4-4-17 Wind Direction Explainer

Idlib CW Massacre 4-4-17:
The When and Where
Wind Direction Explainer
May 14-16, 2017
last edits June 17, July 8, 2017
March 11,  Dec. 1, 2018 

There are many serious problems with the opposition story of an April 4 sarin airstrike in Khan Sheikhoun, Idlib province. While many of these are quite clear and worth pursuing, perhaps the best chance of smoking gun proof is the wind direction problem, as first explained at The When and Where and as I show it again below.

According to the alleged survivors and activist/terrorist sources, the sarin was released from one particular and well-known spot in the town's northeast, and the people affected and killed lived, mainly or largely, to the southwest of that spot. Naturally, released sarin would spread on the prevailing wind, so the opposition/terrorist story line requires a wind blowing to the southwest.

However, in all presented videos of the alleged airstrike, the prevailing wind was clearly to the northeast. So their presented evidence has a serious fatal flaw - all these stories are connected to an alleged wind direction that's not just different from but opposite to the visually evident one.

<add 5-16> The direction has to be changed slightly, a few degrees further to the north. Revised range: 30-45 degrees.  The "opposite" part is slightly less clear now, but otherwise...<end 5-16>
<add June 1>Belated graphic update - this is the revised map, replacing the original, which showed a narrower core range from 40-50 degrees. As noted below, an incorrect filed of view placement caused that error. </end>

All those presented stories they have in the purple area must be false (or at least some part of that story, as it stands, must be wrong). To me, this seems like the argument to lead with in countering the terrorist narrative. Clearly this matters enough to explain in some detail. That's emphasized by the resistance I've gotten from fellow opposition-story doubters, who try to claim there's no wind, etc. It's become clear to me most people don't understand how to read things like this, even with tips. So for this case, and for future reference, here's how you say what direction the wind is moving in a post-attack blast-plume video, as I did it with input from others, in the Khan Sheikhoun case.

Reference:
Tips: (to Orbi on Twitter):
* Look for left-right changes, for any sign of movement towards or away (harder to find and be sure of), know basic FOV,* estimate.
* field of view
* Find other views if possible. When we know it's moving, but appears not to from another angle, that's invisible movement, towards or away.
* Crunch all the considerations, like a 3-D puzzle in 2-D, which it is, draw a few lines, measure, consider, check confidence, and set range.
* Precision math is ideal, but I usually skip it. Visual-based accuracy gets you in the right sector at least, and not randomly confused. End.

Main videos:
* North camera, 3 plumes: https://www.youtube.com/watch?v=MYOMEDK_uVs
* west camera, plumes 1 and 2: https://www.nytimes.com/video/world/middleeast/100000005063944/syria-chemical-attack-russia.html
* west camera, plume 3: https://www.youtube.com/watch?v=yTx0kmQnZLw
* North camera, fog: https://www.youtube.com/watch?v=DWvDisOxJi0
* CNN, N cam, 3 plumes and fog scene, better resolution, at start of horrible video report.
http://edition.cnn.com/2017/05/09/middleeast/syria-chemical-attack-ward/index.html

The Scene: for reference throughout, the big map of most things ... mapped. The three camera angles shown (roughly) in light blue are especially helpful to refer to in understanding the material below.Note: this is just a basic idea graphic - it's fairly accurate but not precise, so don't cite the exact angles shown here. Better notes are included in the text below.

Same-day note: there's a strong tendency to assume all relevant video must be irrelevant, and actually from various wrong days. That's possible, but for many intensive tries, I've seen some patently bogus arguments and no good ones to prove that contention. Usually, wrong-day videos get exposed and proven within a few days or never, and it;s been over a month.

But luckily for us debunkers, that's not the only debunk available. Even if terrorists faked all their videos on the right day, they're still fake and that can still likely be shown. And even if it's just "according to the opposition's evidence," that clearly matters enough to consider, being so widely accepted and triggering US military strikes, etc.. So far it seems the wrong-dayers are scraping for reasons to not seriously consider it, leaving themselves completely empty-handed, and often with poor understanding and confused logic when it comes to trying to analyze these videos like they deserve. So... I'm doing it largely by myself, but taking scattered input from others.

And here's what we can say about the wind direction in the video presented as showing this alleged sarin attack. Here's the fatal internal contradiction in that evidence package, even if we *accept* or have it proven that it's relevant to April 4 after all...  

The Crater

Note in the map above the dark blue marked crater is not in the northern camera's field of view. Since that video never pans far enough left, we never see if there's a blast plume or anything at that spot at the time of attack. Most likely then, there was no dramatic blast plume, or they'd show it (and it would destroy the alleged sarin, and leave a bigger crater than we see).

Perhaps something else, like the alleged non-explosive sarin release, happened here. But if so, on the northeast edge of town, the wind would blow it further off frame and away from all the alleged victims. It provides no supporting clues about the wind, but all the other points available prove what it would show too. So this allegedly central spot is irrelevant to all the visual wind-based evidence considered, and will be almost totally absent from the analysis, except when it comes time for comparison between the opposition story and what we see.

Plumes 1 and 2, North View

This is the video of the two northern blast plumes, filmed from a distance to the north , facing mostly south, or about 21 degrees east of that. The following comparison of frames is from the better resolution CNN video.

Lined-up like this, it's clear all features of these smoke clouds angle to the left as they rise. The effect is quite minor at the bottom (already some distance above the actual ground), but takes hold around then, so each step up is a half-step to the left. This slant shows a mild but steady wind applied all the way up. There's clearly a higher speed at the upper levels, more to the left, turning the slant into a curve.

The same wind will prevail as it can everywhere, even at the cluttered surface. See too the low white cloud at right that seems to be expanding in all directions, for some reason, but on the left side we see what looks more like thinning drift further to the left. Being so low, it's a mild effect.

That all says this wind is partly at least to the (approximate) east. But this is just one component in the 2-dimensinal mapping. That's the clearer movement you see on the left-right axis, not the less obvious changes along the towards-or-away-from-the-camera axis. There may be a huge amount of relatively unseen movement, defining its actual direction and adding to its actual speed; that's a movement along both axes in real space, NOT left-to-right on your screen. But still, it can be gauged some from there, especially if you get more views... 

This is harder to show, but the smoke in this view seems to be moving slightly to moderately towards the camera, as well as clearly to the left. I'll leave it there for now, as other points below show that was a correct impression.

Field of View notes: relative to the minaret right of center above, the camera is facing app. 21 degrees east of due south, So left = 69 degrees on the compass, not due east, and towards the camera would be about 339 on the compass, not 360/0. The true movement of the wind is some mix of these two directions

Plume 3
More distant blast plume #3 is seen in the northern view, but not in much detail. A nearer camera somewhere in the west or northwest of the city gets some very useful views. The line of view and filming spot aren't certain, but about as mapped at top. This is facing south, or S-SE to see plume 3 in the distance - and the unexplained black smoke cloud further to the southwest.

The shape of plume 3 is interesting. I think that Mushroom cap means it was quite hot and rose fast, and might be a fuel-air or thermobaric explosion (as with the above). Also note the bend of the heat column the particles are rising on - first almost vertical-seeming, but angled barely to the right, when it was a fireball of the projectile's detonation. (slightly right from this point of view means south - possibly fired from the north - but above it's seen from more like due north, and bends more to the right, suggesting a westward ban - so fired from the north and east).

