Impact Site Analysis

If this is your first time here, I recommend starting from the conclusion page.

This post estimates the direction of the rocket trajectories, by analyzing the impact sites in Zamalka. 

Impact Site 1

This impact site has a rocket dug into the ground which was investigated by the UN, and which they believe was undisturbed. The UN visit to the site is well documented (Video 1, Video 2, Video 3), and seen in this photo:

This location was accurately identified here. 

The video below shows the rocket from several directions, and it is clear its angle with the wall is 60-70 degrees:

A few screenshots:

Drawing a 60-70 degree angle to the wall on a satellite image gives us a northern trajectory (The red trajectory is what the UN mistakenly reported).

Impact Site 2

This site was matched to a location 100 meters east of site 1. This photo of the site (and more here) clearly show the crater is oriented at a right angle to the plough lines:

Placing this on the satellite image gives us another northern trajectory:

Impact Site 3

This site is a direct hit to a building, examined by the UN investigators in the following video:

At minute 1:40 we see the impact hole at the northern wall and no signs of an angled impact. At minute 1:28 a shadow line is seen, indicating sunlight is coming from behind and slightly to the left. In this video the team is seen exiting the building from the other side. At minute 2:11 shadows are seen on a balcony in the background:

These confirm the angle of the sun as coming from behind and slightly to the left of the impacted wall (which is on the opposite side).

Based on the UN report (Page 26) this visit occurred at August 29th 13:35. Since the sun at that time was at South-South-West, this means the balcony is facing north, and providing us a third indication of a northern trajectory.

Conclusion: The rockets were launched from a location north to Zamalka.

Victim Count

This post will try to estimate the number of victims in the attack.

At this point it is used to coordinate evidence collection, and does not contain any analysis. Please post your evidence in the comments.

Suggesting Scenarios

This page was used to discuss possible scenarios and their fit to the evidence. A conclusion has since been reached and is available here.

===

After weeks of meticulous collection and evaluation of evidence, it is time to reach a conclusion as to what exactly happened on August 21^st in Zamalka.

This will be an ongoing process in which suggestions will be brought forward and examined in an open discussion. Suggested scenarios will be examined on two criteria:

• Fit to Evidence – The scenario should explain ALL the evidence well. In other words, a good scenario is one that is not contradicted by evidence.
• Scenario Plausibility – The scenario should be plausible by itself. This means that regardless of the evidence, it should describe events that are likely to happen in a situation such as the Syrian civil war.

To take two extreme examples:

1. Claiming that aliens launched the chemical attack and then planted all the evidence to confuse us, provides a perfect explanation for all the evidence but is not in itself plausible, and therefore an unlikely scenario.
2. Claiming that no chemical attack occurred on that day is a very plausible scenario (that’s what usually happens), but does not explain the evidence well, and is therefore also an unlikely scenario.

The Evidence

The following lists all the evidence that so far withstood scrutiny, and should therefore be addressed by proposed scenarios:

1. Hundreds of videos, social media reports and eyewitness testimonies describe a mass poisoning event in Zamalka starting at August 21^st 2:00 AM, with symptoms that are highly consistent with nerve agent poisoning: miosis, headaches, nausea, dizziness, foaming, convulsions, loss of consciousness, apnea and death.
2. Many eyewitnesses describe multiple rockets or shells landing in Zamalka prior to the poisoning event. A few hours after the attack videos and images were uploaded showing several rocket remains in their impact sites.
3. The rockets (codename “UMLACA”) show signs indicative they were used to deliver chemicals: small impact craters with little damage, relatively intact rocket bodies, and remains of a container that seem to explode on impact (e.g. Page 22 of the UN report describes container remains found on a roof while the rocket body penetrated to the floor below).
4. Rockets of a similar design but with a conventional warhead were launched by the Syrian Army on multiple occasions.
5. Rockets that seem identical in design to the ones used on August 21^st  landed in rebel-held territory in January, and showed signs indicative of a White Phosphorus payload. None of the UMLACA reports prior to August 21^st are consistent with a sarin payload.
6. The rockets found on August 21^st were launched from an area that is about 2.5 km north of Zamalka.
   A scenario that assumes a different launch area must show how all UMLACA range calculations are incorrect and provide an alternative range calculation that exceeds 3.5 km. Alternatively it would have to show why the three impact sites analyzed here do not indicate a northern source.
7. This area is not under regime control, with parts of it rebel-controlled and other parts contested 
8. The UN investigation team has found multiple indications that sarin was used in the attack.
9. The sarin used in the attack was of low quality.
10. Zamalka is the first alleged chemical attack targeting a residential rebel-held territory. In all previous cases victims were males of fighting age. It is also the first attack in rebel-held territory to result in a large number of deaths.
11. On March 19^th in Khan Al Assal, Syrian troops and regime-supporting civilians were poisoned, possibly by sarin or chlorine, killing 26 and injuring 86, making it the deadliest chemical attack prior to August 21^st.
12. On September 16^th three videos were published anonymously which claim to show Liwa Al-Islam launching UMLACAs at regime forces in Qabun and Jobar, on the date of the chemical attack, while wearing gas masks.
13. These videos contain information indicating that its claimedlocation is very similar to the real launch site.
14. A scenario should choose whether to treat these videos as real (implying certain discrepancies), or as a fabrication (implying another set of discrepancies), or possibly provide some other explanation. See full discussion here.
15. A scenario should choose whether or not to claim a chemical attack in Moadamiyah, again with each explanation implying a different set of discrepancies. See full discussion here.

