More on the Khan Sheikhoun Radar Track

March 5, 2018

One of my sources tipped me to some interesting points regarding the Khan Sheikhoun CW incident and investigation that I had overlooked, spurring me to do some visual analysis on the alleged radar flight tack that I couldn't do previously. Betweeb it all, I find new reasons to doubt bot this evidence, and also my own reading of it so far.

This has been one of two most-central and oft-repeated points in my own case, the "smoking gun" points, more or less:

1) radar (and some other evidence) shows the jets never flew over the city, so could not have dropped any bombs there.
2) the wind needed to spread any sarin blew exactly opposite of the direction people were allegedly affected (see explainer).

I put this one first in writing because it was easier to see and explain (as it turns out, a little too easy) But I always knew the radar part had a limited basis that could be erroneous, whereas the wind aspect always had an undeniable depth of proof in my mind. Now the radar issue can be shown - belatedly - to involve error and confusion and is now demoted. At least temporarily, I consider there to be only one clear leading point in the wrong wind, with many strong supports, probably still including this one...

That revision is worth explaining for anyone interested in these issues, and so this post. First, some background that almost deserves its own post, but helps put all this in context anyway.

The "Jeff" Flight Track Graphic: How it Came to Be

Capt. Jeff Davis, a Defense Department spokesperson, gave a statement on April 6th to announce the missile attack on Shayrat airfield that morning and the reasoning for it. He notes the targeted base "was used to store chemical weapons and Syrian air forces," but not that the attack was launched from there, despite the presumable availablility of radar data. Was that just an oversight, or did it reflect a lack of evidence for that - or even the presence of proof against it - at that time?

That night, apparently, a graphic was prepared to accompany the more explicit claim aired the next day. This is the image I've been citing, wrongly as from the White House report, and/or from that press conference. It came separately from and later than both, with a manner of entry worth a brief consideration.

I haven't found a proper defense department publication of this image, even on their social media accounts. An April 7 DoD article adds "an official said there is "high confidence" the airfield is where the Syrian regime’s aircraft took off for the April 4 attack." It shows base damage, but not the flight track. No other article I found on the DoD website features it.

All April 7 media reports showing the map just say it was released, giving no original source except other media reports. ABC News (1 2) - BBC - LA Times Maybe it was sent to several media outfits directly, or directly to ABC news, whose report adds a comment from the same Capt. Davis: "Officials decided to declassify the photo to prove that Syria's Shayrat air base was linked to the chemical attack, Capt. Jeff Davis, a Defense Department spokesperson said Thursday night." That should be after and apart from the press conference. Further, ABC's image, perhaps alone, contains relevant exif metadata:

title: ht-released-flight-track-jc-170406

Authors: Jeff
Created: 4/6/2017 11:46 PM (time presumably local)

And probably the earliest publication of the flight track is in an ABC News tweet from 8:52 PM (California time) on 6 Apr 2017. This shows the image in hand at 11:52 east coast time, some 11 hours before their next-day article, but just 6 minutes after image creation.

So creator "Jeff" is presumably the same Capt. Davis, lodging comments that night, through someone at ABC. It's not clear why a spokesman would be tasked with graphics preparation at night. The use of first name suggests a home computer was used, not one at the Pentagon, which would be interesting. As I gather, Jeff Davis isn't just any spokesperson, but the overall director of media operations for the pentagon. Someone that senior shouldn't normally sit up late preparing graphics on his own computer, when there should be many people under him with much better technical skills in that kind of thing.

This seems to be outside normal channels, maybe avoiding the usual audit trail and oversight. It's worth wondering if that chance was taken to alter the data. If there were a discrepancy to cover, it would seem to be the excessive distance issue (see analysis below). If so, it might be best to just insist the jets were over/near/around/over the town, and never show the proof. No one would probably rake them over any coals about it. Trying to show the problematic evidence altered to look like positive proof sounds tricky and dangerous.

But if the jet was 5km or more from the town as some evidence suggests, perhaps the map was altered to fudge it north and then shown. As it is, it wasn't by enough to really work, but ... but why then? Hm. Maybe they would decided to show something fake instead of nothing, just to placate the public with some hopefully indecipherable pixels.