At a certain height, that slant changes. That will be the wind. From there, it all drifts steadily to the left.

In close-up - detail on the churning base suggests movement towards the camera, but that's more likely from blast and fire dynamics. Higher up, its movement on the towards-away axis isn't clear. If the wind were to the northeast, it would move almost totally left from this view, with less movement towards or away.

Field of View notes: filming location and exact line of sight to plume 3 are not set, but seems very close to the angle above - so left = around 70 degrees E-NE, and the towards-away axis is likewise rotated about 20 degrees from north-south.

Plume 1 and 2, Side-View

Videos sent to the New York Times show plumes 1 and 2 from a nearly perpendicular view, from the west-southwest facing east-northeast. Some note here there seems to be almost no wind. It does appear almost motionless, but it's not - it's moving just as fast, and in just the same direction as it was in reality, and as partly seen in the other view.

There's a mild movement to the left visible, but that's too slight to be its true speed. So an unclear but major amount of its movement must be on the axis of towards or away from the camera. Consider with the north views a clear drift to the left, so here it must be x amount away from the camera. I thought there was movement towards the lens as seen from the north, and here we see a consistent, but slight drift to the left. That left aspect - and part of the forward aspect - both equate to "towards" a northern camera, confirming that impression.

Since so much of the movement is away from the camera here, the wind direction is close to its field of view, and seems the best measure to determine the proportion of north and east movements into a defined direction. So I tried to set that.

Note 5/15: But I did it wrong (it happens sometimes), and it slightly affects the map above ...

<Add 5/16>I'll give the whole story, for lessons learned. I first set up proportions of view between minaret, plume 1, and plume 2 (in left-to-right order), using Q's located damage associated with attack, then found an angle that matched those distances they marked, but forgot to remember WHAT each mark meant, and traced a view that would show plume 1-minaret-plume 2. That mistake gave me a heading of 56 degrees. Then I did it more "carefully" but just as wrong to get 57 (bottom image here - marks on gray bar, yellow circle to mimic rotation I can't do in MS paint). 

I noticed the basic error when I tried to correct Michael Kobs, who showed a line of sight from further to the southwest, facing 49 degrees (map on Twitter - camera 3). That may not be exact, but at least he drew his lines in the right order (see top - purple points to minaret, plume 1, plume 2). 

Re-considering now, I did a good proportional set-up or looking again, set the plume originss too far south in the view, affecting lineup, so minaret and plume 1 would roughly line up. At least as I copied them here in purple lines (top image) Kobs' lines for minaret and plume 1 are even further apart. So 49 seems a good angle, but perhaps 50 or 51 is better.  My original 56-57 is out.

Still, as I found before, the wind is blowing mostly away from this camera, so mostly northeast, just 7 or so degrees more to the north than I first thought. The slight left drift relative to that line means the compass heading is some degrees lower/more northerly. By how much is hard to say, but probably 10-15 degrees or so. In that way I earlier estimated 40-50 degrees. As I refine it north, I'll take the chance to loosed it a bit too. Now, I estimate the wind movement we see is in a simplified range of 30-45 degrees, northeast to NNE. <end 5/16>

Field of View notes: app. 50 degrees view would mean away from the camera has the same compass heading, obviously, while left = 320 degrees NW.

Plume bend and northern-firing note: Also note here the same kind of bend between the lower and upper parts of the smoke as seen with plume 3, but even clearer. This is an initial slant to the right, so to the south. In fact the view is rotated a bit (the minaret isn't straight up) so that the south direction is actually more pronounced when looking at it level (it appears slightly more vertical here than it really is).

<add 6-17>There are some who see this lower slant to the right and decide this must be the apparent wind direction. This was put best by Bubslug in this ACLOS discussion where I held my view and came up with this sample graphic to explain why we must be looking at two forces to explain what we see - which is actually a curve, not a simple slope. The force from the right that seems to increase at altitude must be wind. The force from the left that prevails near ground could be a drastically different wind or, much more likely, the trail blazed by the ignition fireball, due to blast dynamics and perhaps the munition's trajectory.
<end 6-17>

This slant is perhaps because the rockets responsible were fired from the north, as we saw with plume 3. This bend isn't visible in the north view of the smoke columns because it's roughly along that line of sight. That could mean the rockets were fired form about where this cameraman captures the aftermath, putting the bend in an invisible direction (away from the lens). And of course he would know that, if they had flown right over his head.

We don't discuss the famous crater here, as it has no attack time visuals to consider wind by. But the pavement damage suggests it may have been hit by some rocket that was fired from a bit west of north. This is pretty consistent, and interesting especially if the Syrian jets only passed south of the city, as US radar tracks seem to show. 

This might be me just trying too hard, but fact is, all these possible signs point the same way, and that pops out for me as potentially quite interesting. Maybe I imagined it, but maybe not, and someone else can see it clearer and make a real case out of it.

<add 3-11-18>Important change of thinking note: I was being dense here, fearing any right movement meant SW, etc. All that angled impact chimney effect stuff is interesting and might play in, but this might be the wind after all. As Michael Kobs noted here, for example. the wind in this NE view seems to push differently at upper vs. lower levels - the shape of this dispersing smoke suggests it, and actual movment can be measured between frames in the 4-second clip. So the same wind mainly or totally explains that lower slope. The wind difference here could maybe be sparked by that heat, but anyway, it continues for some time.

But this doesn't mean, as Kobs first said (in the cited tweet), that this lower wind is to the south nor, as he's put it later, to the southeast. It only means it's somewhat to the right of this camera's line of sight, besides somewhat towards or away from the camera. And that baseline right-ahead direction we're starting from is northeast. Using Kobs' big and useful lines of sight map, with some adjustments I think valid, marking features as I can (easily), I can show the full view of smoke seen is all to the northeast, mostly even of the original smoke plumes. Both upper and lower smoke extend off frame by this time, but probably not by much. Most likely, this is a wind in the range of east to northeast at the lower level.

<end 3-11>

Plume Diffusion and Wind Change

The plumes should eventually drift away on the prevailing wind, with the tops going first and getting fuzzier, and finally as the outpouring from the blast site peters off, the last columns will fade and slant away.

This isn't shown much, but among the western views sent to the NYT  is this, with the same basic view, of plume 1 and/or 2 mostly gone, nearing the last moments of visibility. Here we see a heavy slant to the left, so the wind is more to the north than before. (note: rotation of view, as noted above, may apply here and affect the true slant, by a little bit).
<3-11-18>that's not a slant, but the difference/middle of the lower smoke drifting right and the upper smoke drifting left, as explained above. And it was all drifting away, apparently, so primarily northeast. Here's a new graphic:

<end 3-11>

The breeze may have sped up in both directions, or just shifted more of its speed onto the north axis. That's not clear. But higher speed would have those plumes drift off-frame faster than you would expect. Also, far below that, we see some kind of dingy white smoke or mist seeming to spread pretty far to the left/north, don't we? Noting this view does look over the area of the famous crater, it's possible we're seeing mist related to that spot, as well as to another we'll discuss next as we consider this stuff.