Please comment if you think I missed any verified evidence of importance.

I will start off the discussion with a few suggested scenarios. After each scenario I will examine its fit to the evidence and its plausibility. 

Within each scenario, each piece of evidence receives one of five scores:

• Match – The evidence is what one would expect to see under this scenario.
• Medium Match – The evidence is not what one would expect to see under this scenario, but still reasonable.
• Weak Match – The evidence is unlikely under this scenario, yet not impossible.
• No Match – The evidence directly contradicts the scenario. For the scenario to be considered, this evidence must be disproved.
• n/a – Not applicable in this scenario.

Scenario 1 – Regime Attack / The Mainstream Scenario

This is the scenario dominating most media reports, and promoted by western intelligence agencies. 

According to this scenario, the regime has grown so frustrated with its inability to rid Damascus suburbs of rebels, that it decided to use chemical weapons to achieve a breakthrough. The regime was confident that there will be no international implications, since its previous smaller-scale chemical attacks had no repercussions.

The attack was launched from Syrian bases on Mt. Qasiun, using rockets fitted with sarin warheads – a chemical agent that the regime has been known to develop in large quantities for over 20 years under its advanced chemical warfare program. 

The area of Zamalka was attacked by UMLACA rockets, while Moadamiyah was attacked with M14 rockets.

Later, in an attempt to relieve international pressure, the regime fabricated videos implicating Liwa Al Islam in the attack.

Fit to evidence:

1. Match
2. Match
3. Match
4. Match
5. Match
6. No Match. The source of the attack is nowhere near Mt. Qasiun.
7. n/a
8. Match
9. Weak Match. The Syrian chemical program is a strategic asset intended to counter Israel’s nuclear weapons, and is considered very advanced. It is unlikely that after 20 years of development they cannot manufacture high quality sarin.
10. Weak Match. The scenario claims the regime expected no response to the August 21^st attack, based on the responses to previous attacks. This is unlikely, given the dramatic differences between this attack and previous attacks.
11. Weak Match. Victims in Khan Al Assal are Syrian troops and regime supporters.
    Note: The mainstream scenario does not claim an accident or false flag in Khan Al Assal.
12. Medium Match. The Liwa Al-Islam videos are a weak fabrication job, and were published when international pressure was already relieved.
13. Match.
14. Medium Match. The Liwa Al Islam videos have several discrepancies when considered a fabrication.
15. Medium Match. The Moadamiyah scene has numerousinconsistencies which make it unlikely to be a chemical attack site.

Scenario Plausibility:

The first thing nearly everyone noted when this scenario was proposed was how suicidal it seems. This is best understood when examined on a timeline:

1. In August 2012 movements of chemical weapons were detected by the west, triggering Obama's 'red line' speech, which clearly conditions foreign intervention on the deployment of chemical weapons. (a reasonable explanation for these movements was protection from opposition advances, and indeed later reports indicated that all weapons are currently in regime controlled areas).
2. In December 2012 Obama goes further and issues a direct warning stating: "if you make the tragic mistake of using these weapons, there will be consequences".
3. During this time, the regime constantly assures Russia that chemical weapons will never be used.
4. Following the Khan Al-Assal attack in March 2013 which left 16 Syrian soldiers dead, the regime pressed for a UN investigation.
5. When the UN investigation got delayed over disagreements on its mandate, the regime invited the Russian investigation team (who blamed the rebels).
6. In August 2013 the UN team finally arrived in Syria, and upon its arrival the regime decides to launch an unprecedented mass-scale sarin attack on a residential area.
7. The regime allows the UN to visit the attacked sites within two days of being requested to do so.

It is hard to read this as anything but an attempt by the regime to bring international military intervention to Syria. This seems even weirder when considering these points:

1. Even if for some reason the regime has decided to kill hundreds of innocent civilians, it could have easily done so with conventional weapons.
2. No explanation is given as to why, after months of fighting and the regime making steady gains, did its “frustration” suddenly become so unbearable that it had to launch a chemical attack.

Summary: This scenario is highly implausible, and is in strong contradiction to the evidence.

Many analysts felt uncomfortable with the mainstream scenario, and proposed some variations on it. some of these somewhat increase its plausibility and explain some of the evidence, but are still far from likely. A few examples:

1. The attack was ordered by a rogue officer. This may explain the poor strategic choice, but imposes other difficulties such as gaining access to sarin, reaching the launch location in rebel-held territory and not explaining the Khan Al Assal attack. Furthermore, rogue operations on this scale are very rare to begin with.
2. It was an intentional regime attack, but was not meant to be so deadly, using the wrong concentrations of sarin. As discussed here sarin is intended to kill. You use it when you want an attack to be as lethal as possible. If the intention is instead to terrorize there are much better weapons, such as napalm.
3. It was an intentional regime attack on a military target, but missed its target. Looking at the map of the attack shows this is implausible. If the target was not the residential areas but the front line, it would mean the regime sent the attackers to infiltrate rebel territory only to attack the front lines from behind – an act that makes no military sense.

