fighters are running along from 1:03 - 1:09 min. (see frame #2 above). This wall does not appear in the 2004 imagery, which is the most recent in Google Earth, but Open Source GEOINT has/had acquired imagery from October 2010, in which the wall is clearly visible.
update, about the MiG:
(march 23, 2012)

From David Cenciotti I have learned that there are things that indicate the MiG was already taken out of service, and that it might have been placed as a decoy. Indicators of this would be, that the afterburner-feathers and the nose have been removed. After comparing the very blurred frames, that I have been able to freeze, to images of MiG-23s I tend to agree (that those are missing). But even if the target turns out to have been of low value, it's still my impression that the video as a whole is not staged.
update, more attacks:
(september 17, 2012) 

The Abu Ad Duhour Airbase was reported attacked on the night of August 29th. The reports said that 10 fighter jets were damaged/destroyed on the ground, and that one (MiG-21) was shot down as well. I have mentioned the attack with some links in the post 'Syrian Airforce Unleashed', but a  recent video, documenting damages after the attack, have relevance to this post.
damaged MiG-23. Image source: video on شبكة أخبار حلب و إدلب's YouTube channel. Uploaded 16/09/2012.

Yeps, a clear view of a clearly damaged MiG-23 in the same position, and in front of the same hangar, as the one attacked in the March 2012 video.

In theory it could be a different jet, but I think it's safe to assume it's the same MiG-23, that was attacked with the Metis-M missile. It can be seen, that the nose cone is actually there, but it's also clear, that the afterburner feathers have been removed. I don't have the expertise to say if that makes the decoy thesis confirmed though.

The video also shows a group of planes that seemed to have been moved to one side of the airbase. I assume the idea is to show, that the jets have been damaged in the recent attack, but only the one to the fare right have damages, that are clearly visible in the video. 
five jets on the ground. I believe the four that still have their tail are MiG-23s. The position of the wings on the jet to the right makes me think it's a MiG-23 too, but less certain. Image source: video on شبكة أخبار حلب و إدلب's YouTube channel. Uploaded 16/09/2012.

More material on the latest attacks/siege of the airbase:

video: Idlib - Abozaor | | Syria Martyrs Brigades - revolutionaries around the airport 16-9
video: Idlib - Abozaor | | clashes battalion Ahrarmarh Numan 09/16/2012
video: Idlib - Abozaor | | process of armor battalion Sunni at the airport 15-9

relevant posts:

- Syrian Airforce Unleashed
- Hama Military Airport
- Syrian Airforce use of the L-39 Albatros
- weapons used on Homs 
- mortar position, Homs - (A similar study)
- Syria: Local photo sources

- any comments or corrections are welcome, either as comments here or to my Twitter account.

3 March 2012

drone in Iran

With regards to drones, this is a layman's post. I'm an architect and do know something about perspective drawing and buildings though. Also, before you decide what to think of my writings: please read through my latest post (plead added 03/06/2014).

In December 2011, when a U.S. RQ-170 Sentinel drone turned up in Iran, I followed and took part in the debate and speculations about the incident.

At some point I made an attempt at determining the size of the drone by analysing some of the images available on-line. This study lead to an estimated wing span of the drone as well as a possible size of the building and a sun-angle analysis that could work as a tool to verify suggested positions. 

Back then I shared my results with aviation journalist David Cenciotti who published the material on his blog The Avaitionist – this is the nerdy version:

the mess of black lines is to be imagined as drawn on the brick wall, and the purpose is to transfere known sizes for measuring.

determining sizes:

There are different elements in the photos that can be used to determine sizes. In this case bricks are visible. These can be counted to give the length of the constructive module of the building. Unfortunately bricks have different standards in different countries, and I didn't find any certain information of Iran having just one size.
What made it possible to get a credible handle on the sizes in this case was the markings on the floor. I found it reasonably certain that the circle in the middle is a standard basket-ball marking (diameter 360 cm) and that the white lines are international standard volleyball markings.

By extending the floor-markings through the wall and having them meet in their vanishing point the distances between the markings can be used to measure the wall.
And from the vanishing point new lines "parallel" to the floor-markings can be drawn to points that needs to be measured. The wall can then be used to determine their relative distances, either by counting bricks or by sub-dividing rectangles of known sizes.

markings on the floor extended through the wall to their vanishing point. Rectangles sub-divided by drawing diagonals.

- constructive module and gym-space:
from the analysis I found that the constructive module is 600 cm, which gives the gym-space a length of a little more than 36 meters (6x6m + distance from last module-lines to the end-walls).

The width of the gym-space a lot less certain. If the court is placed symmetrically within the space I estimate it to be about 20 meters.

- bricks:
the brick module that could be calculated from the constructive module was 23 cm (22 cm + 1 cm mortar). This matched one of the standards I had found.

The other brick-size I found was a little shorter, and since that would make the volleyball court smaller than the standard size, in a gym-space with plenty of space, I consider it unlikely. In other words: this length for the constructive module is close to being cross checked.

- the roof:
the roof will be pitched with a central ridge. And an estimate of the exterior length of the roof would be about 39 meters (36m + 0.5m to get to the end-walls + 0.75m for the thickness of the end-walls + 1.5m for roof-overhangs). Theoretically one or both of the end-walls could be partition-walls instead of an "exterior walls", but I consider that very unlikely.