Maybe it's not fake and Davis, or someone through him, released this unnecessarily to show the problem, as per a whistleblower? This and other options might make sense, depending on the personnel involved, a lot of unknowns, and the real-world validity of these radar tracks. Mainly, let's stick to the content of the image and what we can learn from and about that. 

What It Seemed to Show

As the ABC report notes "The flight path appears to have taken a northerly track from the Shayrat airbase near Homs towards Khan Shaykhun in the Idlib province. According to the graphic released by the Pentagon, the plane appears to have circled the area around Khan Shaykhun and the area between the two cities multiple times."

As I've repeated, it appears to pass "around" Khan Sheikhoun but not over it as required. The opposition story has two jets or at least two passes, one dropping conventional bombs, the other sarin (disputed: one or two jets, order of drops, number of chemical bombs and whether they also contained chlorine). The distance to sarin crater is around 3km, only about 1 km from the edge of town. There's no way a gravity bomb would drift that far from straight down, off to the left. Out ahead to the east, with the jet's trajectory, maybe. But not kilometers north. Here again is how I've been showing it - red line fudged a bit north from my imprecise reading, to be fair (in fact it might pass this close or closer, by the graphic).

This issues was largely ignored until the UN-OPCW Joint Investigative Mechanism (JIM) final report last year (see here for copies and notes and here and here for my reviews). Based on the "Jeff" graphic and "another aerial map" they were shown by someone, the report declared roughly what I did, but even more so: 

"The Mechanism had access to another aerial map depicting the path of an aircraft alleged to have been in the airspace around Khan Shaykhun between approximately 0644 and 0651 hours on 4 April 2017. The aircraft was depicted as flying in a circular loop pattern in the vicinity of Kafr Zayta and north-east of Khan Shaykhun. The map indicated that the closest to Khan Shaykhun that the aircraft had flown had been approximately 5 km away." (Annex II, point 28)

That seemed like the same but rounded up - the tracks loop to the northeast of Khan Sheikhoun after passing right over Kafr Zita to the south (see composite views below), never coming closer than 3 km from the sarin crater, and just about along the town's outer edge. They saw it as even further out, which should be fatal to the case, but they had an expert vaguely make it sound plausible (as my reviews explain, the effort clearly fails and they're caught in a lame dodge). 

So I had a visual and then this confirmation of a serious problem. But now that I look closer at the visual part...

Analysis Basics 

I didn't realize there were free programs comparable to Photoshop, which had on an old computer, but finally downloading the useful graphics program GIMP 2, I can finally layer images again and do a proper analysis. Now, nearly a year after the event that's still so murky, and after no one else I know of has done this kind of analysis (some like Bellingcat's Triebert have done overlay comparisons but little analysis).

To line up the flight path with reliable satellite imagery, I compared the roads, city outlines for Homs and Hama, and town placements for Khan Sheikhoun and Al-Latamneh (everything given on the DoD map). I used GIMP 2 with the flight tack on top (varying opacity) and an underlying layer from Google Earth. Here is the analysis for the starting portion at Shayrat airfield, with roads partially painted in manually (I forgot to turn that on), Homs city outline painted roughly in black. 

A marked horizontal stretching was evident, apparently on the map and the radar track as if both were layered into one image that was then stretched. (see how the returns change from oval to circle with the adjustment). This was pretty clear from the start, with even the lettering being stretched in the "Jeff" graphic. For what it's worth, here is just the flight track part, roughly corrected (no comparison here, just narrowing it 'til the dots look circular)

As for the radar returns (red dots) - it's hard in spots to make out which dots connect into a flight path, and in some spots it's hard to see how it could connect to anything. This might be an unprofessional mess in fact - I'm not expert enough to say.

There's also irregular spacing of returns, some far apart and others lumped together. They should be pretty regular, barring speed changes. I don't know if maybe distance or weather, etc. cause some returns not to come back or come back a little scrambled. This could also mean - for example - the path was re-traced by hand, with the creator placing dots with mouse clicks and doing it irregularly. But why re-trace the map instead of showing the original he traced off of? This remains an area of question.

