Airwars is a collaborative, not-for-profit transparency project run by a team of professional journalists based in Europe and the Middle East. We track and assess claims of civilian non-combatant casualties and ‘friendly fire’ deaths from Coalition and Russian airstrikes. We also monitor and archive official military reports of the war against so-called Islamic State (Daesh) and other parties in Iraq, Syria and Libya so that they can be measured against the public record. Our data is drawn from a number of sources outlined below.
Civilian casualty sources
A key aim of Airwars is to assess all known claims of civilian non-combatants killed or injured in Coalition, Russian and other international airstrikes in both Iraq and Syria. We also track international airstrikes in Libya. Though we often use the term ‘civilian’ as shorthand, it should always be assumed – unless otherwise stated – that we are referring to civilian non-combatants who were taking no active part in hostilities.
With 14,300 airstrikes reported in the first two years of Coalition bombings in Iraq and Syria – many in urban areas – non-combatant casualties remain an inevitability. Where claims by credible reporting organisations are cited here, they are cross-referenced where possible to specific airstrikes reported by the United States or its allies.
When flagging potential civilian deaths or injuries from airstrikes, Airwars draws on a wide range of sources. These can include international and local news agencies and NGOs; and more fragmentary social media sites including local residents’ groups, Facebook pages (for examples martyrs’ pages), YouTube footage of incidents, and tweets relating to specific events. On occasion we also include links to militant and terrorist propaganda sources which are directly pertinent to an event. These are always clearly marked as such.
Many credible claims of non-combatant deaths are reported by casualty monitoring NGOs based in Iraq, Syria and elsewhere. A number are members of EveryCasualty’s Casualty Recorders Network, as are we (also see About Us.)
It is important to note, however, that particularly in the context of Iraq and Syria, casualty monitoring presents significant challenges. We report information on alleged strikes and casualties in good faith, though are often unable to follow up or to further verify such claims.
We publish an provisional assessment of likely civilian casualties from international airstrikes based on our findings.
Because of wide variations in the quality of casualty recording, Airwars employs a grading system for events alleging non-combatant or ‘friendly fire’ deaths from Coalition and Russian airstrikes. This represents our own assessment of allegations, and we urge those using the site to make their own judgement based on available sources. Our own grading system is as follows:
- Confirmed: Where an international belligerent has accepted responsibility for the killing or injuring of non-combatants or allied forces in a particular incident.
- Fair: Where there is a reasonable level of public reporting of an alleged incident from two or more generally credible sources (often coupled with biographical, photographic, and/ or video evidence). Crucially, there are also confirmed Coalition or Russian strikes in the near vicinity for the date in question. We believe these cases in particular require urgent investigation.
- Weak: These are presently single source claims. Nevertheless, these can often feature biographical details of victims along with photographic evidence from a reputable source – and with international strikes also confirmed in the vicinity on that date.
- Contested Events: These occur where there are competing claims for the origins of a violent incident. For example, both Coalition and Russian aircraft might be blamed for an attack in Syria; while in Iraq casualtes might be attributed to ground forces as well as to Coaltion aircraft.
- Discounted: Those cases where our researchers or others can either demonstrate that those killed were combatants; that an incident likely did not result in any civilian casualties; or that other parties (eg the Iraq government or Assad regime) were instead most likely responsible for reported casualties.
Reporting in Iraq
In Iraq as in Syria, areas occupied by so-called Islamic State or Daesh are often inaccessible for researchers and journalists, making independent verification of casualty reports highly problematic. The repeated lethal targeting of casualty monitors by a number of parties to the conflict is also of major concern.
Most reports of non-combatant deaths reportedly caused by the Coalition are found in Arabic-language media and social media, and we employ a researcher in Iraq to gather and analyse such material.
Reports are often fragmentary. Instead of a regular news report, details of a fatality might appear on a Facebook martyrs’ page, or in an Arabic-language tweet. A large number of local media sources have also appeared in Iraq in recent years, of variable quality. Some are run by political parties or militias, while other outlets are maintained on behalf of so-called Islamic State.
Local groups might also monitor violence in particular Iraqi towns and cities – often at great risk to those involved. Mosul Ateka provided vital details on civilian casualties for example, until Daesh destroyed the group and murdered many of its members in 2015.
At Airwars we also draw on reports by other credible monitoring groups, although there are few such agencies operating in Iraq. The United Nations Assistance Mission in Iraq – UNAMI – issues monthly reports on civilian casualties from all parties. Iraq Body Count is another key source for information on non-combatant deaths.
It should be noted that claims of casualties from Coalition airstrikes represent only a very small proportion of overall fatalities attributed to the conflict in Iraq. According to IBC for the period August-November 2014 for example, some 6,800 civilians reportedly died in Iraq’s violence – of which no more than 100 were allegedly caused by Coalition air operations.
Reporting in Syria
Airwars employs two Syria researchers – based in the UK and Lebanon – who collate and assess claims relating to civilian deaths. We also seek clarification from international belligerents; and engage regularly with local and international NGOs, and with news agencies in the region.
