Thanks for the cool website and write up, I’ll be sure to use this for the next local eclipse!
Cesium does track Moon and Sun positions which can easily be converted to Earth fixed frame, in fact I use this in the 3D compass demo. What about creating a sun_moon vector with CC3.subtract(moonposition,sunposition,sun_moon) Then use that vector to point a cone whom’s base is at the moon? (Or cut off the first 180 or so mega-meters of the cone base and just show the part hitting the Earth.) This would allow people to see why the shadow is of a particular shape (intersection of a cone and oblate ellipsoid.)
It’s odd that the 3D shadow volume is conical, perhaps this is due to light bending around the moon? If the 3D shadow volume were cylindrical the shadow would be much larger as the Moon is at least 2/3 the size of the United states.
Creating a cone with the sun/moon vector is a pretty interesting idea. I haven’t read up enough on the variables in play for the Besselian elements, but presumably that same sun-moon vector probably comes into play. Certainly since Cesium can display a partially transparent cone (a volumetric shadow) I could get away with showing just that as a preliminary or additional step to finding that cone’s intersection. I hadn’t considered that, so thanks for the good idea!
As for the shape of the shadow the moon’s gravity does bend the light but imperceptibly slightly. It’s the fact that light source is not a point but a disc that creates conic shadows. There are actually two shadow cones in play; the narrow umbral cone and the wider penumbral cone, as shown here:
I’m glad you like the idea, I thought it would be a nice addition to the excellent program you’ve made. Perhaps this could show near misses as well, eclipses that almost happened!
That makes sense about a disc light source, the Sun is simply gigantic relative to the Earth. To get a good sense of the size of the Sun check out this SandCastle example https://groups.google.com/d/msg/cesium-dev/psRJ4XBUWZ0/Q5DmV2fuFH0J where the distance between the Earth’s center and the Sun’s center is only 1 GigaMeter (rather than the actual 150 GigaMeter distance.)
To calculate a translucent umbra shadow cone with an accurate shape you’d probably have to consider the distances between the Sun, Moon, and Earth since these vary somewhat throughout the year. I think the cone angle can be determined by tan (half cone angle) = opp/adj where opposite = (radius difference between Sun and Moon) and adjacent = (distance between Sun and Moon.) The bottom base of the cone being at the Moon center with the same radius as the Moon. I’m not sure about the prenumbra though.
There was a hoax ISS Solar Eclipse picture posted today
While this may have been fake, it is possible that the ISS could be in an Eclipse, though briefly due to the speed of the ISS. Sometimes the shadow cone from the Moon might miss Earth but intersect some satellites.
The 2015-03-20 event on http://eclipsetracks.org/ only runs from 09:18:00 to 10:14:00 UTC. Proba-2 catches the eclipse before and after this time period, apparently when the shadow cone misses Earth. It would be neat to see the moving shadow cone and satellites it hits in Cesium.