So the mainstream story and its variations don’t seem to make any sense, but is there another regime attack scenario that was not yet suggested and is more reasonable? Here’s the best I can think of, but if anyone can come up with a better one, please share.

Scenario 2 – Regime Attack / Double False-Flag Scenario

In this scenario, the regime decided to launch a chemical attack on rebel residential areas, but to avoid international response tried to make it look like a rebel false-flag attack (hence a double false-flag).

They decided to use the UMLACA, a proprietary design which is not documented and will be easier to attribute to the opposition, and filled it with sarin that was intentionally manufactured in a sloppy manner.

On the night of the attack an UMLACA team and launch vehicle infiltrated rebel-held territory to reach the launch site, launched over 10 sarin rockets, and returned safely to base.

After the plan failed miserably and the regime was blamed, they tried to divert the pressure by fabricating the Liwa Al Islam videos.

Moadamiyah – In this scenario the M14 could not be the chemical weapon, since it is easily associated with the Syrian arsenal, and in general attacking two locations far apart undermines the credibility of the operation. So in this scenario Moadamiyah was attacked only by conventional weapons, and the sarin findings there are a result of contamination and patient evacuation from Zamalka.

Khan Al Assal may have been a false-flag to gain international support or an accident during a previous double false-flag attempt.

Fit to evidence:

This scenario fits all evidence well.

Scenario Plausibility:

1. Probably the first double false-flag operation in history (anyone knows otherwise?)
2. Seems like a very big risk to take. The western narrative was pro-opposition, and it seems over-optimistic to assume the west will believe this was a false-flag attack, rather than a regime attack.
3. If the regime intended to convince the world this is a rebel false-flag, they should have prepared an extensive Psychological Warfare campaign, including high quality fabricated evidence. In practice, the evidence presented by the regime was so weak that it is very unlikely to be a fabrication. This includes two ambiguous phone calls, discovery of a rebel cache with some multi-purpose chemicals and gas masks, and undocumented claims of soldiers “suffocating” when entering Jobar.
4. As long as three weeks after the attack, Assad still did not want to commit to a specific theory, stating “We’re not sure that anything happened”. 
5. It was only four weeks after the attack, when international pressure has already subsided, that significant evidence was published, namely the Liwa Al Islam videos. However, these too would make for a very poor fabrication job that was unusable for propaganda.

So at this point it doesn’t seem like there’s any plausible regime-attack scenario that fits the evidence.

But what about a rebel-attack scenario? Let’s examine one option.

Scenario 3 – Rebel Attack

According to this scenario, extreme fundamentalist factions in the Syrian opposition have been building chemical capabilities for some time (possibly related to groups carrying chemical attacks in Iraq). The motivation may have been to counter the Syrian Army’s military advantage, or for carrying out a false-flag operation that will meet the US’s red line requirement.

While they were relatively successful in producing sarin, they still did not have the technology to build an efficient delivery device.

In one of the many raids on Syrian Army bases, they captured one of the regime’s UMLACA launcher and rockets, including some with incendiary warheads. They realized these can be easily refilled with sarin, creating a very efficient chemical rocket.

Prior to August 21^st, they made several chemical attacks on Syrian troops, including in Darayya, Khan Al Assal and Barzeh, and possibly many more which the Syrian Army chose not to report. It is also likely that many of these attacks did not cause casualties due to high preparedness among Syrian troops. Since we don’t have direct evidence on when the opposition gained access to the UMLACA, it is hard to estimate whether these attacks used the UMLACA or more primitive devices.

On the night between August 20^th and 21^st the regime launched a wide scale attack as part of its Rif Dimashq Offensive. In response, a Liwa Al Islam chemical unit positioned itself in a clearing west to Irbin and proceeded to attack government forces advancing in Qabun and Jobar, using the chemical UMLACAs. Part of this attack was recorded on video, and later shared among acquaintances until leaked.

At some point the unit decided to target Zamalka. Several sub-scenarios may explain this decision:

• Scenario 3.1 – Misreading the battle situation. In this scenario the group is not composed of locals, and were misinformed to think Zamalka is an abandoned area (like Jobar and Qabun). They detected (or thought they detected) Syrian Army advances into Zamalka and attacked them.
• Scenario 3.2 – A mistake in orientation. In this scenario the group believed it is targeting Qabun and Jobar, but they miscalculated their orientation by 90 degrees, attacking Zamalka instead. Important to note that the whole area has been disconnected from electricity, which would make it easier to make such a mistake.
• Scenario 3.3 – False-flag contingency plan. According to this scenario, the team was trained in advance to perform a false-flag operation on rebel residential areas. This was a contingency plan to be activated in case of a major deterioration in the battle.
• Scenario 3.4 – Rogue false-flag. According to this scenario, the group took an unauthorized decision to use their chemical capability to bring international intervention and change the course of the war. In this case the videos were intended as an alibi, to give the impression of a mistake. 