Identifying the Starting Offset

The main thing, however, is the  starting offset at Shayrat base - the one concrete problem with the map line-up I can identify so far. It might be the only one ever;  this is the only spot where we can put the jets and the ground together somewhat reliably, and we can see they aren't lined up right.

This shouldn't even show the jets on the runways, let alone back before them, almost on the highway. They should appear only after takeoff, upon reaching the right altitude to reflect radar impulses. How far out isn't certain, but perhaps in the white circle here, as they clearly depart E-SE. In fact between the first returns is a short line each along the same line as each runway - maybe a bit of each jet's takeoff, crossing over each other. The dots here are closer together, I propose because the jets were using much of their speed to climb, before they leveled out and started moving further across the map. (note: the extra shape muddled in here - either a plane or a square cross, happens to line up with my white circle, but according to the returns, it's just some spot a ways east of the airport...)

I measure the offset between first return where it is vs. about where it should be at 7.1 km, 103 heading to correct. Again that's inexact, but probably close to right. I may update this if I ever get a more useful (expert) opinion on where the white circle should really be.

That seems quite interesting ... but what it means is a bit of a headache.

What That Was Hiding

I did the same layering and stretching for the area in question near Khan Sheikhoun, getting this similar result. Labeled cities KS and Latamneh here boxed in black. Highways painted on their own layer this time (I traced highway from Maardes not highlighted on Google Earth - it matches until I veer off tracing another road). Some town labels added for reference. 4 returns closest to KS are marked in red on the right-hand view.

Reading the nearest flight path: This was never done fully, and even now I can only make out three returns (B, C, D as labeled at right) forming a clear, gently curving flight path just southwest of Khan Sheikhoun, but from there - the east end could connect to either of the two dots boxed in lime green (one of which is point E) and branch off how it does - it's little matter. Considering our correction to the east, the west side matters more, and there's more variation between the points that could be A, with a possible missing return on the way to point B. Presuming a continuous motion, either aqua green line might fit, the northern one perhaps better. It might possiblt connect to one of the dots further south, but that implies an unlikely maneuver.

The path curving more from the north looks better and has the added feature of putting the jet right over the sarin crater, once corrected. This graphic explains it all:

Additional issues may exist, but would be harder or impossible to tease out without anything further to compare them to. For example even this adjusted set of returns might be stretched, skewed, scrambled, or possibly even made up, besides perhaps genuine and naturally imperfect. But if this offset is the only issue, ironically, the jet's pass over the release point is made to look like a pass south of town, making that a false lead after all. This obviously needs to be noted.

Is it Incorrect Even After Correction?
It's fairly possible there's a real discrepancy underneath this, as the JIM seems to have found, that was "corrected" to put the jet over the crater. If so, it seems this fix was obscured with another distortion, that offset creating the appearance of about the same obvious problem - the jet seems too distant. But this version of that problem falls apart on analysis of the offset. We might take that as the only alteration, but it could be a clever sort of bait-and-switch trick to obscure the real forgery. One effect it's had, intentionally or not, is to trick some (like me) into promoting this flight path, only to have it show what they want us to see in the end. 

The JIM's basis for calling 5km is more crucial now, and I don't trust them. Maybe they saw another map with the same basic issue, and just measured the apparent distance like I did, rounding up. That's not in their immediate apparent interest, nor in the DoD's to show it to them. Presumably, the JIM would have done everything possible to minimize the distance, so if they say 5 km from the town, it's most likely 5 km from the edge of town, and they're most likely rounding down, not up. That would mean the other map we've seen was different - which they didn't note. 

There are some reasons they might be erred in that conclusion, but let's consider they might have seen better data and read it right. Again, here's what they describe: "a circular loop pattern in the vicinity of Kafr Zayta and north-east of Khan Shaykhun," never passing closer than "approximately 5 km away." That's from the city, apparently, and not from the sarin crater. That sounds just like the non-adjusted track I've been citing, but corrected to be even further out. Below: closest passes in thick purple, other confused possible paths (area of other returns) in lighter lines - then that with transparency, overlaid with a green version shifted to fit the JIM description. In fact a shift of 5.3 km to the S-SE gives the best fit for 5 km out on all sides - and pretty much the only fit. It can't be further out and still loop "northeast" at all. Therefore, it's most likely this is about what they meant.

