Just to verify[1] the revised size estimates:
Uriel’s revised estimate:
This puts the structure (be it mobile or im-) at ~20-21km front to back. Obviously, we can’t get an Avatar in there to sit next to it like this. However… the latest video does give us an extremely convenient comparison. Why?
Because of how long a Gila is.
For reference, this is the frame of the video which gives us our benchmark:
It’s an almost perfect profile shot of a 334m-long Gila, near a segmented cowling on the underside of the construction, that just happens to also line up as a profile image of the construction itself. And, because we know how long a Gila is…
That 1km result of 3 Gila-lengths is just extremely handy.
From there, we can then determine the size of the cowling:
This gives us the length of the cowling at just under 1.6km.
2km extends from the back of the cowling to in-line with a smaller area of highlight below and fore of the cowling:
And that, in turn, allows us to make use of a somewhat wider-angle shot:
In this, the green line on the inset reproduces the measurement from the narrower frame, while the red is 6km long. One final frame from the video gives us a less neatly-lined up viewing angle, but a more complete view of the front-to-back length overall:
The highlighted area is the cowling, giving us the rough size of the thing. Taking this view, we can see that the entire construction, front-to-back, is at least 12-15km long. Given just a bit of distance built into that initial frame (which is highly probably, given the position of the Gila as the camera moves), an additional 30-40%, putting the estimate into the range of 16-21km, is very possible.
- I don’t think Uriel’s estimate necessarily needed verification, but having public record of a second method for an estimate can be a useful pile of data, especially when the two estimates line up.