The White Fog Shows the Wind
This is an area of some difficulty for me. I can't help but connect the attack-time "white cloud" of low vapor to the later-seen "winter fog" blanketing the whole town, but so far no one agrees with me, arguing it's a natural morning mist, or simply trapped smoke from the attack (and I agree it's partly the latter, and maybe a bit of mist too, but... we can skip all that here).

This is the scene at attack time (top, 'white cloud' visible to the left of the tel, or flat hill). Below is the same view after the big plumes blew away and the fog had spread. Higher elevation sunlight says this scene is later, but not my much - it was a rapid change, in perhaps 15-25 minutes. It's hard to be more exact.

The way it's spread shows the same NE wind prevailed in that time. On the the left side (nearer, more northern part) direction is not not very clear - if we take the original 'white cloud' as its center, it's drifted a bit to the right, a lot to the left, and some amount to the north (towards the camera). If this were vapor pushed out by a machine or a very hearty chemical bomb, it might crowd the air with particles, pushing them to expand more than a drifting plume would, including against the wind.  Again, it's hard to show in a still image, but I can see movement in the fog here, and it's to the left and perhaps towards the camera.

But the southwest view always showed the wind more clearly, just by its shape. On a distant slope, we can see this area in more of a foreshorteded overview. The improved version from the CNN video, below, clarifies what I already saw there, and labeled here - a wind to the northeast spread the copious white mist emerging from a small, unknown area SW of Khan Sheikhoun. As it spreads to the left and towards the camera, the massive column is divided by a hill into a more eastern and a more northern stream.  (see topo map here, thanks to a comment - that must be the larger e-w ridge south of a smaller one at the south edge of town.)

The yellow lines somewhat mark an area of thinner vapor spread more to the north, perhaps due to the later north-shift in the wind noted above. If so, it seems the bulk of the flow we see has already spread in a pattern more like what my 40-50 measure would predict. Where things go from this is unseen and uncertain. I have a hunch the "fog machines" were turned off by the time of that video and all wee is the maximum size just drifting. Odd, in fact how little of the actual spread they show, Maybe they filmed it but decided it looked bad and so never released it?

Wind Speed:
I have no experience measuring the speed of winds ...I'd say it's not fast but not that slow either. Remember each visible movement is just part of its true speed along TWO axes, one mostly invisible. If you don't think you can gauge the towards-away direction, try doubling the left-right speed you see. That's likely closer to the true one.

A credible source estimates a 2 or a 3 on the Beaufort scale describes what we see (light breeze/gentle breeze - Wikipedia). That seems right to me. At Beaufort 1, 4-7 mph, leaves rustle, wind vanes move. Beaufort 2, 8-12 mph, leaves "in constant motion; light flags extended." The middle speeds (6-8 mph) seem a good range, for what's at ground level. The upper level winds are clearly faster, but in the same direction, and not as relevant to fog spread. (even at Beaufort 1, "Direction shown by smoke drift but not by wind vane.")

Other Spread Factors:

Topography, fog density, and obstacles (houses, walls, etc.) will all provide their own directives to any spreading fog. They can't ignore and override the wind, but would 'negotiate' with it, and cause some general spreading in non-wind directions, and into lower elevations when different ones are available. So the spread pattern will never be just a straight narrow band right along the wind. 

But most would say, and many have said about this attack, and it's mostly true - the wind is the main difference in who lives and dies. That includes a wind that never appears, dooming anyone who lives near the site in all directions, and especially anyone downhill. As the CBC reported, after hearing from Alaa al-Yousef, he and his family "were lucky, the wind went in the other direction, Alyousef said." 

Weather Predictions 

(June 11) Added Dr. Postol found a wind to the northwest, rather than northeast, by using an online prediction, I'm not sure from what source. Charles Wood advises me (here and here) that such predictions aren't to be trusted; "Fake news meets FAKE WEATHER! ... Only METARS are actual observations. Waste of time to discuss." Unfortunately, METARS reading aren't available for anywhere closer than Latakia and Damascus. 

I agree predictions shouldn't be read as gospel (like Postol does, declaring the attack video must be from another day, since we can be sure the prediction it correct). However, it's possible there's some value in them anyway. Say there was a real weather pattern they identified, but screwed up on just when it would appear. In the video, we see a wind to the north-northwest, later shifting more to the north. In proper terms (origin, or blowing from), that would be S-SW shifting to more south. 

Interesting then that whatever Postol cited, it shows that same pattern on April 4, but at the very start of it, in the midnight to 3 am time frame - S-SW, then south. It's possible this is the same wind, just predicted to hit the area about 4-5 hours earlier than it did. Clear conditions and cold temperature seem to match the video. But the speed is predicted a bit lower - 4 mph and then 3, when observation suggests it was about 6 to maybe 7 or 8 mph. Still, this could be the pattern we see guessed from afar. And it could be a fluke, a stopped-clock-almost-right effect, within a totally useless record.

Summary of Wind Findings: 

* All visual signs says the day's breeze, at the time of attack and much of the fog spread afterwards, was to the north and east. The best measure is the view around 50 degrees, from which the wind was first blowing mostly away (line of sight almost = wind direction). So at a bit less than 50 degrees, the best simple, rounded range is 30-45 degrees. That most likely contains the true, initial direction.

* An initial wind speed of 6-8 mph at ground level seems reasonable. At upper levels, the speed is considerably higher, but in the same basic direction and not very relevant.

* The later fog video (interval unclear - probably 15-25 minutes) shows the same basic direction prevailed during the minutes of fog spread.

* By then, however, it seems the wind had shifted to a more northerly direction- the effect of that on the overall spread is unclear, but likely minimal.

* These findings are the best yet, being based on a direct visual recording of the day's winds, not on any kind of prediction, guess, or erred reading, and using several placed views, all carefully considered together. The broad range (lime green lines) is more than certain to contain the initial breeze, and most likely includes the most north-blowing variation as well.  Any relevant winds outside that range won't be far off. This can be used, with reasonable confidence, to compare other evidence in the opposition's presented case. This is just what I've already done, finding the mapped "downwind" homes where people are said to be affected and to die, were placed just about exactly in the opposite direction for the way they should be.

So ... the Opposition Narrative Got it Backwards

Back to the image we started with comparing green and purple areas - knowing the wind is undeniably to the northeast - the other half of that picture is what opposition acticvists and alleged survivor-witnesses claim to be the affected area, at least via the homes they've shown that we can place on the map.

Here, I'll try to clarify that reading, or see if it's already "too clear" and needs adjustments. The wind direction is never spelled out per se, that I've seen (as in "the wind drifted the sarin cloud to the southwest...") But this is just what they would say, it seems so far.

Alaa al-Yousef: "the strike was really close to my house" (HRW) He also said the blast woke them up, being "only a few hundred metres away." (CBC) That's not close enough to matter, nor specified which way except to say, "They were lucky, the wind went in the other direction, Alyousef said." So they didn't live to the northeast. "The" other way could be any other way, not necessarily the opposite way. (Alaa also says 'There was no bad smell," agreeing with his famous cousin he runs media relations for, and disagreeing with most other witnesses)

Mohamed Nejdat al-Yousef ran into toxic fog that then "was blown downwind toward his farm outside the village," badly affecting his family. (NYT) So by the accepted video record, carefully read, his farm is northeast of town. Could be, since that's farmland, but it's not specified. Furthermore, it's possible some witnesses - maybe him - are telling it like it is, while others (likely a majority of those put forward) are working on a false script.