In this scenario, Moadamiyah was attacked only by conventional weapons, and the sarin findings there are a result of contamination and patient evacuation from Zamalka.

Two side notes about false-flag attacks:

1. The term false-flag is often associated with fringe theories, and rightfully so – false-flag attacks are a rarity. However, in the case of the Syrian civil war it is actually the expected behavior: Once the US made the explicit link between a regime chemical attack and military intervention, it would be unreasonable to assume that of the many extremist opposition groups operating in Syria, not even one would try to exploit it.
2. Many believe a scenario involving an opposition attack “on their own people” is implausible. (a) In the context of a war where hundreds die each week, there will definitely be people who would find a sacrifice of this size justified in order to stop the killing. (b) Some of the fundamentalist groups are in direct war with the more secular FSA factions. They could have viewed the FSA-dominant Zamalka residents as legitimate targets. (c) These groups have a high percent of foreign fighters, who may not feel as attached to the local population.

Fit to evidence:

This scenario fits all evidence well.

Scenario Plausibility:

Unlike the regime-attack scenarios this makes perfect sense in terms of motivation. However, it has two weak points:

1. While theoretically possible, producing sarin on this scale is something that was not done before by a non-state actor.
2. While the opposition seems to have gained access to practically every heavy weapon in the Syrian Army’s arsenal, there is no evidence yet of them seizing UMLACAs.

Update - Foreign Intervention Scenario

In the discussions below, several contributors suggested scenarios involving foreign intervention on the side of the opposition. This could be relevant in three aspects of the attack: (a) Sarin production, (b) delivery devices, and (c) execution.

The last two do not seem very likely:

• Execution – There is no doubt that a large part of the opposition is composed of foreign fighters, especially among the fundamentalist factions. However, these are usually individual volunteers or militants from similar organizations in other countries. While it is theoretically possible that foreign troops or covert agents are assisting the opposition on the ground, there is no evidence that this is the case. There is also not much reason to bring troops into Syria and risk war, when there are enough combatants in Syria that can be trained abroad.
• Delivery devices – Since there is strong evidence that the attack used UMLACAs, and that these were developed by the Syrian Army, there doesn’t seem to be a likely scenario where foreign countries provided the delivery devices.

This leaves us with the option of foreign support in sarin production.

Since there is not much evidence supporting or refuting foreign involvement in sarin production, we can only evaluate it on its a-priori plausibility. There are a few possible sub-scenarios to consider here:

• Help in acquiring chemicals necessary for sarin production – Chemicals that are further along in the sarin production process are more heavily regulated, and a seller receiving large orders of these should alert authorities (as Al-Nusra discovered in Turkey). If buyers know that authorities are willing to look the other way, they could order the more advanced chemicals at larger amounts. This would result in an easier production process, less failures, larger amounts and higher quality.

• Providing lab equipment – High quality lab equipment is essential for mass production of sarin. Obtaining such equipment would be easier for a country than for an underground group.

• Guidance and know-how – Countries may provide access to chemistry specialists and knowledge, which are not easily accessible to an underground organization.

• Actual production – Last, a country can produce sarin (or more likely, its immediate precursor Methylphosphonyl Difluoride) in its facilities and provide it ready made to the opposition.

Doing any of these would be a violation of the Chemical Weapons Convention, and constitute a significant risk on the part of the supporting country. This should mean that more “obvious” support, such as sarin production or specific guidance is less likely. However, ignoring acquisition of suspicious chemicals and lab equipment is more easy to explain as unintentional, and may be done by government officials without receiving explicit approval (e.g. similar to the Iran-ContraAffair).

So which countries could be relevant here? The three immediate suspects are those most involved on the opposition’s side: Qatar, Saudi Arabia, and Turkey. Turkey has arrested Al-Nusra operatives trying to produce sarin, which indicate it is probably not cooperating in such an effort, and Qatar seems to have stepped down its involvement. Saudi Arabia, however, does seem like a possibility, and especially its head of Intelligence, Bandar Bin Sultan, who is reported to be directly involved in toppling the Syrian regime and convincing the US that Syria used chemical weapons.

In terms of evidence relevant to foreign intervention, we only have the Syrian report of a chemical cache found in Jobar. This included bags of Lye manufactured in Saudi Arabia. Lye could be used to destroy sarin in case of spills, but is otherwise a multi-purpose chemical not related to sarin production.

It is important to note that while the whole foreign intervention scenario is mere speculation, it does have some value: The complexity of underground production of large quantities of sarin is currently estimated to be the weakest point in the plausibility of the rebel-attack scenario. The addition of the foreign assistance option provides another way for the opposition to acquire sarin, and therefore strengthens this scenario and increases its plausibility.

Summary: While there is not much evidence to either support or refute foreign assistance, it extends the range of options for sarin acquisition, thereby increasing the likelihood of a rebel-attack.

Conclusion: At this point the rebel attack theory seems far more likely than the regime attack theory. Please help by sharing your feedback and scenario proposals.

