Physics Forums

Physics Forums (http://www.physicsforums.com/index.php)
-   General Discussion (http://www.physicsforums.com/forumdisplay.php?f=14)
-   -   Atlantis, the lost continent: Please debunk/interpret these photos (http://www.physicsforums.com/showthread.php?t=582649)

jay.yoon314 Feb29-12 05:25 PM

Atlantis, the lost continent: Please debunk/interpret these photos
 
3 Attachment(s)
Hi, I'm Jay and I'm new to Physics Forums. I'm an undergraduate student studying physics and computer science. I was wondering what your thoughts were regarding the following photos I took on Google Earth.

Edit reason: Maximum number of photos.

Thanks very much. My personal opinion is that whatever the structures and patterns that are underwater, they are not created by natural processes because natural processes are "nonlinear" in their nature. For example, a mountain, canyon, ridge, or pattern in nature generally shows a fractal-like self-similarity that makes its features rough and bumpy rather than smooth and patterned distinctively. I'm looking forward to your impressions and thoughts and a lively debate.

DaveC426913 Feb29-12 05:27 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Quote:

My personal opinion is that whatever the structures and patterns that are underwater, they are not created by natural processes because natural processes are "nonlinear" in their nature.
You cannot assume that Google Maps is the authority on the topography of the ocean bottom. You have no idea what resolution the imagery of terrain is. It could be so rough that the shapes you are seeing are simply measurement artifacts. For example, data points every ten mile increments will produce straight lines ten miles long.

And there are plenty of natural processes that create linear patterns.

Evo Feb29-12 06:20 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Jay, you may present the pictures, but you are not allowed to claim that they are anything, that's against the rules.

SHISHKABOB Feb29-12 06:23 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
for example: compare these two images of the Namibian coast line

the first one, http://upload.wikimedia.org/wikipedi..._AU_Africa.svg is fairly smooth. Not exactly linear, but quite smooth indeed.

However a more zoomed in map gives us this: http://upload.wikimedia.org/wikipedi...etailFinal.jpg

much more bumpy



since the ocean floor is 1.) not very widely mapped or studied in general and 2.) not very much cared about to that level of precision, and also what Dave said about Google Earth's merits as a source for topographical images of the ocean floor, I would not get too excited about these pictures

DaveC426913 Feb29-12 07:34 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Now that I see the actual images I can state categorically that they are rendering artifacts of Google Earth software. I've seen Google Earth render artifacts just like that near my home (which is not near Atlantis).

Go to Toronto and glide across the city NorthWest toward Bolton. You will pass over a hill that appears to be several hundred feet high, vertical and razor-straight for dozens of miles. It blithely ignores all landmarks such as roads - which go straight up the "cliff" face.

The artifacts occur because Google Earth is calculating and extrapolating altitudes. Sometimes it drifts from proper values, and must "correct" itself - which is does abruptly.

jay.yoon314 Feb29-12 08:24 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Quote:

since the ocean floor is 1.) not very widely mapped or studied in general and 2.) not very much cared about to that level of precision, and also what Dave said about Google Earth's merits as a source for topographical images of the ocean floor, I would not get too excited about these pictures
Quote:

You cannot assume that Google Maps is the authority on the topography of the ocean bottom. You have no idea what resolution the imagery of terrain is. It could be so rough that the shapes you are seeing are simply measurement artifacts. For example, data points every ten mile increments will produce straight lines ten miles long.

And there are plenty of natural processes that create linear patterns.
To SHISHKABOB and Dave:

I have come to agree that linearity versus nonlinearity, in general, is not a very good heuristic in making a distinction between a human-modified versus a naturally formed landmark.

I don't think Dave got a chance to actually see the pictures because I didn't upload them correctly at first, but to SHISHKABOB:

I very much agree with the fact that the ocean floor has been, and still is, mapped to a lower resolution than Earth's land surface. However, it is necessary to say precisely what we mean when error is discussed, and I argue that the existence of error should be a fact that leads to more questions rather than one that closes the discussion.