Ahmad al-Helou "saw the plane drop a bomb and the bomb falling until it hit the ground. The bomb fell in front of the bakery," probably meaning the place south of the grain silos, so the famous crater. It didn't blow up, but "he saw the bomb kick up yellowish smoke that spread in the prevailing wind." (HRW report) So, as the video proves, he should mean to the northeast, but it's not specified.

An Al Jazeera video report (here via Twitter) shows a map, apparently, of where people died, in three adjacent areas all west of the famous crater (This is shown here with my points added). The big label says "While the eastern areas were the least densely populated." (credit: Saleh) So they don't say it's the wind, like it might've spread everywhere, and this is just where the people were, or maybe it was magnetically attracted to the more populated areas, or the wind blew that way. Their reasoning is unclear.

This map adds whole broad areas that make no sense, but apparently they mean some victims in the northern blocks died, probably on the southern edge. Just by geolocation, from video where that's possible, we have six locations, one in the southern NE block, and the rest in the southern block. So perhaps this image we've formed is too clear in its SW orientation. But a spread much wider makes little sense anyway.  So while they seem to say just west, al-Jazeera's people would probably agree with the prevailing picture of a wind to the southwest, which is roughly opposite of correct.

In more detail: The unnamed family said to die in a sparsely-furnished basement apartment furthest north, could almost be gassed no matter the wind, from being so close to impact. Or maybe not almost... anyway, they probably couldn't be, if the wind was blowing away. Everyone else was even further upwind. 'Abu Rabeaa' has the home furthest south. It should be in the plume as well, but he says he was hit while out in the neighborhood. Abdelhamid al-Yousef, the most famous witness, has the home furthest from the crater, furthest upwind, and unlikely to be effected by the WC - Amira Saleh's children (next furthest site) were said to be taken to Yousef's home before they died there with his family (ACLOS). So that site really matters. Where did that alleged sarin blow in from?

With a northern shift, wind could blow from the "white cloud" spot directly over the easternmost of these homes - but maybe only after most of it already spread a different way - and it would spread wider than directly over on the wind, as it hit and wrapped around buildings, etc. and especially if the fog was dense, as if pressurized (particle crowding), as it seems it was to start with (see the white cloud). I may try a graphic for it, but for now I'll just note either way and from whatever spot, the white fog covers probably this whole area to differing degrees. So these homes being filled with that white mist could make sense. Just what that was remains unclear - from the massive volume, at least we can say it's unlikely to be pure sarin.

But as presented, with this one NE crater specified as the release point, it falls apart quite nicely. It's not just wrong, but 180 degrees opposite of true. And so, as I've already noted, it seems quite likely the opposition/terrorist planners of this operation got it exactly backwards; an accurate measure of the wind was verbally reported as a southwest wind (the proper terminology, meaning FROM the southwest), and mapped their fake story based on a northeast wind (blowing TO the southwest). Because somehow, elaborate victim stories pop up like well-fertilized mushrooms along a strip exactly UPWIND from the highlighted crater.

This would be the same error we were just familiarized with in Dr. Ted Postol's work on this very case. First, he found an inaccurate prediction of a wind to the NW at attack time, and mapped it as to the southeast (then later corrected it to the second wrong direction he's used as the basis for important arguments). Here, the opposition's dimwitted activists may have pulled a collective Postol, or the guy with too much centralized authority made the error everyone else just didn't know to catch.

It's quite possible this is some other kind of error that coincidentally comes out opposite of true, or that it's even a deliberate mistake, as a sort of trap (it occurs to me, but I'm "too paranoid" like that). But I like this basic error explanation. It is an uncanny fit and, as we just learned, an easy enough mistake to make.

A ridiculous allegation like this, with obvious and sloppy lies at its core, was the basis for president Trump's much-hailed humanitarian airstrikes. These killed and wounded some civilians and some soldiers/airmen, and destroyed some homes and some materiel used in Syria's fight against the false-flagging Islamist terrorists the US and its allies have enabled now for years. The strikes and the possibility of more favor the same kidnapping, child-killing, dunder-headed salafist mercenaries, who don't even notice when they've got their fancy wind-based evidence all set up backwards. And make no mistake, the people concocting false cover stories are the ones truly responsible for massacring a reported 100+ civilians, including "beautiful babies."

Apparent Efforts to Address this Problem

<Add 5-16>Bellingcat member Timmi Allen on Twitter has a nicely stabilized video showing the steady leftward movement of everything, except for the 'white cloud,' which he catches expanding mostly to the right for a few seconds (usually it expands in all directions). This was presented as "Different wind directions in #KhanSheikhoun chemical attack footage on ground and in height. @bellingcat @DanKaszeta." But as I and others noted there, it also seems to change directions ON the ground, at the minaret, or anyway right around this swelling cloud of vapor. The wind only blows on it, for a few seconds, while touching nothing else, as that upper wind strangely descends and blows most things, including, as we can see above on this page, this same white cloud, smearing its left side into that damn, problematic upper level wind. Right there at ground level. Weird stuff, this reality seeping back into the void created by another terrorist lie in Syria.

Why discuss wind direction now? It was never an issue for the first weeks. Is it now, that I'm raising it over the last week or so? Is someone of more influence raising the issue as well? Or is it just a coincidence Allen happened to notice a possible wind to the right ... where there isn't even one ... and decided to publish that as important just now?<end 5/16>

<Add July 8>The OPCW's no-wind theory, by which a lack of any sufficient breeze had the sarin roll downhill, gets its own post, citing Allen and several others, from both sides of the blame debate, in disagreement.<end July 8>

<add Dec. 1 2018>Wind Direction Challenge <end add>

Idlib Chemical Massacre: The When and Where

Idlib CW Massacre 4-4-17:
The When and Where
May 6, 2017
last edits June 11

Besides an obvious lack of rational motive, .there's no proof Syria did this and much contrary evidence, about the alleged sarin and its relevance, the possibility the victims were hostages killed by their Islamist captors, prior precedent for false-flag chemical attacks, etc. While this is my overall line of argument (see first post), in this post I'm focused on more narrow technical questions people from all sides of the debate may find of use: Video analysis, geolocation, reading solar angles for a visual-based time for key images, and establishing the most relevant estimate yet of the wind direction - this lets us place things in time and on the map, and expand what we know, narrowing focus for further research.

Flaws in the Tracks
In my opinion, most circulated arguments against government guilt are deeply flawed, easing the job of those sent to debunk such claims and reassure the public that what the "activists said" is true after all.

For example, I've felt sarin secondary exposure dangers are overrated. I hear it tends to evaporate within a few hours, washes off with hosing, doesn't easily re-float off people - it could be most of the reckless behavior supposedly disproving sarin might make sense after all. There still may be none, but it's not so easily proven by pointing to rescue workers with no gloves handling hosed-off people at the White Helmets cave hospital. Seeing unprotected people carrying the victims out of their gas-filled homes with no problems would be a different story. But they haven't shown any of these first-recovery scenes, which is interesting (see way below, near the end).

