December 10, 2013
Apparently I wasn't done with Syria's chem attack yet
There's been a lot of discussion on this, with Sy Hersh weighing in recently, so it's time to revisit. Short story, Syrian government culpability still seems clear from here, but the evidence has taken some twists and turns that are worth commenting on (if only because my graphic several posts below now turns out to have been based on inaccurate UN reporting and would appear to have been disappointingly misleading). Below the fold is a detailed analysis of the recent findings on this, and why I would argue the preponderance of evidence still points at the Syrian army as the culprit.
1. Use of improvised rocket munitions in Syria
Ok, from the top. Artillery rockets are type of weapon that have been common in battlefields since World War Two. The Russians call theirs Katyushas, the Allied troops under German rocket attack called them "Moaning Minnies." They are traditionally a long range area-effect weapon, not particularly accurate, but able to deliver a devastating amount of explosive if fired in large quantities. Unmodified artillery rockets continue to be used for these purposes today, and were used in large quantities in Chechnya, Lebanon, and most recently Libya.
Rockets, like everything sold and used in large quantities, tend to come in a few standard sizes, like dry cell batteries. The obvious reason is you want the replacements you purchased to fit the launcher equipment you have. The most common sizes in general circulation today defined by their rocket diameter, are:
107mm (aka Type 63): Produced by China, Sudan, Iran, South Africa, North Korea, Turkey, Egypt. Length 2.9m, Max range 8 km
122mm (aka BM-21 or "Grad"): Produced by Russia, China, Romania, Egypt. Length 7.3m, Max range 20 km+
Other, heavier calibres also exist, but are much less common: 140mm, 220mm, 240mm, 300mm, etc.
In Afghanistan, the Taliban harassing targets like Kandahar airfield used the 107mm almost exclusively, fired singly, off improvised launchers. While this was not particularly effective militarily it shows how almost any government or insurgent faction could purchase a quantity of these smaller calibre rockets if desired.
One of the reasons the 107mm rocket wasn't particularly effective for the Taliban is it isn't a particularly large warhead. It's really meant to be fired in large quantities, not singles. At 1.3 kg of explosive, it is about the equivalent of only two 81 mm mortar shells, with a lethal blast radius of 2-4m at best. (The only reason the Taliban really had any success at all against Kandahar airfield was the whole place was pretty much prefabricated structures.)
In urban settings in street to street fighting, the basic artillery rocket's use gets constrained. The buildings are tougher, the longer ranges aren't particularly effective, and both sides are looking for less for area effect bombardment of neighborhoods and more building-buster munitions they can use to seriously injure combatants behind thick walls. So it was no real surprise that the Syrian conflict has seen real innovation in the area of so-called Improvised Rocket Assisted Munitions, or IRAM, by all parties to the conflict.
An IRAM is basically an artillery rocket with the standard warhead replaced with a much larger-calibre explosive payload. This enlarged front end tends to make the weapon look a little like an ice cream bar or a popsicle. Ballistically, it's not particularly effective, with the same size rocket motor pushing a heavier and less aerodynamic projectile, and so it gives up a great deal of range in return for the much more significant blast effect.
Syrian ingenuity with improvised weapons has introduced a large quantity of 107mm-based IRAMs, which have been used heavily by the opposition forces in particular. The Syrian army, for its part, has produced both a 122mm IRAM (with an enlarged warhead diameter of 330mm, hereinafter referred to as a "122/330" and a larger variant possibly based on a 240mm rocket. The 122/330 seems to be fired mostly off a two-rocket truck-mounted launcher while the larger variant is fired off a single-rocket trailer launcher. These are both notably different looking from either 107mm-based IRAMs or unmodified rockets, in particular because they are painted the same consistent shade of white, and have the addition of a fin stabilization ring at the end of the popsicle stick tail that have led some to compare its shape to a bicycle pump. These are significantly bigger weapons and their effect on impact would undoubtedly be devastating, akin to a bomb dropped by an airplane.
This would all be military trivia, of course, if it hadn't become clear that in the sarin attack on Aug. 21 on a rebel suburb of Damascus called Zamalka, the delivery mechanism was a modified form of 122/330 IRAM.
2. The Aug. 21 chemical attack
We know this, because two damaged 122/330 IRAM bodies were found in that neighborhood and photographed, still buried in the ground, the morning after the attack. Witnesses reported several more impacts in the immediate neighborhood, perhaps as many as 12 all told. UN inspectors tested one of them, finding conclusive evidence of a sarin poison gas payload, which was consistent with the symptoms reported by hundreds of Syrians.
