Jump to content

Wikipedia:Featured picture candidates/Image:Crops Kansas AST 20010624.jpg

From Wikipedia, the free encyclopedia
Circular crop fields in Kansas

NASA ASTER image of circular crop fields in Haskell County, Kansas from June 2001. I like the geometric patterns and it's a unique perspective on how fields look. Used in Center pivot irrigation, Ogallala Aquifer, Agriculture, Crop rotation, and Agriculture in the United States.

  • Nominate and support. - howcheng {chat} 17:18, 8 May 2006 (UTC)[reply]
  • Support Clear, sharp, lots of detail, and very encyclopedic. SteveHopson 17:22, 8 May 2006 (UTC)[reply]
  • Support. Very abstract image, but gets your attention and makes you want to learn more about it. Diliff | (Talk) (Contribs) 17:48, 8 May 2006 (UTC)[reply]
  • Support Superb. Very interesting perspective, excellent resolution. --Pharaoh Hound 18:31, 8 May 2006 (UTC)[reply]
  • Support, preferably unstraightened original. Great addition. Has the color saturation been increased? bcasterline t 19:29, 8 May 2006 (UTC)[reply]
    • Not by me. It doesn't say anything on the source page, so my assumption is no (if any modifications have been made, it usually says so). The ASTER page doesn't have any information about it either. howcheng {chat} 20:58, 8 May 2006 (UTC)[reply]
    • My guess is that it has been modified, or at least, that is not a natural representation of the colour. A quick look at Google Earth will quickly show you that there is a vast variance in the quality of the imagery. Some satellite imagery isn't even visible colour but rather based on the amount of reflectivity. Diliff | (Talk) (Contribs) 21:16, 8 May 2006 (UTC)[reply]
      • The latter is more likely. Considering ASTER stands for "Advanced Spaceborne Thermal Emission and Reflection Radiometer" it's probably not a natural-color image. howcheng {chat} 21:34, 8 May 2006 (UTC)[reply]
        • However, our own article on ASTER states that it takes visible spectrum images too, and since NASA makes several references to the color without any mention of it being false color, this is probably actually the real colors, and we should not presume otherwise. Night Gyr 22:56, 8 May 2006 (UTC)[reply]
  • Comment strangely, I would actually prefer a cropped version, especially to appear on the main page. You can't actually make any detail out in that thumbnail. Perhaps by cropping it you could at least see the circles and get an idea. Two questions not answered on the image page: What orientation does the image have (is there some reason not to "straighten it" from its current ~10° slant to the left)? Also, what is the road(?) running through the image? Stevage 20:01, 8 May 2006 (UTC)[reply]
    • Unfortunately, that information is not available in the source page. howcheng {chat} 20:58, 8 May 2006 (UTC)[reply]
    • I don't think that is going to help. There have been situations where I've wished that there was a way of allowing a cropped thumbnail to link directly to a full sized image but in this case, I think you just have to accept that it looks abstract and click on the thumbnail to see it at 100%. It wouldn't be any more recognisable when cropped anyhow. Diliff | (Talk) (Contribs) 21:16, 8 May 2006 (UTC)[reply]
  • Support - These are fascinating, particularly if you've ever used GoogleEarth to see how much of Kansas is covered by them. Really makes you think about the demand that feeding the Western world places on its environments. A quick GE of the coordinates shows that the road is Highway 56 which runs 640 miles from Springer, New Mexico to Kansas City, Missouri. --Yummifruitbat 21:39, 8 May 2006 (UTC)[reply]
  • Comment - Also, there's no obvious reason for the orientation of the image as the grid of circles runs N-S/W-E with remarkable precision. --Yummifruitbat 21:43, 8 May 2006 (UTC)[reply]
(getting horribly addicted to GoogleEarth!) - The two settlements alongside the Highway are the small towns of (lower left) Sublette and (upper right) Copeland.
So then the only possible explanations I can come up with for the tilt are (1) that's how the satellite was oriented, or (2) it was an artistic choice by the NASA person who processed the image. howcheng {chat} 22:11, 8 May 2006 (UTC)[reply]
Straightened/cropped version
It's (1), see below. Stevage 11:39, 16 May 2006 (UTC)[reply]
  • Edited version - I've uploaded a straightened and cropped edit as the rotated version is misleading IMO. One of the interesting features of this landscape is the meticulous geometric arrangement. --Yummifruitbat 23:49, 8 May 2006 (UTC)[reply]
  • Support Original, Oppose edited. Straightening the image just makes it look unnatural, I think. The grid is easy to see even angled, and it feels more like reality when things haven't been messed around with just for the sake of perfect alignment. I think the rotated version is more misleading, because it implies that the satellite lines its images up perfectly with the crop grid. Night Gyr 00:20, 9 May 2006 (UTC)[reply]
It is unnatural, and the perfect alignment is part of what makes it interesting. When the fields were originally created, they were laid out on a measured North-South, West-East, 0.5- and 1-mile grid. Presenting the image at an angle suggests that they were arbitrarily oriented - isn't that a bit like drawing a map of the Americas with Canada in the bottom right hand corner? --Yummifruitbat 00:43, 9 May 2006 (UTC)[reply]