Many point to a "Russian explanation," at least as read and reported, where a Syrian airstrike hit an unknown terrorist CW cache and accidentally caused the mass gassing incident. But Syria denies any strikes in the area prior to about 11:30 am, and the attack in question happened around 7 am by all available evidence (exact time is allegedly 6:46). The 11:30 attack is irrelevant: either Syria lied about the morning raid, as that theory and other theories presume, or the opposition lied.

My experience has been it's worth giving Syria the benefit of the doubt, and they have zero motive or less to launch a random deadly sarin attack against strictly civilian targets at this or any time. And while they might deny a conventional strike at 6:46 that accidentally released chemicals, I kind of doubt that's what happened. Some of the objections raised are valid - like that a conventional blast would tend to destroy any sarin nearby, not spread it in a deadly mist (it behaves like water, and needs a low-heat aerosol dispersal to do much damage).

And finally, more to the point of this post, many have excitedly pointed to the work of MIT professor Ted Postol, who has written a string of technical critiques of this incident. Along with fellow MIT scientist Richard Lloyd, Postol is renowned for helping disprove claims regarding government guilt for the 2013 Ghouta chemical attack. That work seemingly stands and is credited as such. But here Postol makes unfortunate errors, at least in establishing the wind direction, on which he based many of his arguments. I'll explain this below as I relate the much different best reading we have to offer. In the end, knee-jerk Assad-blamer Clay Claiborne was able to raise some (seemingly) good points in slamming Postol's work, while George Monibot posed hard but fails to do the same.

Finally, on April 26, Dr. Postol revived the perhaps-discared "Russian explanation" as possibly true. In all this, sadly, he has sown confusion, which I hope to counter here. 

The work here, if not that "attack" on Postol, was done only partly by me, and largely by and in consultation with ACLOS (A Closer Look On Syria) members Petri Krohn, Charles Wood, Pmr9, Q, and others. For future reference, to separate the known physical-visual facts from the unknowns, my best understanding from our collective work so far. This isn't a perfect summation, but I took extra care to iron out these details the best I could in a semi-timely manner.

When Was The Attack?

The White House explanation for its strikes on Syria cites radar tracks of two Syrian SU-22 jets overhead at 6:37 and 6:46 at the latest (graphic at right). Some early reports saidthe attack was at 6:30, while others have said 6:45, 6:55, and the like. Perhaps the reported first pass confused some distance witnesses, who took the later reports as stemming from then. Otherwise, these could line up well enough: blast timed with jet's pass at 6:46, reports of casualties and toxic gas 4-7 minutes later, group getting reports notes it 2-3 minutes later as "6:55." (for example).

The NYT video cites Facebook messages about an airstrike, if not chemicals, as early as 6:55 and 6:57. That would about nine minutes after the alleged strikes, which is a bit slower than one would expect. One of those cites 6:50 as a time, and an activist told them four bomb were dropped at 6:43.None of these seems exact, but not far from a decent center time that's just about when those jets were tracked passing.

Further, the post-attack video of smoke plumes (discussed next) seem to be filmed at around this time by sunlight angles (it's some minutes after sunrise at 6:16 as the sun was still low and from the east). In fact all detailed considerations suggest a time very close to 6:46, possibly late but not by much. The fist Twitter message I could find mentioning a chemical aspect was at 7:22, which is a bit slow. As explained below, we can see alleged sarin victims already at the cave hospital run by the White Helmets by 7:10 at the latest.  So there seem to be some possible delays in reporting, but no real discrepancies on the actual timeline of events.

What Was the Attack?

> Sub-Post: The Blasts
The evidence indicates at an early time around 6:45 am, there were at least three powerful explosions, consistent with conventional high explosives (or perhaps fuel-air explosives, etc. - damage bombs, not chemical spreading ones). Below, the three large plumes of smoke rise above the city, as seen from the north (in a panoramic view from this video).

As that starts,a tearing sound like a fighter jet is heard, trailing off as if leaving the area. That's not certain and could be edited in, but it might line up with what would be valid radar evidence. All considered, I tentatively conclude that's the real 6:46 jet, and this video is filmed about 6:47 and forward. That could be wrong, but it's worth floating at least. But we do not have proof that any passing jet actually caused these plumes.

Add 5-7, Side-note: is this all or partly a different day? As I'll explain, these scenes are timed to about 6:46 am, with a jet heard (FWIW), with recent damage in spots plumes rise from, and a strange white fog as described - it's a perfect set-up for the day's allegations, all set up ... when, other than the day it was clearly tailored for?<end add 5-7>

The other possibility to consider is a false-flag incident, and as the jets passed would be the best time to do that (previously explained at the Shayrat airfield link). If they had the plan and the victims and the story ready in advance, they could get early warning of incoming jets via the opposition's "observatory" system, to make the final preparations to launch their own rockets, release their own gas cloud, or whatever at just the right time. It would take powerful rockets, but rebels have these, and they could be aimed to land anywhere, roughly, to create any pattern, on demand. Maybe that's why no videos show the jet's missiles coming in - because they sounded like incoming rockets, that were maybe too big even to be from a jet.

So if we accept - at least on a trial basis - Syria's stance that they dropped nothing, I propose this version where much else of what's cited against them is true.
* One or both of those jets passed near the city at 6:46
* There were no airstrikes; the jets were probably on a recon mission in advance of the planned strikes at noon
* Someone on the ground with a nasty plan made some explosions happen just then, to provide the explanation for a lot of dead people, including an especially big number of children ...
* Perhaps a toxic gas was also released by those same plotters in the open. That's not necessary when, for all we can see and know, these victims could have all been hostages gassed in some basement gas chamber overnight. But some visual release, perhaps filmed on video, would obviously be helpful to their story (see below).

So ... It isn't completely unreasonable to conclude those jets launched the attack. In fact, it would seem obvious to someone who's accepted opposition claims in the past. No feat of depravity or senseless stupidity would seem past "Assad." But here, we're leaving this central question properly open. There's no proof this was a jet attack, and ample reason to seriously consider the alternate explanation.

When was the Hospital Attack?
Some reports and officials insist there was a follow-up strike on the White Helmets cave hospital where patients were "treated" (hosed off in the mud), perhaps to kill the survivors or cover up evidence. Russian drones were said to survey the area before, implicating Russia in a crime that Israeli Defense Minister Avigdor Lieberman described so: "The cruelest thing...is that those who were evacuated from the area of the bombing to the hospital were attacked a second time with chemical weapons as Syrian army planes bombed the hospital." Both attacks were "by direct and premeditated order of Syrian President Bashar Assad, with Syrian planes. I say this with 100 percent certainty." He sounds well-informed. He also rejects direct Israeli punishment, saying "I'm not ready for Israel to be the dick that the whole world pisses through." He urged "the international community" to take direct action over this clear double-tap gas attack. (YNetNews)

However reports and visuals are clear the hospital attack was with conventional weapons, with no gas reports suggested anywhere I've seen. There was no damage to the hospital at 7, 8, 9, 11 am, while fresh damage to the larger external building appears around noon. A video of the attack is filmed around noon (inside, there's a blast, they run out and see smoke and noontime sun). Most patients had been gone for hours by then, and few if any would remain quite this late. One dead girl has a reunion with her grieving father shortly before, but the place is mostly deserted. Petri Krohn times this (maybe roughly) at 11:55, and the attack at 12:30.