For Syria, a number of local NGOs and casualty monitors monitor the chaotic and violent situation on the ground. Many had already emerged during the civil war, building up extensive casualty recording networks in areas now occupied by Daesh or rebel forces.
A key resource for information on reported casualties from airstrikes is the Syrian Network for Human Rights which tracks civilian deaths through a network of local sources.
The Syrian Observatory for Human Rights also tracks casualties of airstrikes and other volent incidents. Additionally other local monitoring groups like the Violations Documentation Center and Raqaa Is Being Slaughtered Silently regularly report on airstrike fatalities, while international NGOs such as Human Rights Watch and Amnesty International have also issued reports on civilian casualties of Coalition, Russian and Assad regime airstrikes.
Airwars airstrike data is primarily sourced from regular military briefings from allies participating in the US-led Coalition in Iraq and Syria, and from the Russian Ministry of Defence. US strikes in Libya are reported by AFRICOM.
These daily, weekly and monthly reports are then cross-referenced against reported civilian fatalities, and are archived to provide a permanent record of military claims.
The US military remains the dominant source for public reports on the Coalition air war. US air strikes are carried out through Central Command, or CENTCOM, whose area of responsibility stretches from North Africa to South Asia.
Between August 8th and December 5th 2014, CENTCOM issued daily (and later three times weekly) reports on the air war. From then onwards, regular reporting responsibility was passed to Combined Joint Task Force – Operation Inherent Resolve (CJTF-OIR or ‘the Coalition’), which remains the key disseminator for most official information on the air war. The Task Force is a US-led joint command which involves elements from all participating militaries.
In its campaign reports, the Coalition presents overarching data for all participating allies, and lists the number of “strikes”, some details of vicinities targeted, and what materiel, buildings or military units have reportedly been struck. No casualty estimates are provided.
The Coalition also provides a weekly breakdowns of the proportion of airstrikes carried out by the US and by its allies, for both Iraq and Syria.
In addition, Air Force Central Command (AFCENT) publishes monthly Airpower Summaries for Iraq and Syria. These feature tables of data including the number of weapons released, ISR and fuelling missions flown etc.
International Coalition allies
The United States has variously been joined by twelve known allies in its air war against Islamic State. Many issue their own reports of military actions, although levels of openness vary significantly.
As the second international partner to join the allied air war, France’s Operation Chammal initially aimed for a fair level of transparency, with the Ministry of Defence reporting airstrikes within 24 hours – stating what aircraft and weapons were employed; and which locations and targets were struck. France later moved to weekly reporting, and somewhat limited the information it released. General target locations are often given, along with a weekly tally of strike numbers.
Britain’s Operation Shader began reporting with good levels of transparency, noting the aircraft and weapons used; the locations struck, and the targets engaged. However the UK has often limited location details for strikes involving its remotely piloted Reaper drones – a transparency concern.
The UK maintains a single evolving webpage for its reports, which may not represent a complete picture of all UK strikes. The Ministry of Defence has also on occasion amended earlier copy, making the process of accurately tracking some reports a challenge.
Belgium was the first nation officially to suspend its participation in the Coalition’s air war, when Operation Desert Falcon ended on July 3rd 2015. The campaign resumed in July 2016, this time with strikes in both Iraq and Syria.
Belgium remains the least transparent nation involved in the campaign. After reporting an initial airstrike on October 5th 2014 and another on November 3rd, Belgium made no public statement on its ongoing airstrikes until April 24th 2015, and only then reporting overall tallies of targets and sorties. In total, we now know Belgian F-16s dropped 324 bombs and missiles during their first nine-month campaign – an estimated 115 airstrikes.
The Danish government was initially heavily criticised for refusing to state where it was bombing in Iraq, and on which dates. Reporting improved somewhat, with weekly summaries detailing the numbers of sorties carried out and weapons released. Denmark still refused to declare where it bombed however, with a military spokesman telling Dagbladet Information that the Danish military would prefer ‘to hide in the crowd.’
Denmark ended its initial role in the Coalition on October 1st 2015, with strikes resuming in both Iraq and Syria in early summer 2016.
The Dutch initially had a reasonable transparency record, publishing weekly reports stating how many missions had been carried out in Iraq, and how many weapons had been released. The Dutch Ministry of Defence declines to say which locations it bombs, however.
From spring 2015 the quality of Dutch official reporting of the war collapsed, with the number of strikes and weapons released each week no longer given. Airstrikes were extended to Syria in late 2015, though the Netherlands does not distinguish in its reports whether it has bombed in Iraq or Syria in any given week.
Reporting by the government of Australia (Operation Okra) is presently weak, with only monthly reports indicating the number of weapons released and sorties carried out. Pressed on why it refuses to say where it is bombing in Iraq, a spokesman has said that the Australian Defence Force ‘will not release information that could be distorted and used against Australia in ISIL propaganda.’