That would have been pretty good evidence, except for a couple really odd unforced errors by the UN in their reporting, and some basic skepticism that the Syrian government would ever do anything as stupid as an undeniable chemical attack with UN chemical weapons inspectors already in the city on Aug. 21. The only plausible explanation that has been raised for this has been the "use it or lose it" hypothesis, that at some level Syrian ground commanders assumed their use of sarin would soon be extremely curtailed and they acted tactically before that capability was lost to them. But the motivation for the attack still remains until today so opaque that a "false flag" attack by rebels on their *own* neighborhood with captured munitions could not be wholly ruled out on that basis.
How the rebels would have obtained sarin in such large quantities was one issue with that alternate hypothesis, however. Production of hundreds of litres of the substance (with a shelf life once mixed of days, at best) still seems likely beyond their capability, at least compared to what the government, which still controlled several chemical weapon-associated facilities in the Damascus area, would be capable of. But again, there was still some potential for doubt there, as well, so more evidence was required.
The UN investigatory report that resulted complicated the matter in two respects: first by bringing in evidence from a second more questionable site, on the other side of the city from the Zamalka impact points. This attack, in Al Moadamyeh, was alleged to have occurred on the same night. However, skeptics quickly pointed out the evidence from this site had some big error bars. In particular, no actual sarin was found in chemical tests, only sarin precursors, and no intact rocket body was found, only a single rocket motor (140mm, an unusual calibre but one that was both chemical-munitions capable and known to have been used by the Syrian army in the past using standard high explosive payloads). While an actual second sarin attack that night in Al Moadamyeh, in what is effectively a completely different location, should not be entirely ruled out, the evidence in retrospect really was not enough to elevate this event to higher than "suspicious," really, especially given that there was never any question that the Syrian government had occasionally shelled that area with conventional high-explosive artillery rockets from its nearby positions at the Al Mezzah airfield, which could also explain the forensic evidence.
That left the Zamalka attack on that night, and here again questions were being raised. The first related to the rockets' range. Once it was clear the attack had been by 122/330 IRAMs (which Syria civil war blogger Brown Moses has dubbed UMLACAs (Unknown Munition Likely Associated to Chemical Attacks) and others on the Internet have called "Eskimos" (after a popular form of ice cream bar, because of their shape... Syrians on the other hand apparently tend to call all calibres of IRAM "Volcanos"). Using standard ballistics calculations the skeptics raised questions whether a 122/330 IRAM wasn't going to make it much farther than a couple kilometres. This was confirmed recently by this video, which shows an undisputed 122/330 IRAM launcher firing a (high explosive, not chemical) rocket, with about 6 seconds on the tape between the light and sound of the resulting detonation reaching the launch site.* Again, this isn't a bug, it's a feature: in the urban fighting context, raw hitting power, not range, is what's needed. 122mm and 107mm rocket bodies are plentiful on the modern battlefield; targets for them in their traditional form, somewhat less so. It is logical that they would be repurposed for greater effect.
That said, the value of loading a chemical weapon on such a short range weapon seemed... unusual. The attack was at night, but the use of the 122/330 IRAM still involved the government bringing chemical weapons very close to the impact points. Because of uncertainty about where those points were, and where the front lines were at the time, the range information still seemed open to question at first.
This was compounded by the second UN report problem, what now appears to be a gross error in the report. The UN reported azimuth for the most intact rocket is given there as 285 degrees, or very close to due west (p. 23). Numerous commentators on the web quickly pointed out this value was contradicted by the photo evidence which show the rocket tail pointing close to directly away from an east-west wall, making its direction roughly north.
In short, that makes my initial diagram in this post wrong in three key respects, and rules out it being any use at all in confirming Syrian government involvement. In short, the left stick for the Al Moadamyeh attack should probably have had a question mark on it, and the right stick is both pointing in the wrong direction and is way too long. So although the UN report's attack azimuths did have a nice intersection in Syrian military-controlled territory, that is as it turns out, pure coincidence.
That said, I would argue sites like this one go too far in concluding this must therefore be a "false flag" operation. In the next section, I'll explain why.
3. What actually happened on Aug. 21?
When I wrote that original post, good information on point-of-impact locations, or military positions in that area, wasn't available. Fortunately, a lot of that has emerged since, and taken in the whole still points very strongly at the Syrian army being behind this particular chemical attack.