Prior Alleged Chemical Attacks

If this is your first time here, I recommend starting from the conclusion page.

A common claim made for the regime attack theory is that it was just one of many chemical attacks. For example, the UK Intelligence assessment claims lethal chemical weapons were used in 14 occasions. If indeed the regime has used chemical weapons in the past, it would constitute very strong circumstantial evidence for its culpability in the August 21^st attack. To evaluate the strength of such evidence, I analyze in this post the reports of previous chemical attacks.

A major limitation in examining these reports is the high motivation among the opposition to convince the world that chemical weapons are used by the regime. From the moment Obama issued his red line in August 2012, and especially after his direct threat on December 4^th, 2012 ("if you make the tragic mistake of using these weapons, there will be consequences"), the opposition started frequently giving erroneous reports of chemical attacks. For example, in this video  a doctor shows a chemical detection kit and describes it as a chemical weapon captured by the FSA (similar examples here). In this video from one week after the red line speech, Syrian Surface-to-Air missiles seized by rebels are reported to be chemical missiles (minute 1:56). More examples are described below and many more can be found online.

This is further complicated by the fact that many nerve agent symptoms may be caused by other types of injuries (e.g. foaming), or by common medical treatments (e.g. morphine causes miosis), and can also be easily faked. Even sarin blood and urine tests can be faked by drinking diisopropyl methylphosphonate, a chemical available on the market (see the comments for details).

One thing that becomes clear when reviewing videos from claimed chemical attacks is the large number of incidents in which victims show symptoms that seem to be caused by a strong irritant.

Differentiating between an irritant and a nerve agent is not difficult. Nerve agents affect the nervous system and cause loss of control of the body. Victims will often appear calm or sedated and report mild symptoms such as loss of vision, dizziness, and nausea, while the more advanced symptoms such as convulsions and foaming are normally accompanied by loss of consciousness. The progression of symptoms is generally rapid and painless.

Irritants, on the other hand, cause damage to sensitive organs, resulting in aggressive conscious responses such as forceful coughing, vomiting, and noticeable suffering.

To explain this inconsistency in symptoms, it is often suggested that a combination of sarin with a weaker agent has been used, in an attempt to create some less-than-lethal weapon. As described in detail in Brown Moses’ interviews with chemical experts this is highly unlikely: Sarin is designed to kill, and has no other effective use. It is also chemically unstable and should be kept away from other chemicals.

Keeping this in mind, let’s analyze the reports one by one:

Daraya 6-Dec-2012

This is the first claim of a chemical attack, made 2 days after Obama’s direct threat (if anyone has an earlier report, please share).

According to this report “the regime forces shelled Mohasan and Buomar by Phosphorous Bombs and threw toxic gases in Daraya”.

There are no videos from Darayya claiming a chemical attack, but there were definitely clashes going on at the time. There are videos that clearly show a White Phosphorus attack in Mohasan.

Conclusion: No evidence of a chemical attack. Possibly White Phosphorus.

Aleppo 8-Dec-2012

A video showing burn victims claimed to be from a regime chemical attack.

Nerve agents do not cause burns of this level, leaving us with the option of mustard gas (the only chemical weapon in the Syrian arsenal that is not a nerve agent). At high concentrations Mustard gas would cause severe respiratory problems, which are not present here. At lower concentrations blisters would develop over a few days, causing severe pain.

Conclusion: No evidence of a chemical attack. Probably misrepresentation of the cause of injuries.

Darayya 22-Dec-2012

The regime reports losing 7 soldiers to a rebel attack with a yellow toxic gas. No evidence is presented.

Conclusion: Unverifiable

Homs 23-Dec-2012

This is the first of many attacks that are documented by video and seem to show similar symptoms, including severe respiratory irritation, vomiting, and pain. The agent is described as a “white smoke”.

This video (minute 1:14) and this video (minute 1:08) seem to show normal sized pupils.  

While it is theoretically possible to be exposed to nerve agents without suffering miosis, this seems to be very rare in practice: 99% of victims of the sarin attack in Tokyo exhibited miosis.

Some speculated that “Agent 15” was used in this attack, but the symptoms seem highly inconsistent.

Victims are all males of fighting age.

Conclusion: Irritant used on a military target, possibly White Phosphorus. 7 reported dead, but no evidence this was the cause.

Khan Al Assal 19-Mar-2013

This attack is very different from the other attacks described here, in several aspects.

Following this attack, the Syrian government requested a UN investigation. This was delayed following requests by Western governments to allow the UN team to visit other sites, which was initially refused by Syria. During the negotiations, Russia sent its own investigation team which found traces of sarin and blamed the opposition. Of course, the area is under control of the Syrian Army and the investigation was done by a Syrian ally, damaging the reliability of this report. 

An agreement was eventually reached as to which sites could be visited, and the UN team was sent to Syria. However, upon its arrival the August 21^st attack took place, and the team was diverted there. On September 25^th they returned to complete the investigation.

This attack is also far deadlier than all others, with 26 deaths compared to none or low numbers in other attacks. It is also the only one with civilian casualties.

