Incident ID: DHA10002
Location: Sanaban, Dhamar, Mayfa’at Anss, Yemen
Coordinates: 14.426831, 44.664921
Incident Grading: Confirmed
Date: 2015-10-07 YYYY-MM-DD
Time (AST): Approximately 2200 AST
– We identified the location of the alleged strike as the house and courtyard located at coordinates 14.426831, 44.664921. Using open sources, including images, videos, and satellite imagery, we believe that the strike took place in an area primarily inhabited by civilians, while a wedding party was taking place.
– By comparing satellite imagery, media reports, and user generated content (UGC), We were able to determine that the strike occurred around 2200 AST on 07/10/2015.
– Contextual evidence and witness accounts indicate that this was an airstrike. The Joint Incidents Assessment Team (JIAT) confirmed that an airstrike did take place in this area on this date. However, it denies that the wedding was targeted.
Searches were carried out using English and Arabic variations of the name “Dhamar” and “Sanaban” and words such as “wedding”, “airstrike”, etc in both English and Arabic. These search terms were entered into Google, with specific site searches for sites of relevance, such as YouTube and the Saudi Press Agency. The search terms were also used in Tweetdeck.
Initial Google searches led to this article, which identifies the strike as having occurred on 07/10/2015. From this point, searches focused on the days immediately following this date. Results from these searches can be found organized below.
Sample search string: site:https://www.spa.gov.sa حفل زواج ذمار
The Telegraph 15 fatalities, 25 wounded
The Irish Times 25 fatalities, 50 wounded
The New York Times 23 fatalities
Samaa TV 13 fatalities, 38 wounded
Yaman Yoon 54 fatalities, 31 wounded
Al Hayat 28 fatalities, 20 wounded
CNN Arabic 30 fatalities
Eje Central 26 fatalities, 40 wounded
Al Jazeera 28 fatalities 38 wounded
List of casualties (with names) published by the Security Information Center, General Directorate of Moral Guidance and Public Relations, Ministry of Interior of Yemen (43 fatalities (although only 42 names are listed), 65 wounded)
Human Rights Watch 43 fatalities
Yemen Peace Project 43 fatalities
Social Media (sorted by upload date and time):
07/10/2015 2255 AST: Tweet – First report of attack
08/10/2015 1457 AST: Tweet – Images of aftermath
08/10/2015 1702 AST: Tweet – Video of aftermath (referred to as “aftermath video”)
08/10/2015 1827 AST: Youtube video – Yemen Today report (referred to as “Yemen Today report”)
08/10/2015 2123 AST: Tweet – Video of rescue efforts
09/10/2015 0254 AST: Tweet – Images of aftermath
10/10/2015 1355 AST: Youtube video – News report on incident
12/10/2015 0325 AST: Youtube video – Yemen Today news report from the next morning (referred to as “post-incident report”)
12/10/2015 0344 AST: Youtube video – Yemen Today news report
16/10/2015 1746 AST: Youtube video – What appears to be Yemen Today footage edited to be more dramatic (referred to as “Haitham video”)
11/08/2016 0621 AST: Youtube video – News report on commemoration
Where did this incident take place?
Topography to the southwest of Sanaban matches that captured in video from the incident and led us to examine that Sanaban in greater detail.
We discovered that Google Maps had associated images with this location, although the exact source of these images could not be established. We noted that the horizon did appear to match that seen in the aftermath video.
Both these views also match the Google Earth Pro landscape view of this location.
We believe the coordinates of the building which was destroyed are 14.426831, 44.664921. This location will be referred to as “Building 1” for the rest of this report. Much of the UGC (like the video panorama below) is captured from coordinates 14.427050, 44.664794, which is immediately to the north of Building 1. We built the panorama below by stitching together still images from the aftermath video to give an impression of the scene.
In the aftermath video we can see three primary structures (Building 1 on the left, a minaret, and the building on the right) from the filming location facing southwest toward the mountain. We see that the remains of Building 1 are tall and thin, which matches satellite imagery taken after the strike.
We can also match up distinct features of Building 1 and its surroundings with features seen in satellite imagery.
We can stitch together two images from another social media post to form a panorama which gives us a more detailed view of the area to the south west of Building 1, which we have done in the image below. When we compare this panorama to satellite imagery we can see that the key features match. We know this location also depicts Building 1 as the pair of distinctive half-moon windows on the base of Building 1 can be seen on the far left hand side of the panorama.
We can also clearly see the damage to Building 1 in Google Earth satellite imagery.
What was the location being used for?
On Wikimapia, a crowd-sourced mapping service, many of the tagged sites in the village al-Sanaban are local homes, a primary school, and a cemetery. No locations are marked as military on Wikimapia. The majority of these locations were added 5-11 years ago.
Similarly, on Google Maps, nearby locations include several mosques, a garden, a shopping mall/market, and a poultry farm. It is not currently possible to accurately identify when these locations were added.
Satellite imagery from the time does not appear to show any fortifications or obvious military objects. It should be noted that it is not possible to confirm the absence of Houthi soldiers in this area, only that open sources indicates the area is inhabited by civilians and was likely being used as both a residential and commercial area.
Was there military activity in this location?
Analysis of the sources available to us do not indicate that there was any military activity in the immediate area of Building 1. According to several sources, Muhammed al-Sanabani was “a tribal leader known to support the Houthi rebels.” However, footage of the aftermath of the strike and from inside the hospital, as well as witness reports, indicate that most of the victims of the strike were civilians, including women and children.
According to the Yemen Peace Project, “only one military post could be seen 15 kilometers south of Sanaban, on the way to Dhamar City” and “the presence of the Houthis could be hardly seen in the village.” According to Human Rights Watch, “Al-Sanabani and other residents said there were no military targets in the vicinity, no Houthi forces, not even a checkpoint.”
In images posted to social media depicting the aftermath of the incident there do appear to be some people carrying weapons. It should be noted that these images appear to have been taken the next day, hours after this incident, and that carrying a weapon in Yemeni culture can be regarded as a mark of rank, and not necessarily a marker of a combatant.
There also appears to be at least one person dressed in military fatigues, although again it should be noted this was taken some time after the incident.
No armed persons or people seen in military uniform could be seen in the videos taken at night immediately after the attack.
All media reports confirm that the strike occurred on 2015/10/07. Sentinel Hub satellite imagery appears to support this date, as we can see a structural change at the location of Building 1, marked in the white square, between 04/10/2015 and 21/10/2015. Although the imagery is indistinct, we can see that there has been enough change for this relatively low resolution imagery to demonstrate something significant has happened. The structure appears slightly smaller and has a darker area surrounding it, which may be caused by the debris field resulting from this incident.
Notably, these videos appear to have been taken immediately following the strike, as we can see people digging through rubble attempting to rescue people who are still alive and evacuating people to the hospital. Most of the UGC content captured during the daytime is people walking through the rubble with much less urgency, indicating it was taken some hours after the strike occurred and after rescue efforts had been abandoned.
Furthermore, one of the earliest reports of the attack found through a methodological Tweetdeck search was posted at 2255 AST. This supports the argument that the strike occurred at around 2200 AST.
Report continues on next page