Now, that's just based on a description of a set of paths that might look different, but taking that as a reasonable basis for an implied JIM reading ... here is a conceptual graphic showing how their implied plots (as read by me, above), compared with what the DoD "Jeff" graphic seemed to show,  and what it shows after correcting the obvious offset (what the DoD wanted us to see in the end?). A triangle of given/read/decoded versions of each data point with the truth likely somewhere in the middle, probably closer to the south end. Each black line is the right difference between each red dot (right heading, proportional km scale, drawn from measured points on a map). 

So in my complex thoughts above, the original plots (southern dot) would be first be shifted northeast ~6.5 km to red dot "corr. for offset." That puts the jet over the sarin crater at one point. Then the offset in the "Jeff" graphic shifted it all in appearance ~7km to the W-NW, making it seem to not line up, perhaps as bait to get us swallowing the hook of that first alteration. 

In Review

If the 5km measure is closer, there must be another error in this version, like the flight track rotated a bit, skewed, etc. The other unseen track is preferable, but not gospel either and of course - it's unseen. So the JIM's unverifiable call becomes the best argument as to distance, with supports but no certainty. I don't like having to trust them for anything.

The point is weakened somewhat, but remains strong anyway. Many things agree in suggesting a proximity problem we just can't prove or quantify. We also have the Syrian pilot in the same JIM report swearing to a distance of 7-9 km (from what?), suggesting 5km (from town's edge?) was rounded down. But if the above is correct, he probably passed about 5 km from town twice. Maybe he was rounding up? Maybe he was the other pilot? One of the two had been later shot down by foreign-backed Islamists and went missing before the JIM could interview him.

So the case remains, but it doesn't wrap up neatly. And who "outside the choir" would believe the pilot, and who should trust the JIM?

I open the floor for thoughts on how to consider all this, but for the moment I'll be making no more calls to simply "look at the radar track it's south of town!" because that picture actually seems to show it over the sarin spot. Wind direction is undoubtedly strong point #1 now. I'm not sure if this issue is even #2, but maybe things will clarify in time.

The situation isn't clear at this point, but the possibility is raised the radar does show one jet passing over the alleged sarin impact crater and that, despite what the JIM thought, this is even correct. I rather doubt this, but it can't be ruled out. For those inclined to leap there, and those willing to consider the possibility ... That certainly doesn't mean the government dropped sarin gas and killed people. Consider the following points that would remain standing: 

- There's no sign at all of the second jet pass claimed by all witnesses. Is it the sarin bomb run or the conventional bomb run with no radar track remotely near the town?

- The opposition story still has enough problems (like the exact wrong wind direction) to possibly be true.

- The government still had less than zero motive and probably no sarin, while Islamists had all the motive and have never even been asked to give up their supplies of sarin.

- the victims still seem to be hostages gassed in a managed massacre (primarily not with sarin).

- etc.
...

Side-notes:

A contact recently suggested "I now think the story about smoke from an airstrike on chemical stores was a deliberate feed of disinformation via a source trusted by investigative journalists like Hersh and Porter, helping to divert people from asking the right questions." I note it can't be true and had that effect, so agreed. In fact, it helps cement the idea that a jet was over town and attacking stuff. That's an important point, supported by such false stories being disseminated, and also by this seeded image of dubious reliability.

Where was the sarin at Shayrat? From the cited April 7 DoD article:  "The two senior military officials who spoke to reporters at the briefing said the Shayrat Airfield in Homs governate in Syria that was targeted by the U.S. strike has a "history" of having chemical weapons stored there. ... The U.S. was not tracking the airfield as an active chemical site, an official said, but the military did take precautions not to hit anything that could possibly lead to the inadvertent release of chemical munitions or chemical substances." 

Does this suggest they already knew where the CW were and weren't? Like - not on the base at all as far as they actually knew? Investigators first refused to visit Shayrat to look into these sarin claims, saying it would be hard to know where to look to prove or disprove the claims. Later they did go, but found no sign of sarin, not that they knew for sure where to look - or to not bomb... so ... ?