Videos from the attack are available but seem to have been taken a long time after it occured. This video shows one victim foaming at the nose. In this video symptoms are reported to include “immediate fainting, convulsion and death”, which are consistent with nerve agents and not with irritants.

The victims seem to be a mix of Syrian soldiers and regime-supporting civilians (One blames ‘terrorists’ for the attack, and another is wearing the official Syrian flag on her hand). An opposition source blamed the attack on the regime, claiming it was a false-flag attack on the regime-supporting town.

Some eyewitnesses report smelling chlorine, which is a simple chemical weapon to deploy, and was used numerous times by Iraqi insurgents. However, the symptoms reported (convulsions, immediate fainting) are inconsistent with chlorine, and chlorine symptoms like skin or eye irritations were not reported. The chlorine smell could theoretically be the result of low-budget sarin production, but in that case many other smells should have been reported.

Conclusion: Seems like a lethal chemical attack, possibly sarin, possibly chlorine. Not enough evidence to determine culpability, but a rebel attack seems much more likely.

Otaybah 19-Mar-2013

Three videos from this event: One showing an unconscious victim foaming at nose, another showing victims with no specific symptoms, and another showing two bodies claimed to have been killed by a chemical attack. All are males of fighting age.

Conclusion: Not enough evidence.

Adra and/or Douma 24-Mar-2013

According to the Le monde article 39 people were affected and two died.

This video shows mild convulsions, and its title describes “chemical and phosphorus weapons". Another video shows a bloody discharge from the nose. While nerve agents can cause bleeding from mouth and nose, this seems to be rare and all records of foaming from Zamalka mentioned white foam.

Victims are males of fighting age.

Conclusion: Not enough evidence. Possibly white phosphorus

Jobar 7-Apr-2013

One video found claiming sarin or VX, while showing a victim coughing violently, and describing nine victims fainting, losing eyesight, and coughing severely. Another victim has miosis and is filmed from three different angles and lighting conditions, possibly in an attempt to look like multiple victims. Since he is not exhibiting any other symptoms, this seems very dodgy.

Conclusion: Probably an Irritant misrepresented as nerve agent.

Jobar 12-14 Apr 2013

This event was not documented but it is reported as one of two events for which the French government received blood samples that tested positive for sarin.

This is probably related to an event described in the Le Monde article:

On April 13, the day of a chemical attack on a zone of the Jobar front, Le Monde's photographer was with rebels who have been waging war out of ruined buildings. He saw them start to cough before donning their gas masks, apparently without haste although in fact they were already exposed. Men crouched down, gasping for breath and vomiting. 

These symptoms suggest the use of an irritant rather than a nerve agent, and the article gives no explanation why a professional photographer did not document an event of such importance (adding to many other inconsistencies in the Le Monde article which I’ll hopefully get to in the future). It therefore seems there is no way to know the source of these samples.

Conclusion: Unverifiable. Possibly irritant, possibly sarin.

Sheikh Maqsoud (near Aleppo) 13-Apr-13 and Saraqeb 29-Apr-13

These two attacks were analyzed by Brown Moses here, here, and here. Both attacks used the same small white canisters, and in Saraqeb a helicopter is seen dropping something that leaves a trail resembling that of White Phosphorus.

Symptoms shown in videos include severe coughing, vomiting, miosis, foaming and convulsions.

Samples from 13 Saraqeb victims were taken to turkey and tested negative for sarin. France, on the other hand, reported finding sarin in samples from Saraqeb and Jobar (13-Apr-2013, see above). The manner in which the samples were taken is unclear.

Following the Saraqeb incident, the regime accused the opposition of carrying two false-flag chemical attacks in the town. One in which powder was thrown on civilians, and another in which they brought hostages to a site of a chemical accident. No evidence was provided for either claim.

Conclusion: Contradicting evidence makes these events difficult to figure out, but the evidence for sarin seems weak.

Barzeh 26-Apr-13

The regime reported that Syrian troops were attacked by a chemical shell. The symptoms mentioned include asphyxia, nausea, foaming of mouth and nose, loss of consciousness and indirect contamination of medical personnel. Of all cases examined here, this is the only report where many sarin symptoms are described accurately. Of course, all we have is the news reports and no evidence.

Conclusion: Unverifiable.

Qasr Abu Samra (near Homs) 14-May-2013

According to this article, this is one of four instances in which the US believes chemical weapons were used. I could not find any evidence of such an attack.

Conclusion: Unverifiable.

Adra (patients treated in Douma) 23/24-May-2013

This incident has a large number of videos, which show unconscious victims, one case of convulsions, one case of miosis (also seen here), another case of foaming at the mouth, another case of disorientation accompanied by miosis. Other victims do not present miosis, which as mentioned above is atypical to nerve agent exposure.

This incident is unique in that in one of the locations medical personnel are treating patients while wearing gas masks. Something that was not seen in any other video, including Douma videos from the August 21^st attack.

All victims are males of fighting age.

Conclusion: Unclear. These are the only videos showing significant symptoms consistent with nerve agent exposure. However, not enough evidence is available to determine this is indeed the case.