The estimated width of the roof will be about 22 meters (20m + 2m for walls and overhang). This is if the court is placed symmetrically. If not the roof will be wider.

placing the drone:

Since the drone was placed with its axis at an almost right angle to the wall a good estimate could be made already by drawing lines from the wing-tips to the vanishing point of the floor markings. But to try to get some more information I also tried to determine the drones position and orientation in the room. The method I used was to extend the line of the leading edge of the left wing to find where it reached the wall and to estimate the position of the nose of the drone. After finding the point where the axis of the drone reached the wall, the axis could be drawn and used to mirror the line of the left wing.

line of leading edge of left wing extended to the wall.
depth are marked along the line touching the centre circle to the right - sizes transferred by drawing diagonals (imagine lines are drawn on the floor).

The most difficult part of this was to measure the distance from the nose to the wall. The photo best suited for this was a low resolution screen-shot from a YouTube video.

The resulting angle of the wings turned out to match the shape of the artists impression in Wikipedia - which was interesting even if it didn't prove anything. The wing-span however - according to my study - is between 13 and 14 meter, which is significantly less than what it was thought be. The thought-of wingspans that I found was between 18 and 26 meters.

the position and orientation of the drone in the gym-space. White lines are volley-ball markings, dark double lines are the steel-frames/constructive module. The coloured circles refers to same colour vertical lines in images above. Inserted image of drone is an artists impression from Wikipedia.

The width of the gym-space in this drawing is based on the assumption that the court is placed symmetrically in the room. However some things could speak against this: it is thinkable, that the court has been placed closer to one side-wall to give room for spectators along the other. There is also an elevated camera position in some of the recordings that could suggest some kind of bench-arrangement.

sun angles:
(updated march 5.)

During the discussion of a specific suggested location of the drone photo-session the direction of the sunlight was taken as taken as proof (or dis-proof). The arguments were in the "looks like" category, but if the height of the sun (vertical sun-angle) can be measured accurately together with the compass-direction to the sun (horizontal sun-angle) this - together with the date will be a very useful tool for verifying a location.

- explanation:
at any particular position; over the course of a day the sun will be in a certain vertical angle only twice (once before noon and once after). Therefore - if the date is known and a specific location considered - the vertical sun-angle will point to just two moments. And since the horizontal sun-angles at these moments can be found/calculated the angles can be checked against each other. However; the combination of date and sun-angles is not specific to a single position, only to a single latitude (on each hemisphere). Different sun-angle calculators can be found on-line. Here is one.

- in this case:
measuring the horizontal sun-angle relative to the building, and combining it with the vertical sun-angle, will give the two possible compass-orientations of the building in the photos if it is to be in the suggested position. This can then be checked against the aerial photo (if another position is to be checked, the moment specific to the vertical sun-angle in that position will have to be found and the compass-orientation re-calculated).

sunlight on the end wall can be analysed.
this diagram is meant to explain the principle and it is not in scale. I have written the measures that I came up with through my calculations, but they are all to some degree uncertain.

- method:
I found that the sunlight on the end-wall was where the vertical sun-angle and the horizontal sun-angle relative to the building was best measured. The diagram above shows which distances were needed to calculate the sun-angles.

The vertical distances were measured by counting bricks. And since the corner was obscured by the steel-frame the horizontal distance along the end-wall (5.2 m)  was calculated by using a calculated climb-ration of the sunlight. The climb-ratio too was determined by counting bricks.

the AM and PM if-in-this-position orientations of the gym-space.

In this drawing I have shown the two calculated orientations of the building if it is to be in the suggested position. And below the AM-orientation is used for comparison to the suggested location and a more randomly chosen location. The orientation is close.

comparison to possible locations: on the left a position that was suggested by a source to The Aviationist.

summing up:
(updated/changed; march 5.)

These findings have some margin of error (of course).
In my own opinion the measures of the constructive module of the gym and the wing-span of the drone are likely to be quite accurate, while the angle of the wings of the drone is more uncertain because of the poor material for measuring the depth (width).

When it comes to finding ways to make comparisons to buildings in aerial photos things get less clear. What can be established is a likely minimum width (22m), a length (39m) and two within-some-margin-orientations (that will have to be re-calculated for different latitudes).

In the sun-angle studies the main source of possible error is the height of the brick-module, and if I was to give myself an advice it would be to investigate this further. But as it is now: If the brick module is lower than the figure I used the calculated vertical solar-angle should be lower. This will push the AM and PM-moments further apart, making the AM-moment earlier. Given that the horizontal solar-angle relative to the gym-space is less likely to be flawed, this will turn the calculated AM-orientation even closer to the orientation of the suggested building. In my opinion, this places the orientation of the suggested building within the margin of error.

roof outlined and overhang indicated at south-end. The ruler-result is total length of the roof (yellow line).

Therefore - since the width of the roof is uncertain - in relation to the suggested location the length is what can be used.

New aerial imagery has appeared in Google Earth - and the length of the roof I get from there is (still) a figure between 43 and 44 meter. But in my estimated length of the roof of about 39 meter  a roof overhang of just under 1 meter was included. In this "new" image a detail, that looks like damage to the overhang, leads me to believe the overhang could be about 2.5 meters. If this is the case, then it leaves less than 1 meter unaccounted for, which must be considered well within the margin of error.

This means that I have to change my position about this building: it is now a good candidate for the location of the drone photo-session.

the not-low-res photo at the top of the post was released by the web site of the Iranian Revolutionary Guard (if I remember correctly). All other images from the gym-space are screen-shots from different YouTube versions of footage released by Iranian state media on December 8. (2011). The aerial images are screen-shots from Google Map and Google Earth. Drawings and lines on images etc. are by me and are for the main part not made as illustrations - but are from my work-sheets (sorry). In the plan view of the gym-space an artists impression from Wikipedia is inserted.

Someone made a similar size study. It seems we agree.

other studies on this blog:
- mortar position, Homs (Syria)
- air base missille attack (Idlib, Syria)

- all comments and corrections are welcome, either here on the blog or to my twitter account.