It's a picture, not a map, and I think we can trust our readers to know that. I've been on cross-country flights before, and seen the pattern, so I know how it runs, and it seems more natural to me for a picture of the pattern to not align perfectly, since even the satellite isn't seeing it straight on. This picture is not a map. The change is more comparable to taking a picture like the blue marble and spinning it to line up with a map--completely unnecessary, insulting to our readers' intelligence, and distorting the compositional appeal of the original. Night Gyr 03:52, 9 May 2006 (UTC)[reply]
I disagree with putting "the original" on a pedestal in that way. The original rotation was totally arbitrary anyway. Why not rotate it to 45 degrees, 37 degrees, or 344? If you find the roughly 10 degrees to the left more appealing than ramrod-straight, then that's one thing. But considering 0 degrees (the apparently "true" alignment) to be "arbitrary" and the 10 degrees left to be more canonical is, well, very arbitrary indeed. Stevage 08:16, 9 May 2006 (UTC)[reply]
I do find the original more appealing, and I don't like the rotation because it takes away everything about the image that distingushes it from a generic satellite photo or map. There are plenty of places out there that can show you the precise north/south grid, but I feel like showing it at an angle makes it feel less like a map and reminds you that you're looking at a photograph rather than a drawing. Night Gyr 08:41, 9 May 2006 (UTC)[reply]
Comment Google earth can show you the same location here Night Gyr 04:07, 9 May 2006 (UTC)[reply]
It is a generic satellite photo, but it's the subject that distinguishes it, not the 10º rotation which adds nothing to the encyclopaedic nature of the image. You may have been lucky enough to fly over this scene but the vast majority of readers/viewers will not have, nor can one assume that they will have found the circles using GoogleEarth and happen to have had the Lat/Long overlay switched on at the time. Straightening the image increases its informational content - this is supposed to be an encyclopaedia after all. Should we refrain from straightening, say, a landscape nominated for FP with the horizon tilted 10º because the photographer's tripod was lopsided? --Yummifruitbat 11:19, 9 May 2006 (UTC)[reply]
It doesn't really add to the informational content. Why would a straightened image lead the viewer to assume that the fields themselves were straightened along a grid? Personally, I don't think I would ever make that assumption. It would need to be stated somewhere (the caption, for example), which would be equally necessary for both versions. Fitting to a grid is different than leveling a horizon. bcasterline t 12:59, 9 May 2006 (UTC)[reply]
It's a good idea to include that information in the caption. Perhaps my choice of phrasing was poor above - straightening it doesn't really increase its informational content, but IMO rotating it by an arbitrary 10º reduces the information conveyed because then the rotation becomes a feature of the image, even though it is of no real relevance to the content. Lots of images (eg. of buildings) are either opposed on FPC or edited, because walls aren't perfectly vertical or exhibit perspective distortion. It strikes me as a case of double standards, then, to say that rotating it to align a clear North-South grid with North is "insulting to our readers' intelligence". --Yummifruitbat 13:47, 9 May 2006 (UTC)[reply]
Maybe there is a double-standard (although I don't myself oppose the straightened version). But nature shots, except where there is a visible horizon, aren't always straightened with relation to anything in the picture -- see Image:Tulip - floriade canberra.jpg below, for example. The focus of this image is the crop fields, not the gridding of Kansas. And, as a matter of personal preference, I find them more interesting when not perfectly aligned in a square. bcasterline t 14:12, 9 May 2006 (UTC)[reply]
Well put, I can see your point. From a purely aesthetic point of view I quite like the angled shot, but I personally think the straightened one is more encyclopaedic. --Yummifruitbat 14:55, 9 May 2006 (UTC)[reply]
  • Weak support original, support edit, and I would strongly support an edited version that lined up the edges of the image and the edges of the outer crop circles. It just looks more striking when perfectly north-south and east-west. Cuiviénen (talkcontribs), Tuesday, 9 May 2006 @ 01:29 UTC
I had a go a this but it didn't look all that great as the pattern doesn't line up perfectly all the way along. I think there's either some distortion in the corners (which I tried to correct but couldn't) or the guy with the measuring tape had been in the sun too long... --Yummifruitbat 11:23, 9 May 2006 (UTC)[reply]


Vote for your favourite angle

[edit]

Since the rotation of the above image is totally arbitrary, I propose we vote on which way to rotate it for the final image. Consider that 0 degrees is the straightened version above, and -10° degrees is the version as originally nominated. If your preferred angle is not included, feel free to add it in the list where it fits. If you oppose the image, please say so above this section.