Ottowa’s Operation Impact had the most transparent approach of all participating militaries (operations ended in February 2016), releasing details as they happened of where, when and what was bombed in both Iraq and in Syria. Senior officials also hold regular media briefings.
Canada’s practice of using a ‘rolling summary’ format to describe its strikes can present challenges however, since the MoD deletes detailed comments after a time and replaces them with one-line summaries.
Ankara began airstrikes in Syria and Iraq on July 23rd 2015, though did not officially join the Coalition until August 28th. Despite claims that its targets would be Islamic State forces, most Turkish airstrikes appear instead to have focused on Kurdish irregular forces (which Ankara views as terrorists).
Press releases relating to particular actions are occasionally released via the Ministry of National Defence. Civilian casualties have been alleged from a number of reported Turkish events, and Airwars is presently assessing these claims.
At times assisting the United States in its airstrikes in Syria have been the combat aircraft of Saudi Arabia, Jordan, the United Arab Emirates and Bahrain. However all such operations had reportedly ceased by September 2015, with the GCC states switching their efforts to the war in Yemen. According to an Airwars estimate, 135 anti-Daesh airstrikes have been carried out by the Arab allies.
The Russian Federation entered the conflict in Syria on September 30th 2015, unleashing a wave of airstrikes against multiple targets across the nation. While Russia initially claimed its strikes were aimed at so-called Islamic State, it later conceded that numerous armed groups in Syria were its target. The Ministry of Defence provides occasional updates on the campaign in both Russian and English.
Reported civilian fatalities from alleged Russian strikes have been high in comparison with the US-led Coalition. Airwars runs a separate tracking and assessment project which compiles data on Russian airstrikes and casualties.
Israel has carried out a number of airstrikes on Syrian territory, which appear primarily to have targeted Hezbollah fighters from Lebanon. Civilian casualties have been alleged from a number of these reported Israeli events, and Airwars is presently assessing claims.
The Iranian government backs the Assad regime in Syria, primarily with ground troops. However Tehran has also declared carrying out drone strikes in support of Assad, for example during the 2016 siege of Aleppo.
Reporting issues: Coalition strikes, sorties and locations
As already noted, members of the US-led Coalition use differing methods when reporting on military actions in Iraq and Syria. The primary counting method is the ‘airstrike’ – employed for example by the Coalition as well as by the US, the UK, France and Canada. This reporting language can be problematic, however.
The term ‘airstrike’ is one the US Air Force does not employ internally, given its imprecise nature. The preferred USAF metric is actually ‘Sorties with at least one weapon release’ – see for example AFCENT’s ongoing Airpower Summaries.
As the Coalition itself notes of airstrikes in every daily press release:
‘A strike, as defined in the CJTF releases, means one or more kinetic events that occur in roughly the same geographic location to produce a single, sometimes cumulative effect for that location. So having a single aircraft deliver a single weapon against a lone ISIL vehicle is one strike, but so is multiple aircraft delivering dozens of weapons against a group of buildings and vehicles and weapon systems in a compound, for example, having the cumulative effect of making that facility (or facilities) harder or impossible to use. Accordingly, CJTF-OIR does not report the number or type of aircraft employed in a strike, the number of munitions dropped in each strike, or the number of individual munition impact points against a target.’
A clear example of this imprecise term can be seen on December 5th 2014. CJTF-OIR reports only two ‘strikes’ in the vicinity of Mosul. In contrast, the French Ministry of Defence notes that 15 aircraft from 7 nations struck 20 targets in this particular Mosul raid. Only three nations (France, the United States and Canada) have actually confirmed carrying out attacks on that date.
Variations can also exist between CJTF-OIR and allied military reports, even when relating to strikes at the same location and on the same date. Again on December 5th, the US reports that allied aircraft striking in the Rawah area destroyed ‘an ISIL tank.’ Yet the British – who appear to lay claim to that strike – instead report an armoured personnel carrier being bombed. In such cases we are unable to reconcile contradictory accounts.
Any given locations for airstrikes must also be treated with some caution, with US officials conceding that stated locales should be taken only as guides to where strikes may have taken place.
On October 18th/19th 2014, the British reported carrying out an airstrike in Ramadi, while the French noted in the same period a strike at Tikrit. Neither of these cities was mentioned in CENTCOM’s overall summary of bombings for that day. An official later conceded to the compilers of Airwars that strikes described as ‘south of Bayji’ and ‘south east of Fallujah’ most likely referred to the UK and French operations that day.
On another occasion, officials initially denied that a Coalition strike had been carried out at Al Bab, Syria on December 28th 2014 – a problematic event in which at least 50 non-combatants reportedly died. Only on January 11th 2015 did the Coalition concede to the news agency McClatchy it had carried out the attack, raising fresh questions about the trustworthiness of its locational reporting.
Errors and corrections
We strive for accuracy and transparency of process in our reporting and presentation. Our casualty datasets are continually evolving, representing our best current understanding of any alleged incidents.
If you have new information about a particular event; if you find an error in our work; or if you have concerns about the way we are reporting our data, then do please reach us at firstname.lastname@example.org