There are some serious questions about the videos of this event, and it's not certain it was even attacked then, or if some planned fake attack was again tacked onto a real Syrian air raid. This time the Syrians acknowledge attacking something, but not that it was the hospital. But if so, they probably thought it doubled as a militant base of the Al-Qaeda affiliate al-Nusra Front, and it quite likely did. I believe Syria says their first attack was around 11:30, not 12:30, so this wouldn't be the first strike, anyway.

But this is all a side-issue. Let's get back to the events of the real and singular alleged chemical attack at 6:46 am.

Mapping the Attack
The famous crater where everyone says the sarin was released is well-placed on the map, but the rest of the attack not so much. Again, refer as needed to the 3 plumes video (with the panorama view above, inset small below, and then shown again below). For consideration, here's what it shows: Left to right: (off-frame, unseen) the space above the famous crater - smoke plume from a high explosives blast southwest of that (plume #1) - another the same way (plume #2) - a strange white cloud, apparently originating just northeast of the tel (big flat hill) - the tel, continuing, with a blurry minaret in the middle - another standard explosives blast from a bomb or rocket, roughly geolocated to the southwest of the tel (plume #3) - (not visible here) a smaller black plume rising from a located spot in the town's southwest.

The opposition-supporting Bellingcat investigative group did a good line-of-sight analysis for this scene, which I copied (roughly) to make a rough map (ACLOS, just improved to the below) of where these 3 plumes were rising from (dark red circles, locations narrowed down at ACLOS with a lot of help from Q). This also shows where the white cloud might be, plus the black plume to the south, and other features indicated, including alleged places where people died from the gas (little red squares - see explanations at ACLOS here and here, again mostly by Q).

The famous crater is marked with a blue circle on the map above. At right: activist reporter Hadi Abdullah standing by it, around 1:00 pm on the day of the attack. That's  long enough I think most sarin would have evaporated, and it would be fairly safe to stand there - but you wouldn't want to shove your hand into and shaded soil under the surface).

There are some valid questions about when and how this crater was created, with its bent metal tube jammed in. Dr. Ted Postol proposes a sarin-filled pipe underneath a ground-based bomb. I suspect it was a rocket fired in; bent tubes like that are common - when the front of the rocket tube becomes buried, the remaining part bends forward with its remaining momentum. Postol will have seen this with the rockets linked to the Ghouta attack. Most clues of the damage and that bend suggest if it was fired, it came from the north, or a bit west of that. (note, the Syrian jets also circle around the city, including to the north, so it's possible they fired from there).

The crater munition does not look like an air-dropped bomb, as alleged by Human Rights Watch. Further, as I explained at the Shayrat airfield link, the U.S. provided flight-track of the jets shows they never got closer than about 2-3 km south of this crater, when they would have to be right above it to drop that bomb there. That mapping could be wrong, but it seems accurate, and raises problems for this allegation; if the debris is conclusively linked to the alleged soviet-made KhAB-250 sarin bomb, we'll need to know how it got there. I'll suggest the distinctive fill cap was planted, or perhaps an old KhAB was re-jiggered to be fired as a rocket.

In the distance views, the area above the crater is simply not shown. Bellingcat's analysis shows this, as does my panorama, both trying to show the furthest views each way. On the big map above, the light blue cut-off line is for a field of view that just never pans that far left. Perhaps nothing was happening there at this time, or it just wasn't obvious enough to see from a distance. Still, it's quite possibly something did hit here and released something here that tested for positive for the kind of impure sarin used in prior CW attack.

The other blasts were located with great work by Q, noting in various videos and other sources where damaged house were located, that, with a side-view of plumes 1 and 2 from the southwest, puts those homes in the cross-hairs of two perpendicular fields-of-view (NYT video). Hence the nice, accurate markers on the map for those plumes. Note in the side-view the plumes both "bend," with the lower part rising first on a rightward angle, likely from blast dynamics, while the actual wind takes over at a certain height, as the smoke cools. That initial angle is to the south (roughly), suggesting these munitions might have fired from a somewhat northerly direction.

The "white cloud" to the right of these  is simply along that line of sight in that zone near the tel, hence the big rectangle for possible area. But it may cast a shadow on the tel, allowing for a more exact placement (it would have to be almost due east of the tel, so in the southern end of the identified zone). This is one of a few things still not done.

Plume 3 is far to the south, but placed with less certainty than 1 and 2, and there's even less precision for the faint black smoke cloud further southwest. This seems to be attack related but different - perhaps a munition that didn't detonate, or something with just a weak starter charge; black smoke suggests gunpowder, I think, and not high explosives.

Most analyses, like Dr. Postol's, seem unaware of the black and white clouds, just considering the big and explosive blasts and/or that crater as possible origins for sarin gas/vapor - even though, as I understand it, such a blast would destroy the stuff. In that light, the nonviolently expanding  low white cloud everyone seems to ignore, between plume 2 and the tel, looks interesting.

The White Fog
> Sub-Post: white fog explainer
Below are two panoramic composite views from two videos associated with the attack, taken from north of the city. They both seem to share the same exact point of view (features line up the same in both, except the lower view was taken maybe on the roof). Therefore, we can make a direct reading between them, as I lined everything up - if this is all the same day.  I'm presuming so, but noting it might not be. And I have pretty good reason.
Already in the top view is that white cloud at the left shoulder of the tel. As a possible sarin source, it's been ignored, but it has the advantage of not being a hot, violent, explosion.  Below, from another video, we see something like a white fog, with no inherent color I can identify. What it is, we don't know. There's a hell of a lot of it, and it seems to be densest in, and maybe spreading from, at least these two locations:
* roughly the location of that white cloud, spreading north (towards us), maybe a bit to the right, and mainly off to the left (east).
* another spot southwest of town in the fields (far right distance), seeming to spread both east (left) and northeast from there, perhaps divided by some hill in between. (this is further southwest than anything mapped above)

It's not clear this fog is sarin, or whatever people described as doing the sarin-like killing. Witnesses describe a "yellow dust" with a "foul," or "really disgusting odor" that's like "rotten food" but also strange and hard to place. This was even smelled from far away. It caused burning eyes and chest tightness (lung damage), they say. A 14-year-old girl "described an explosion like a yellow mushroom cloud that stung her eyes," and something “like a winter fog.” (New York Times) A farmer mildly poisoned  also described “a winter fog — not quite yellow and not quite white.” that stung his eyes, and caused supposed sarin symptoms (also New York Times, different report). That might be the same thing we're seeing - "not-quite-yellow" close up, basically white at a distance, in sunlight's glare.

So this white fog probably is, or is supposed to be, the sarin people describe, also causing people to pass out and die, convulse, and foam at the mouth. Those aren't quite the right symptoms anyway, and really I suspect they had some white effects fog from giant fog machines, perhaps filled with a nasty chemical that was "sarin or a sarin-like substance." Seeing just two extra dense spots means maybe just two locales have coated this whole area, and are still pouring out more. Eventually, they would turn the machines off and "Assad's sarin attack" could end. But that's probably too copious to make much sense as pure sarin - it would be a huge amount of a hard-to-make substance, and there might be 2,000 dead, not 100, if they had done that.
* Perhaps it's a concoction including trace amounts of the their nasty variety of sarin, to tigger the tests
* a simple stand-in visual effect of their allegation, with any sarin worked into the evidence some other way
* or that plus something actually caustic and similar to the impure sarin they decided to go ahead and blame the Syrian government for, as they already did back in 2013 anyway.