  • 0° (straightened)
    • Comment: I support either versions,but I think it would be more encyclopedic to represent it in its conventional orientation with the North pointing upwards. The original picture has a totally arbitrary orientation, and even if it just looks nicer, Wikipedia is primarily an encyclopedy and not an art gallery. Glaurung 06:30, 11 May 2006 (UTC)[reply]
      • IMHO, the best thing to do would be to present the angled version as the FPC, but probably use the straightened one in the actual article, but I suspect that would be against the spirit of FPC. There is obviously an inherent conflict between what is aesthetically pleasing ("striking", "best work") and what is encyclopaedically useful. A simple example is that often heavily cropped versions or even details of pictures are more useful for actual articles, whereas a full, uncropped image would be better as an artwork. Stevage 14:21, 11 May 2006 (UTC)[reply]
  • -10° (original)
Local magnetic north for that region would be 7.7 degrees from true north according to NGDC. What's 'global north'? --Yummifruitbat 03:10, 10 May 2006 (UTC)[reply]
Global north is apparently what I say when I forget the term "true north". ;) --Dante Alighieri | Talk 05:09, 10 May 2006 (UTC)[reply]
North is up. --Yummifruitbat 02:56, 10 May 2006 (UTC)[reply]
Good idea - it'll be interesting to see what other details about the image they can give us. Even if there's no reason for the rotation, I'll go with the consensus and support the original as this image deserves to be FP in one form or another. --Yummifruitbat 16:50, 10 May 2006 (UTC)[reply]
No response as of yet. --Dante Alighieri | Talk 22:01, 12 May 2006 (UTC)[reply]
Response from NASA. Details below. --Dante Alighieri | Talk 20:32, 15 May 2006 (UTC)[reply]

NASA Response

[edit]

I had sent an email querying the ASTER team as to some details about the image. They graciously responded and answered all the questions:

I can answer your questions:
1) the image is aligned with the satellite orbital track, which is in a 98
degree tilted orbit. North is not "up". Rotate the image about 10 degrees
clockwise to align the roads north-south.
2) The image is a false-color presentation made to simulate natural color.
The 3 bands that were used are in the green, red, and near infrared parts of
the spectrum. ASTER does not have a blue channel, so we have to create one
from the other bands.
Michael Abrams
ASTER Science Team Leader
NASA/Jet Propulsion Laboratory

Now that we know the reason for the tilt, people may be more comfortable deciding which tilt they prefer. Also, we should add the bit about false-color to the picture summary. --Dante Alighieri | Talk 20:32, 15 May 2006 (UTC)[reply]

Done. howcheng {chat} 20:40, 15 May 2006 (UTC)[reply]
Good work Dante, interesting to hear about the false colour. --Yummifruitbat 23:12, 15 May 2006 (UTC)[reply]
Thanks for doing that. Interesting that my original hypothesis about the arbitrary angle is essentially true - it's basically akin to looking out of a plane, pointing straight down, and taking that angle as gospel. However, the fact that it's false colour really wrecks it for me. It's now little more than a computer generated image of some geographical data. Along with an arbitrary angle, we have arbitrary colours. Hmm. Stevage 11:37, 16 May 2006 (UTC)[reply]
Well, not precisely... I mean, the green and red channels are real data via visible light, presumably... given that they also have infrared data, and they probably have other data from the Terra satellite (or other satellites), they can probably do a lot more than "guess" at the blue channel. I'd say that rather than "false color", we might want to call the image "adjusted color" or something like that. The base image is still real visible data. As for taking the angle as "gospel", it's not so much that, it's just that that's what the satellite really "saw"... so why bother altering it? --Dante Alighieri | Talk 14:45, 16 May 2006 (UTC)[reply]
Well, when someone takes a wonky photo of a building, we straighten it. But yeah, we've talked about the issues of straightening this image for long enough, and most people prefer the unstraightened version. Stevage 18:03, 16 May 2006 (UTC)[reply]
  • Support original. The "false" color are common in satellite images - scientifically, there's less use for a blue channel. But, it is entirely possible to create a "synthetic blue", as is (was) done with blue-screen special effects filmig, so I'd say the colors actually look pretty natural. --Janke | Talk 18:39, 16 May 2006 (UTC)[reply]

Promoted Image:Crops Kansas AST 20010624.jpg ~ VeledanTalk 19:21, 16 May 2006 (UTC)[reply]