Khan Sheikhoun Wind Direction: Southwest Fog Spread

Idlib CW Massacre 4-4-17:
Wind Direction Explainer
Southwest Fog Spread
April 3, 2018
(rough, incomplete)

I've decided I might split-up my wind direction explainer post that's already grown unwieldy with notes and additions, even as the work still needs some more refinements.

Here we'll re-visit the mysterious fog coating Khan Sheikhoun during the alleged sarin attack of 4 April, 2017. It's more than any bomb would release, and comes from at least two areas no sarin or any bomb was reported to have landed. There's field of fog in the north of town apparently starting as a huge expanding cloud of white vapor that's barely been mentioned, let alone explained well. That will partly be dust and smoke from the two sizeable bombs that went off nearby, but I think it's mainly that expanding vapor, having expnaded for about 20 minutes in between videos.

And there's also this other expansive field of the stuff at the southwest corner of town we'll look at now. There's no sign or word of bombs going off this way, and yet smoke or fog appears in vast quantities here at the same time. Some researchers see in this a natural fog on some other day, but I think it's all related, not natural, and relevant. It's probably not toxic, but produced as a visual effect, using a couple large-scale smoke machines. I think it's on the alleged chemical attack day, and for the reason of helping to paint that picture. The picture isn't very realistic, for what it's worth, but I think it was painted. See my dated but still-relevant overview of the subject here: white fog explainer.. 

First, an enhanced view of this area, from Salloum's video filmed from north of town, facing south (from CNN video, citing a different guy though). The way the fog splits on the left is of great interest below. The gold/yellow box marks some taller, highly visible buildings. The red circle is my prior guess for fog release point .. or maybe not. That's my main point here today. As painted here, it's the spot the fog seems to emanate from, perhaps. But as it appears on overhead maps below, the red circle is the area suggested by oddities discovered at that site consistent with something odd happening there this day. As it happens, that spot must be revised or at least questioned.

Next, a new field of view study - green for bottom edge of the tel, or big flat hill at town's center (approximate), purple for a located minaret, the yellow box buildings. Three orange lines mark where the fog starts to split into an east and a northeast stream, and where it splits wider and wider yet. This will be a hill it was blown up against. The far edge of fog (thinner at the very end) is marked in white (likely fog origin just west of that. This marked image scaled down and rotated onto line of sight so the distances match. Extension lines drawn on the overhead view, up to where the fog seems to be. My old red circle is marked here, and a new estimate in green (explained below). In between, a Google maps topographic overlay (enhanced colors, partial transparency, carefully matched at the tel and all roads) (yellow box buildings are between the narrower yellow lines - the wider lines were from how it had seemed, but it seems to only include some of those buildings - unclear).

The topographic overlay helps set the fog spread in white and gray. It's about what I saw before - splitting at the nearer small hill (orange lines for increasing split). Note how well this matches the contours here (thin black line in original image).

This also shows how there's no topographic reason for the general triangle shape to form if this were natural fog, as some think. There's no valley of that shape or anything to cause a natural fog to take this shape. The splitting in the middle, yes, but not the outer limits. There would be more fog continuing all across the green and blue area if so, but it appears at the start of an upslope and spreads or moves only uphill from there. We can see why it stops going uphill, but not why it doesn't even exist downhill. So I propose it's what I see: a discrete point of origin and a spread shaped jointly by the ground-level wind and topography.

For reference, here's what I decided long ago without such nice layering options, and even before identifying the red-circled likely origin. Below, the gold box was inexact, too wide. But the new green circle makes this original version closer to the mark, maybe by accident, than I was with that red circle.

The wind direction here at ground level, where any sarin fog would be spreading, is east and north, with the proportion debatable. The hill distorts the shape, making part spread more north than the wind alone would do, and another large portion drift more east than it would otherwise. Note in the video view, off to the left and behind that hill, the fog seems to "pile up" higher or, in context, it extends deeper to the south. That could be a south trend to the wind if we didn't seem to have a north trend as well. Or, see the topography for that (south of the split is only a little lowland before it bumps into the same hard-to-crest yellow turf - a continuing slope to the south). We can't see or gauge how much fog has amassed that way vs. what spread north of the hill. 

Part of the fog's shape - the left-hand point of the triangle pointing right to its origin - seems to be missing by the time we see it.  Just where that red or green circled spot is will help solidify the overall shape and what it says about the wind, be it more north or more east.

To that end, it also seemed worth trying a comparison with the roads seen that way (off to the upper right in the panoramic view). This was a bit tricky, but I think it lines up nicely. I used a Google Earth perspective view facing south, enhanced a bit, and traced the roads most likely to be seen in white. One I didn't trace at first appears, so I added it (top view, but not bottom, and it's included in the graphics below)

I outlined the traced roads on their own layer, and with some additional skewing, perspective, and "cage transform" in GIMP (for the curve) I get a pretty nice fit. The under layers are enhanced views of the scene pre-fog and post fog (Salloum's videos via CNN), carefully composited so all details match. Both scene are stretched vertically so each road and the overall pattern is a bit more visible, and to compromise with the outlines trying to fit without squishing to unreadability. Not every road clearly appears, but enough do I say this is probably just the view. There are issues with the foreground buildings, and that nearest road is invisible behind a mild slope those building are up on. The main issue here is the fairly flat expanse beyond, the roads, and how that helps frame the fog origin. 

The red circle ... interesting as that spot is, it now seems to not be the source. By this, it would be further out, past the next road south, or about where the green circle is here, or perhaps further yet to the west or south. (At the available resolutions, TerraServer's free previews don't show any evident changes here between 2-22 and 4-6, but it might merit a closer look.)

If the mist came from there, this would be the last puffs released when the machine was switched off moments ago, now drifting with the rest off to the east and north, with that unsettled proportion. This is significantly different - about 600-650 meters south of the red circle. Clearly, if this is more the corner of the triangle, its overall north spread is more pronounced than from the red. If it's green, the wind direction is almost surely north of due east.

But that red spot was so intriguing - and it is hard to gauge how much blew south of the hill. Could that be the last bit drifting mainly to the southeast? It's also important to remember there's no one surface level wind to consider. At any given moment, yes, one direction. But from moment to moment the direction can change. southeast, then northeast, then east. Given all the variables, I still can't call it with certainty.

Except I can more clearly than ever rule out a few things:
1) any wind trend to the west,
2) any serious trend to the south,
3) nothing like a southwest wind (the direction alleged and required for the opposition's story to work).
2) any relative lack of wind that leaves topography as the main driver of fog spread (a backup argument floated by the OPCW). Downhill is the one way it doesn't get to move, because of a decent-force wind that happens to go the opposite way. Too bad they don't care and let "activists" convince them there was no wind and also just the right slope, and didn't bother checking to see both parts of that are untrue.

The conditions in the north of town are quite similar. The fog does not expand southwest from the grain silos, or to the west, depending (topography varies right there), but rather - it seems - north and east from  a spot next to the tel (large flat hill in town center), in a pattern suggesting it started in the tel's wind shadow, as the expanding vapor cloud seemed to be - if the wind was to the northeast. The wind existed, it mattered, and it's quite wrong for what the activists describe. (my graphic, to be updated perhaps, but still just about right.)