While witness accounts are never to be taken as obvious truth, it's likely that a noxious gas like they describe was released in Khan Sheikhoun that day. It's not only plausible but seemingly supported by visual evidence. And it's entirely possible the same was released at the crate location, and it's just not visible on these videos.

they say it was corrosive, stinging the eyes and lungs - evidence of caustic exposure was found in lung tissue in autopsies done in Turkey, alongside positive sarin results - how that all adds up, not clear. They seem to be saying this "Assad's" other sarin ... "let's get it out in the open: yes, it's Assad using that nasty improvised "kitchen sarin," maybe to make us look bad..."(see the Sarin Evidence)

What was the Prevailing Wind Direction?
> Sub-Post: wind direction explainer
May 20 note: the following is now more fully explained and revised at the above link. The core point is the same, but the exact direction is estimated a bit differently now, as insert-noted below.
---
Dr. Theodore Postol - using a weather prediction of questionable value - first read the wind direction backwards, confusing to and from (April 11 report). Winds are given properly by their direction of origin, and the direction it pushes things to is opposite. To me, Dr. Postol, and others, it's counter-intuitive, but a SE wind blows to the NW. The top view below shows the erred reading for 6 am, with a wrong 3 am reading to the south as well. After a tip-off from one of my research associates, Postol then corrected that to show a breeze to the west-northwest at 6 and 7 am, using that to make an argument about expected death toll (bottom image,  April 21 at TruthDig).

Take 1, top: wind to the southeast at 6 am. Take 2, bottom: to the northwest at 6 and 7 am.

However, a better way to say what the wind would do at 6:46-7:00 might be ... what the wind was actually doing then. If they disagree, video of the real wind would trump that prediction. Such video evidence has been available the whole time, and it disagrees.

In all distance views from the north (three), over a short but very relevant span of time, smoke or gas clouds seem to be blowing to the left and towards the camera, so north and east, in a ratio that's hard to be sure of (see video, and just the shape of the plumes, slanting mainly left). This video filmed from the west part of town facing south shows plume 3 in some detail (and the black smoke, briefly, at the very start). Zooming in on its base, the apparent direction here is clear: it rolls towards the camra, and to the left - to the north and east.

Properly, this would be a southwest wind, and the direction it moves might be around due northeast (45 degrees on the compass), or closer to east, or north-northeast. The wind will shift a bit over time. But from what we see, it's very unlikely a wind to the w-nw as predicted for 6-7am ever prevailed enough to matte at the time of the attack. An almost opposite direction probably did.

As discussed here at ACLOS, there was a later find of a view from the west-southwest, with the first of three views already shown above. I estimate the line of sight of this as about 56 50 degrees on the compass. The most useful view, shown above, has plumes 1 and 2, drifting little that we can see. We know they were moving at a moderate speed, so most of the movement (aside from up, ignore that) is invisible here: That means it's moving towards or away from the camera (must be away, as the northern view's left movement clarifies) and a bit to the left. It's a very short clip, but it seems mostly to move away from the camera, and less to the north. So most likely it's blowing at an angle less than (about) 56°.

I'll estimate the wind was blowing to a most likely range of 40-50 30-45 degrees on the compass, due northeast (add: to NNE), with a wider range also feasible. It's not close to 0° due north, or fully east at 90°, but safely between them. A broad range from 20-70° is certain to include the dominant average. That's narrowed down to the correct sector of the compass at least. In the big graphic a little ways down the page, I mapped the direction range before this final decision, using a more easterly central range of 45-55 and a bottom of 30°. The difference is not big, and the graphic is quite useful 9will be updated).

First, I'll note Russian news site Rusvesna got it right in an analytical piece (apparently by an unnamed external expert in some areas), along with other interesting points, some of which I disagree with. (English, Russian original). Their graphic (right) shows a nice, cautious range centered a bit east of northeast. This might be just it, or it might be more northerly as I think. A

And we have to note the fog spread - that video seems to show the same - east and north drift has prevailed long enough to blanket this wide area. Further, the light seems more diffuse or overcast in that view, as the big smoke plumes have disappeared. Where did they go? On the wind, largely in the direction of the sun, clouding the light here (sun was almost due east, about 87, at the time). Noting how winds can change, this would suggests the same basic wind directions we saw earlier prevailed in between, and would be the clear direction to consider in evaluating the other evidence. 

Here ... early views are clear the wind is in the northern part of this range (45-55), but shifts after that could fall anywhere in (or near) the range marked in lime green, and momentarily perhaps further out of this range. Note: this isn't exactly "affected area." There will be additional dispersion, especially with that massive white fog we see, to fill space well left and right of the initial cloud as it drifts further out. If the direction is in that green core, the lime lines might mark a potentially affected area. 

May 10, new version:

So ... as the Rusvessna piece already showed and as I show again here - if anything was released at the crater everyone cites, it would probably blow over the grain silos and the fields north of them, dispersing where no one lives, and not towards the identified homes where people died from the stuff.

In fact, the wind blows almost exactly away from these homes. I measure a line between them as 225-229° on the compass, making that the best wind direction in purple above (range in lighter purple). This is directly opposite of 45-49° on the compass, the southern half of of my estimated central range (or northern half as shown, whatever). It's as if the people deciding where to say the victims died had someone send them the correct wind reading, and then read it backward like Dr. Postol did, and picked houses in the opposite direction of expected drift from the crater. That's so uncanny, it might be just what happened. While Postol's personal flub might be embarrassing, this would be a case of several stories and actors all working on a wrong script, with no one noticing (early enough anyway) to halt or change the elaborate set-up.

I'm not certain on deadly range: Postol shows a span of about 500 meters long as the expected "area of very high casualties." The Rusvessna analysis is based on a 55-meter circle of deadly effect, and a secondary range to 220 meters out on the wind where lesser effects are to be expected. In my graphic above, I mark with pink arcs the ranges of (approximately) 200 and 400  220 and 440 meters on the actual wind direction, and 200 and 300 meters in the (alleged?) direction. (I made the map too small to show 400, but this isn't the real direction anyway, so who cares?) Between them is a smaller core of 55 meters radius in red. Rusvessna's expert had it 55m in diameter, but I felt expansive here - anyone in that whole intersection or those first couple houses on the left might well die. This too would stretch a bit in the real wind direction, but still be more of a circle right around the site than the drawn-out plume running out to the pink lines and past.

Considering the wider scene, the reports of something dispersed over the this area, and the fog video that might show that, let's consider the other impacts. Maybe "Assad" had worked some unknown indestructible sarin into the conventional bombs of plumes 1, 2, and 3? Plume 2 came from a killed family's home (whether the blast or sarin is blamed, unclear), and another 2 homes clearly said to be gassed could be downwind of that, besides downwind of plume 1. But that's probably not possible, and 3 other located homes are upwind of both plumes and of that crater.

Further, it's not likely any of the located homes is downwind from the white cloud either. Thus, every story attached to gas poisoning and deaths in that area is questionable.

