Redo of this and this and all the others except curvature is overrated so this time we are scaling to something else instead.
Width of Japan = 17.1 px
Except the perspective makes vertical lines seem smaller than they should, so assuming its a perfectly circular hole we calculate the "eccentricity" as:
294.5px / 136.3px = 2.16067498
(Using this eccentricity yields a lower end result than not using it by the way)
17.1 px * 2.16067498 = 36.9 px
Therefore Width of Japan = 36.9 px = 233.3 km
233.3 km / 36.9 px = 6.3225 km per px
I feel like at this point I should say that while this method of scaling is certainly not perfect, in my opinion it's still far more accurate than using curvature as the people behind the show would certainly have put more thought into drawing this hole in relation to their home country than calculating the proper curvature of the planet for a scene that lasts like 5 seconds.
Top of Hole Diameter = 494.1 px = 3132.94 km
Top of Hole Radius = 1561.97 km
Bottom of Hole Diameter = 294.5 px = 1861.97 km
Bottom of Hole Radius = 930.99 km
Hole Depth = 77.23 px = 488.27 km
Truncated Cone Volume = 2.4342e+18 m^3
95% Hollowness Volume = (2.4342e+18 m^3)*(0.05) = 1.217e+17 m^3
Density of Silk = 1310 kg/m3
Mass = (1.217e+17 m^3)* 1310 kg/m3 = 1.594e+20 kg
Distance is average of the two Radius'
Distance = (930990 + 1561970)/2 = 1246480 Meters
Timeframe is actually closer to 12 Seconds as Ryuko starts stabbing through the Cocoon at 14:04 and the hole is seen on her background at 14:16.
Velocity = 1246480 Meters / 12 Seconds = 103873.333 m/s
KE = (1/12)*(Mass)*(Speed of Movement)^2
KE = (1/12)*(1.594e+20)*(103873.333^2) = 1.433e29 J = 34.2 Exatons (Moon level)
Very close to Baseline, so with some slight changes it could still fall to High 6-A. An excelent result nonetheless.
Update[]
Talking with a certain user I realized that the previous calc assumed that the hole was as close to the viewer as Japan. Which is incorrect as the Cocoon is portrayed as being quite high into the sky, meaning the result is marginally smaller. I'll do it anyway for the sake of accuracy.
Distance from pov to Japan = object size * panel height in pixels/[object height in pixels*2*tan(35deg)]
Distance from pov to Japan = 233.3 km * 768/[17.1*2*tan(35 deg)] = 7482.0925 kilometers
(The height of the whole panel was actually 768 pixels, the one you see above is only 648 because I cropped it before scaling)
I think it's safe to assume that the sphere starts at the end of the atmosphere (100km)
Distance from pov to Cocoon = 7482.0925 - 100 = 7382.0925 km
By simple algebra this means the radius of the hole is actually 7482.0925/7382.0925 = 1.01354629 times smaller than on the previous calculation.
To not do the entire calc again, I just use the fact that the radius of the hole influences the energy result by a fifth power (Volume-Mass by ^3 and Speed^2 by ^2). So to get the new result just calculate 1.433e29/(1.01354629^5) = 1.34e29 J. Which is still baseline Moon level.
Update 2[]
Was told to use a different way of calculating the depth of the hole. As the previous calc was kind of inflated.
This new measure is kind of angled in a way that lends a deflated result, but I couldn't think of anything better to do. And its ultimately better for the calc to be deflated than inflated as it was previously.
New depth is: (34.2 px / 77.23 px) = 0.443 times the size of the depth in the previous calc. Since depth in this case scales linearly to the energy result, we just multiply the previous result by this new factor.
1.34e29 J * 0.443 = 5.94e28 J which is only high-end Multi-Continent level to the dismay of all our audience around the world.
KLOL506's version (ACCEPTED)[]
OK so here's my version of the calc with the height being measured according to Mitch's recommendation- measuring the height between the base and the topside diameter.
Length of Honshu[]
OK so Eyeballing alone and by comparing it to google maps we can see that the width of Honshu starts from Kyoto and ends just after Wakayama on some small islands, and I got around 265 km for this.
Pixel-scaling[]
Honshu (Orange) is 60px. Height (Yellow) is 102px. Base (Green) is 422px and Topside (Blue) is 702px. This is a truncated cone.
Height: 265*(102/60)= 450.5 km
Base: 265*(422/60)= 1,863.8333 km diameter. Halve this to get a radius of 931.91665 km
Topside: 265*(702/60)= 3100.5 km. Halve this to get a radius of 1550.25 km
Volume and mass[]
Put the values here (It has a nifty Truncated Cone calculator as well) and you get a volume of 2.225043044781423597e+18 m^3.
95% hollowness so 2.225043044781423597e+18*0.05= 1.1125215223907117985e+17 m^3
Silk has a density of around 1310 kg/m^3, so (1.1125215223907117985e+17 m^3)*(1310 kg/m^3)= 1.457403194331832456035e+20 kg
Speed[]
OK so I watched Episode 24, Ryuko starts penetrating at 14 minutes 04 seconds and the hole is visible immediately in Ryuko's background in 14 minutes 16 seconds, so timeframe is around 12 seconds.
Distance is gonna be the average of the base and top radiuses.
(1550.25+931.91665)/2= 1241.083325 km or 1241083.325 meters.
Speed= 1241083.325/12= 103,423.6104167 m/s
Calc[]
KE is omnidirectional, so (1/12)*mass*speed^2
(1/12)*(1.457403194331832456035e+20)*(103,423.6104167^2)= 1.299086e+29 J or 31.0489006 exatons of TNT (Moon level)