Image talk:Earth tilt sample.jpg

From Wikipedia, the free encyclopedia

Following some discussion on the reference desk, I've created a rudimentary image in Blender to show the tilt of the earth (Image:Earth tilt sample.jpg). It strikes me that an animation might be tricky, though: if I allow 24 frames for one rotation of the Earth, it'll take ~8760 frames for a complete revolution around the sun (almost 5 minutes at 30fps). So clearly an animation would need to be carefully crafted to show significant transitions during the revolution, namely the equinoxes and solstices. Any suggestions on how best to use this illustration? (I'll definitely brighten up the Earth and add some reference lines for the equator and whatnot). -- Wapcaplet 21:55, 16 May 2004 (UTC)

Super. Even a few (or one) static images will make a huge difference, and I'm not sure an animation is necessary. But certainly this concept just can't be explained in words (I can tell I'm failing miserably) and the math is inaccesable to most. Some suggestions:

• the purple line is confusing (it looks like an axis, but I figure it's merely a normal to the eclipic). Is it possible instead to actually draw the ecliptic instead, perhaps as a translucent nonreflecting purple disk, with its outer edge at the orbit of the earth, or something like that? I think the confusion some folks seem to have is they thing the earth's axis of rotation is perpendicular to the ecliptic, which makes for a regular (but dull) system.
• I think the terminator should be harsher (I'm guessing blender's lighting model has a "falloff" parameter, or similar)
• I think the sun is too big (I mean artistically), and maybe the stars are a tad cheesy :)
• You're right that it's wasteful (and boring) to show all the earth's orbit, and indeed the solstices and equinoxes are the important thing. I think just a single image with the earth shown in four (or maybe eight) places around its orbit would be great. An animation would need lots of popup text, which is probably too much work (and too much effort to watch too).
• The thing that really made the idea stick for me was the idea that the whole solar system was like a giant clock - that a given date corresponded with a given place in the earth's orbit. Is it possible to write the months somehow around the orbit (whether radially or around the edge) to show this?

But it looks great already - thanks! -- Finlay McWalter | Talk 22:28, 16 May 2004 (UTC)

People seem to be generally ignoring my homely line drawings. I still have the source files for those. Are there any changes I ought to make?

As regards Wapcaplet's image, I think the terminator is fine (twilight does tend to last a while), and there's no problem with the stars. If you were going to make it into an animation, it might be best to have each day be approximately 3 seconds long, and vary the point of view so that it initially shows one of the poles and ends up nearer the other. --Smack 05:54, 17 May 2004 (UTC)

Comments noted. Smack - there's nothing wrong with your drawings! They serve their purpose quite nicely, but I think the concept we're attempting to show here may require a different perspective. I believe having some kind of 3D element may help people to better visualize what is happening. I like the idea of showing the poles separately; I will definitely try to work that into it somehow.

I've revised the image a slight amount. Tilt has been removed for now while I work out how to get animation in order (even if I don't actually do an animation, it's good to have the keyframes, armatures etc., the "clockwork" so to speak).

I was in the process of going on a rambling tangent here about a puzzlement I had with the discrepancy between actual day length and the time taken for the Earth to rotate 360 degrees, but I have just found my answer in sidereal day. That's another concept that may benefit from some illustration... -- Wapcaplet 20:41, 17 May 2004 (UTC)

You say you've removed axial tilt, but the axis still looks tilted. It also looks like it's attached a little too close to Greenland.

Shall we move this discussion to Image talk:Earth tilt sample.jpg? --Smack 22:48, 17 May 2004 (UTC)

Yes :-) The tilt really is gone; could just be a trick of perspective that it doesn't appear so. The axis/poles should be in the right place; I downloaded a spherical image map of the Earth at high resolution and just slapped it onto my model, but I'll check to make sure. Wish I had an actual globe to verify it, but this image shows that Greenland is pretty far up there. I'll keep you posted... -- Wapcaplet 22:11, 18 May 2004 (UTC)

Update: New image showing Earth in four positions. Now that I see all four together, I'm thinking it'd be best to render this in an orthogonal view instead, since the pink tilt axis looks out of whack due to perspective distortion. I've sharpened up the Twilight Zone in order to more clearly distinguish day and night, particularly around the polar regions since that's kind of the goal of this exercise. Made polar regions white. Verified that the axis is in the right place with respect to Greenland (easiest to see in the right-side image - the dangling white part is Greenland, the rest of the white blob is the arctic). I finally got the right mechanisms in place after beating my head against Blender's animation system for a while, so now you can basically just pick a date, and I can give you a rendering. Comments? -- Wapcaplet 01:52, 19 May 2004 (UTC)

Yes, I realized after I posted that last comment that the perceived axial tilt was a trick of perspective. I believe it would be optimal to make the point of view infinitely distant. Try messing with the size of the Sun again; it doesn't look quite as big with four Earths orbiting it. Finally, a question. Issues of visual appeal an animation aside, what does your image convey that mine do not? --Smack 16:56, 20 May 2004 (UTC)

New version uploaded. View is now orthogonal, and eight positions are shown. You make a good point; I'm not exactly sure what this image illustrates that yours don't, aside from possibly making the poles' transition from light to darkness a bit clearer to the 3D-impaired. I believe with appropriate labelling, this illustration could help to show what Finlay McWalter was originally getting at, that the poles do indeed have days and nights like everyplace else; they just stay dark/light for long periods of time due to axial tilt. I'll work on some labels... -- Wapcaplet 14:40, 22 May 2004 (UTC)