There's plume 3 to consider, the black smoke area, and the apparent fog origin point further to the southwest. All of these are possibly upwind of the stricken homes, but at a much grater distance. Any noxious gas they encountered from those locales would be quite dilute. It would not likely to re-pool in homes,  even in a basement,in the concentrations needed to kill dozens of people.

So ... there is evidence, if still debatable, for some kind of gas release that mattered, that had a white fog in perhaps the right parts of town to fit  that story. But for what it's worth, the wind says that didn't come from the famously fingered sarin release point, that crater at the edge of irrelevance with wind drifting even further off-frame. The crater with the red poison sign is a red herring. No one has identified the real chemical release point. But I may have some leads above.

When and Where did People Die?
Allegedly, the victims mainly died in their homes, still sleeping or just waking. Some alleged homes are shown on video, six of which are placeable in town, all in small area in the town's north (map above, the red squares).

But those homes are well upwind from the alleged sarin release point (though the noxious stuff might have come from somewhere else). No victims are shown at their homes, and they tend to appear fully dressed in street clothes, not any kind of sleeping garments. All this raises question marks over those stories, and leaves open that some victims we see are actors who dressed before work, and others are perhaps hostages, who usually don't get pajamas, who were gassed - somewhere, with something. 

First Appearances: We can't say just when the earliest view of the victims was. Petri Krohn found the earliest indentified video (dead children in a truck, the famous scene) was uploaded by Assi Press Center only at 9:33, nearly three hours after the alleged attack. That first-seen scene could be timed by sunlight, but not easily; it's got low light, probably no more than 20-25 degrees (?), so not long after 7 am, or 'about the right time.' There's an apparent delay there.

Around this same time, lots of dead people had appeared at various collection points, with the noted ambiguity as to how and from where. Any link between these bulk corpses and the passing jet is at least as unproven as with the attack plumes. In fact it's possible they were simply collected and gassed at one of these sites (most notably the cave hospital) and then dragged outside for "treatment" and display.

timed at about 7:20-7:25 am

The first time-able visual of victims is at the White Helmets cave hospital, in scenes I time visually by sunlight angles as starting around 7:09 or 7:10 am, through 7:20 and continuing to 8:30 and beyond. I only looked at the longest-shadow scenes - I'm not sure how long it runs, but from no earlier than 7:10 that we see.

Method: NOAA solar calculator (pin set to Khan Sheikhoun, time set to GMT +2, DST on, correct date (4 April, 2017) - enter times to see az/el (azimuth, compass direction to the sun, elevation of the sun above the horizon).

This seems fairly plausible - perhaps a bit swift to have victims gathered, but presenting no obvious timeline issues - people could be found and trucked there by 7:10 if the bombs fell at 6:46 as alleged. This doesn't mean the whole story is true, just tht it could be, or could be a reasonably well-crafted false narrative that lines up better than many of them do.

Later Movements: Dr. Shajul Islam in Binnish, an hour's drive north, was taking in patients at 9:43 by clocks there, with later scenes included at 2:00 pm and past, treating alleged victim nearly eight hours after the attack (Jesuraja on Twitter). This was an all-day event, it seems with new victims cycling through, for fresh photo ops for media coverage running in shifts. That's abnormal or, in rebel turf in Syria, fairly standard.

One time-able video of dead bodies in a truck (screen grab at ACLOS with notes) has a solar elevation around 36-38° = time range app. 9:20-9:30 am. The corresponding azimuth would make this a roughly E-W street, tending a bit NW-SE, so perhps (from other details) here on Wikimapia, almost next to a mosque.

A later video claims to show a scene in Maarat al-Numan, well to the north of the attack, with gas attack victims unloaded from an ambulance and rushed into some hospital. Sun elevation (via ambulance shadow) seemed around 65°. It never got that high, so I must be off, and it's solar noon. The NOAA solar calculator says solar noon was 12:36 pm, with 160.4 degrees elevation. So nearly 6 hours after the alleged attack, people are still rushing around with dead-ish people and trying hard to save them.

As for why ... we can only speculate. Shajul Islam, aknown accomplice to terrorist kidnappings, by the way, took some patients up in Binnish, while more were recieved by Dr. Tennari in Sarmin (not quite as far), besides Maarat al-Numan as seen, Khan Sheikhoun's main hospital, and the White Helmets cave hospital, at least, each had a hand in handling the alleged sarin victims. And so, many opposition activist "doctors" were brought in for second opinions and wider coverage of this incident. That might be part of the reasoning.

Other Points and Conclusion

In summary:
* There was a likely jet pass south of Khan Sheikhoun at 6:46.
* At the same time, at least 3 powerful conventional explosions across the city, likely by terrorist rockets, with some indications that some were fired from the north.
* A mysterious white cloud appears on the surface in the north of town at the same time, and a puff of black smoke appears far to the south.
* A massive amount of a possibly caustic and perhaps sarin-like vapor appears from at least two spots, one being where that cloud was, and settled over the town. This might also come from the crater area. It's not clear to what degree witness descriptions of this can be trusted, but it's made to sound highly toxic and deadly, with effects very much like a poor understanding of what sarin does.
*  The prevailing wind would blow this all northeast, so any "sarin or sarin-like substance" from the famous crater would dissipate over the fields, and actually blow almost exactly away from the located victims' homes. Anything that did permeate the alleged affected area must have come from another, unexplored, location to the southwest of there.
* Victims appear at roughly the right time, but it's not clear where they came from, and (as explained here) generally they don't appear to be suffering form real or serious sarin exposure. These were then heavily shown off and shuffled around over the day for unclear reasons.

What happened is Still uncertain, but there's still no proof for government guilt, and no motive for government, to ample incentives for opposition forces. I still feel my hunch of a terrorist massacre with a false-flag chemical aspect is quite strong. But then, I'm pretty sure I've seen just this same thing, over and over now. To me, it's not even surprising; this is the grim reality of the world today, and it just keeps stomping horribly forward through more and more of Syria's poor "liberated" people. - it's worth more careful consideration than it's received - by both sides - so far. I hope that these findings are of some help.

Add 5-7: Some people are stumbling over this, and I think a problem is the degree of planning and sophistication required for terrorists and their helpers to pull all this off so well. In my assessment after looking at hundreds of such cases over the years, this is an exceptionally ambitious and well-planned flase-flag event. Proof man Abdelhamid Al-Yousef seems like a leading actor for a leading story of great importance. If there were any case they want to plan out well, it would be Ghouta 2.0, which this apparently is.

In fact, I can see them deciding to say the victims they had listed and laundered lived here, here, here, all downwind of the well-chosen sarin release crater, asexperts could point out. But no experts have (that I've noticed). If there was such a plan, and why not, for Ghouta 2.0... everyone from the people firing the rocket at a chosen spot to the people practicing their lines about how their families died here, here and here... all of them must have been laboring on an erred wind reading, planning this part backwards, and setting the sarin up to blow AWAY from their victims ... Monty Python grade absurdity, that would be. It seems kind of likely to me, either way. If "Assad" chose to drop his sarin bomb here, hoping to kill, the whole chain of command read the wind backwards. Otherwise, it was the rebel propagandists who pulled a collective Postol here. (Or, maybe, that's a total coincidence, but that option is, at the very least, the boring one).<end add 5-7>