Harasta 26-May-2013

This video seems to have been taken a significant time after the attack, with no atmosphere of emergency. The doctors report treating 200 people with symptoms of vomiting and dizziness. Since (a) vomiting is an advanced symptom of sarin accompanied by other symptoms that are not reported, (b) few deaths were reported, and (c) no second-hand contamination to medical personnel is reported, this does not seem to be a sarin attack.

An earlier video shows one victimg shivering and in pain, and another shows several victims with no specific symptoms.

All victims are males of fighting age, some wearing military vests.  

Conclusion: Unclear. Symptoms seem to go beyond those of the other irritant cases, but are still not consistent with a nerve agent. Perhaps some strong riot control agent?

Al-Otaybah 26-May-2013

This video shows a doctor describing symptoms of miosis and convulsions. He shows blood samples which are to be sent for analysis to prove the use of chemical weapons, and then mistakenly describes a chemical detection kit as a chemical weapon.

Conclusion: No evidence of a chemical attack.

Jobar 27-May-2013

This video shows several males of fighting age suffering from respiratory irritation, with the cameraman claiming sarin was used.

A civilian girl is shown breathing heavily, but does not show other symptoms.

Conclusion: Possible irritant use.

Zamalka 19-Jun-2013

Reported here. One video found, showing a victim with bloody foam  – somewhat similar to a symptom seen in the 24-Mar-2013 attack, and not typical to nerve agent exposure.

Conclusion: Unclear. Probably not a nerve agent.

Adra and Douma 5-Aug-2013

These two attacks were already analyzed here. More videos can be seen here and here, with a victim describing a “red color”, “sulfur smell”, blurred vision, and suffocation.

While the Adra victims are males of fighting age, exhibiting similar symptoms to other attacks, the Douma incident seems to be of a much wider scale and involving many civilians, although with light injuries and no deaths.

Conclusion: Adra seems like a typical irritant attack on opposition forces. Douma could be White Phosphorus smoke from a fighting zone reaching a civilian area.

Summary of Findings

1. In multiple cases, opposition fighters are seen affected by some kind of irritant. Phosphorus is sometimes mentioned, but may not be the only chemical involved.
2. These cases are correlated with clashes between the regime and the opposition in the area, and nearly all victims are males of fighting age.
3. The number of deaths reported in these incidents is very small and unverified.
4. In most cases, the symptoms are not consistent with use of a nerve agent, including severe coughing, vomiting and pain that are not accompanied by loss of consciousness.
5. Despite multiple claimed chemical attacks and the high availability of cameras, there is no documentation of an actual attack taking place or the munitions used. In a few cases, evidence of white phosphorus is seen.
6. In multiple cases blood samples have been taken from victims and sent to countries friendly to the opposition. Given the political interests, it is safe to assume that when not reported otherwise, the results were negative.
7. The Khan Al Assal attack is very different from the other cases: (a) victims are Syrian soldiers and regime-supporting civilians, (b) a significantly larger death toll, (c) the symptoms described are consistent with the use of a nerve agent, (d) environmental samples tested positive for sarin, and (e) the regime requested a UN investigation.
8. In general, the effects of these attacks are negligible compared to the extent of the conflict. So if indeed a chemical attack is the cause it would have small military gain for the regime, but significant political gain for the opposition. This of course is not evidence in itself, but it does mean that much stronger evidence is required before accepting that the regime has used lethal chemicals.

Conclusion: The regime has used White Phosphorus and possibly less-than-lethal chemical agents against opposition forces. There are no reliable indications for nerve agent use by the regime. There is some evidence indicating the attack in Khan Al Assal was an opposition attack with a lethal chemical weapon.

Did I miss anything? Please share your evidence and analysis and help improve the conclusions.

In this report I found the compilation of evidence done by Wikipedia, acloserlookonsyria, Brown Moses, and George Washington University very helpful. All evidence was independently verified.

Update:

Gleb Bazov directed me to this research from the Harvard Sussex program on chemical and biological weapons. It analyzes the allegations of chemical attacks (including some that I missed), and reaches a similar conclusion that the Syrian Army has been using less-than-lethal chemical agents against opposition forces.

Sarin and the Syrian Opposition

If this is your first time here, I recommend starting from the conclusion page.

Since it seems that over 500 kg of sarin were used in the attack, and the regime is known to have large amounts of sarin, this is circumstantial evidence supporting the regime attack theory. To evaluate its strength, we need to estimate the likelihood that the opposition could obtain sarin.

Update: The final UN report published after this analysis was written provided strong evidence that the opposition has obtained and used sarin on several occasions (see here).

Direct Evidence Connecting the Opposition to Sarin 

Carla Del Ponte, a senior UN investigator who interviewed victims of previous attacks, shared her personal opinion that chemical attacks were initiated solely by the opposition. I usually ignore opinions brought without concrete evidence, but unlike other reports, this comes from a very reliable and relevant source. Given that the clarification the UN issued following her interview does not contradict her claims, this testimony may have some value.

Update: In a later interview (minute 4:17), she reiterates her assessment, and explains this is what the initial evidence indicates, but recommends waiting for the final report. (Found by Jim Dobbin).