To recap: leaving out the questionable Al Moadamyeh attack reports entirely, we still have conclusive evidence of a volley of 122/330 IRAMs landing close together in Zamalka on the night of Aug. 21, the same night hundreds in that area died in an undisputed sarin gas attack. While likely a larger number were used, 2 new 122/330 IRAM impact points with largely intact rocket bodies were identified and photographed the next morning, with one of those bodies later tested and confirmed to have carried sarin by UN inspectors.
For those feeling like plotting at home, the locations of those two confirmed IRAM impact points, in decimal degrees and MGRS coordinates, are given here below:
Impact 1: 33.5204N, 36.3559W (MGRS: 37S BT 5443 1211)
Impact 2: 33.5207N, 36.3574W (37S BT 5456 1214)
Now, we know from the video evidence of the conventional launch that these probably could only have been fired from 2-3 km away, and the impact photos clearly show a northerly launch direction (I wouldn't put too much more precision than that, as these weapons would already have air bursted and the body started to tumble when they hit the ground, so any estimate of azimuth from the direction the tail is pointing much better than "North" or "Northwest" probably isn't of great deterministic value). So for the "Syrian-army-did-it" hypothesis, one has to find a plausible launch location less than 3 km away to the north, but that involved knowing a little more about the battle lines as they stood on that day.
A key piece may have come with this analysis. A must-read post on many levels, it describes in great detail Syrian army operations in an area immediately north and west of the impact location in August, in the Jobar suburb of Damascus, close to the impact site. It is particularly relevant to this discussion, though, in three respects.
The first is the section "Northward Attack," which shows Syrian army forces attacking into a suburb from the south from a staging area around a specific building from which some of the filming is conducted, some time between Aug. 6 and Aug. 21.
The second is in the section "Splitting Jobar and Zamalka", which shows the use of 2 other prominent buildings in the same block being used as observation and filming posts by the Syrian army, no more than a few days after Aug. 21. The coordinates for all 3 buildings are given again here for reference:
Observation Point 1 ("Northward Attack"): 33.5404N, 36.3440W (MGRS: 37S BT 5338 1436)
Observation Point 2 ("Splitting Jobar..."): 33.5358N, 36.3419W (37S BT 5317 1385)
Observation Point 3 ("Splitting Jobar...") : 33.5368N, 36.3453W (37S BT 5348 1396)
Mean Centrepoint: 33.5377N, 36.3436W (37S BT 5334 1406)
This mean centerpoint is roughly 2250m away from the sarin impact points, within the estimated max range of the 122/330 IRAM, at a reverse azimuth of 325 to 330 degrees, or very close to north. This direction would be roughly consistent with the photographic evidence of the points of impact. (This does not mean that the launch came from this exact centrepoint, of course: a more plausible location tactically would be closer to the large cloverleaf intersection to the northeast.)
Staging areas and forming up points like these tend to be in locations that are relatively secure from incursion and fire. What this footage shows conclusively is that a heavily armed Syrian army unit had firm daytime possession of a relatively secure area within effective range of the weapons system used, in the approximate direction indicated by the point-of-impact remains, in the immediate period both before and after the night in question. Further, given that the rough direction is certainly approximately north, the chemical attack must have been launched from very close to this unit's location at the time, on the day before it launched clearing operations to the south of the position in question. Those operations would undoubtedly have benefited from supporting bombardment of the nearby hostile suburb of Zamalka (the operation is described as intended to split insurgents in the Jobar neighborhood from those same supporters in Zamalka, in fact), whether conventional or chemical, to dehouse and drive out the inhabitants who could support the rebels. That unit, for the record, is identified by Open Source IMINT as an element of the Syrian Republican Guard Armoured Division, which undoubtedly would have chemical munitions in its artillery stores and as the elite unit of the Syrian army would have plausible access to considerable quantities of sarin and the 122/330 IRAM.
Annoying flaws in the original UN report aside (that this blog played a now-lamentable role in supporting), there is still more than enough evidence here to point to the government and no counter-hypothesis that seems more plausible.
*Contrary to what we're taught as children when counting the time between lightning and thunder, sound actually travels around 330 m/s, or a kilometre in 3 seconds and a mile in 5.
"endearingly macho" -- Mark Steyn
"wonderfully detailed analysis" -- John Allemang, Globe and Mail
"unusually candid" -- Tom Ricks, Foreignpolicy.com
Bill & Bob
Ghosts of Alex