In a chemical attack in Khan Al-Assal, whose victims were Syrian troops and regime-supporting civilians, a Russian investigation determined the use of sarin. While the report attempts to lay blame on the rebels, it does not share its evidence and is therefore difficult to evaluate. More here.

On June 2nd Syria reported seizing 2 canisters of sarin. Additionally, on two occasions (Barzeh April 26th, Jobar August 24th) Syria reported a chemical attack on its troops, describing symptoms typical to sarin.

Since evidence was not presented, these reports are of no use.

On January 2^nd an FSA official claims to have chemical capabilities, but does not provide specifics.

This video also shows an activist claiming to have chemical weapons, but does not provide specific evidence.

This well-known video threatens Assad’s supporters with chemical weapons, and demonstrates their use on rabbits. The video uses the term “Reeh Sarsar” which is later mentioned in the Liwa Al Islam rocket launch videos. Assuming it is authentic, it would only demonstrate intent and not capabilities, since poisoning an animal in a small confined space is very different from an operational chemical weapon. Additionally, the chemicals shown are not directly related to sarin production.
Update: In a comment below Charles Wood showed this video most likely shows chlorine poisoning.

Another video shows two Liwa Al Islam operatives coordinating bombardment of a building. One of them says he will bring just one sarin (translation verified independently). It is heavily edited, and can easily be fabricated, but still worth noting.

Two days after the chemical attack in Ghouta, the Syrian TV broadcasted two phone calls, which were presumably intercepted by Syrian Intelligence. In the first a rebel reports to his Saudi sponsor that his group carried a chemical attack in Homs. In the second a transfer of sarin is coordinated.

They seem authentic but are not strong enough as evidence: The first does not include details about the attack, and may be an exaggeration to impress the sponsor, and the second can possibly be interpreted as sarin antidotes.

Evidence of Attempts to Acquire Sarin

The most important evidence for opposition involvement in acquiring sarin, is the arrest of 11 Al-Nusra operatives in Turkey in May 2013. They were initially reported to possess 2 kg of sarin, which was later claimed to be antifreeze. They were eventually charged with attempting to acquire chemicals for the production of sarin. Interestingly, the prosecution listed these chemicals, including:

• Thionyl Chloride (SOCl2)
• Potassium Fluoride (KF)
• Methanol (CH3OH)
• Isopropanol (C3H8O)
• Isopropanolamine (C3H9NO)
• White Phosphorus (P4)

Comparing this to the sarin production process (in the appendix here), shows this is unmistakably a sarin shopping-list. The initial downplay as ‘antifreeze’ probably referred to the Methanol, which is one type of antifreeze.

Interestingly, only one week later, Iraq arrested five Al Qaeda operatives that were planning to produce sarin for use in Iraq and abroad.

Complexity of Producing Sarin

Last, let's examine the difficulty of producing sarin in underground conditions.

One way to examine the difficulty of underground sarin production is by examining the only case where this is known to happen: by the Aum Shinrikyo cult in Japan (source1, source2, source3). The cult started producing sarin and other nerve agents after failing to produce a biological weapon. They made significant investments in lab equipment intended to produce 70 tons of sarin. In practice, technical problems and government investigations limited their total production to less than 100 kg over their year and a half of operation.

However, Syrian opposition groups may have some significant advantages over Aum Shinrikyo:

1. Much weaker government supervision. Treating the opposition as a non-state actor would be inaccurate, as they have full control of some areas of Syria, making them the de-facto state there. This is critical: Aum Shinrikyo had stopped their process several times and destroyed products because of police investigations. 
2. Possible access to former Chemical Warfare professionals from Iraq, Libya or Syria.
3. Access to lab technology that is 18 years more advanced. This was specifically addressed in this detailed analysis of Aum’s chemical program, which estimated that current lab equipment would make a similar effort much cheaper. Quotes:
   “Such an effort might be well disguised or established on a smaller scale, however, by taking advantage of the development over the last decade of powerful, low-cost micro-production chemical capabilities”.
   “… the chemical industry has, over the last decade, introduced modular and flexible designs where reactions may occur in a solvent-free environment, at increased concentrations and in much smaller and less expensive facilities”.
4. International allies. This could prove very helpful when trying to obtain regulated chemicals or lab equipment.
5. Internet access, which makes information on sarin production widely available (e.g. details of Aum Shinrikyo’s process). This can significantly accelerate production - Aum Shinrikyo scientists spent much of their time in trial and error.

This indicates that sarin is not a chemical that “anyone can make in their kitchen”, as is often claimed. However, an organization with sufficient funding and trained professionals can produce large quantities of sarin within several months.

Conclusions:

1. There is strong evidence that Syrian opposition groups have attempted to acquire sarin, and some weak evidence that they succeeded in doing so.
2. There is some weak evidence that Syrian opposition groups have carried out chemical attacks.
3. Production of sarin in the quantities used to attack Zamalka is within the reach of well-funded underground organizations.

Did I miss anything? Please share your evidence and analysis and help me improve my conclusions.