A relevant difference between the first two, and the third image, I think, is that in the first two, there is a very large amount of vertical relief. Even taking into consideration the fact that the resolution of the image is not at all means exceptional, for Google Earth to display an escarpment 1,250 feet high, oriented almost exactly from west to east, and with the vertical relief increasing in equal ratios with equal ratios of distance traveled from the westernmost point to the easternmost point along the entire roughly 30,000 feet length, as a false artifact that's not really there is, I feel, highly implausible.

I will further clarify that image artifacts that are specifically caused by low resolutions must be of a size not too much greater than, and ideally equal to or lesser than the resolution diameter of the image. In other words, an object much larger than the resolving diameter of an image cannot be a resolution artifact, which is a systematic error, unless it is maintained that the resolution of the image is so poor that it is essentially worthless, and therefore must be a random error.

Presupposing that the first two images represent faithfully the actual features present underwater at that location, I'm guessing that it would be a fault line, particularly that of the thrust type; Miranda, a moon of Uranus, has a 20 km high fault scarp: http://www.nasaimages.org/luna/servlet/detail/nasaNAS~4~4~15336~118118:Miranda-High-Resolution-of-Large-Fa

I won't further speculate as to what the formation is, and anyhow, I honestly have no other ideas.

DaveC426913 Feb29-12 08:29 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Quote:

Quote by jay.yoon314 (Post 3791813)

I don't think Dave got a chance to actually see the pictures ...

I did. Please see my post#5.

There is absolutely no doubt in my mind that these are artifacts of rendering.

Remember, the 3D modeling resolution has nothing to do with the resolution of the image you can pick up by zooming in. Yes, indeed, you can be looking at a very large image of what, in the renderer, is a single data point, extrapolated beyond reasonability.

Additionally, just like in their land pics, they must stitch together pics, and those pics have edges. When two pics have discrepancies in their altitudes, they much be reconciled.

DaveC426913 Feb29-12 08:47 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Wow. Google Earth sucks. Either that, or my computer does.

When I first played with it a few years ago, I could glide across the surface like a plane. It was perfectly smooth. Now, it's awful - like 2 frames per second awful.

SHISHKABOB Feb29-12 08:57 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
if you have the free time, jay, I would suggest going around with Google Earth and trying to find more situations like this. If you find a LOT of them and also in places that are well known, like a harbor or something, then I would say that they are probably a common occurrence.

Or even look deep underwater, if you find something like this that is like 2000 meters below the surface, then there is a problem with your explanation.

jay.yoon314 Feb29-12 09:04 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Quote:

Quote by DaveC426913 (Post 3791709)
Now that I see the actual images I can state categorically that they are rendering artifacts of Google Earth software. I've seen Google Earth render artifacts just like that near my home (which is not near Atlantis).

Go to Toronto and glide across the city NorthWest toward Bolton. You will pass over a hill that appears to be several hundred feet high, vertical and razor-straight for dozens of miles. It blithely ignores all landmarks such as roads - which go straight up the "cliff" face.

The artifacts occur because Google Earth is calculating and extrapolating altitudes. Sometimes it drifts from proper values, and must "correct" itself - which is does abruptly.

I did experience that effect, but it was very minor and almost unnoticeable. Even, then it was only during the glide itself, when I was in continuous motion from Toronto to Bolton (I did some of the glides at a constant altitude throughout and some in such a way that I glided "down" at an angle from Toronto to Bolton). I do agree with the fact that there are rendering artifacts, and that calculation and extrapolation of altitudes occurs. But this point does not apply to the photos I took, because they were taken in a static position in which all of the rendering, calculation, and extrapolation had fully adjusted to the camera's particular viewing angle and altitude, whereas the example you describe is dynamic gliding from Toronto to Bolton.

I maintain that if Google Earth ran on a machine with a rendering engine that was powerful enough to correct for altitude changes faster than our eye could notice, there would be no abrupt discontinuities and artifacts even during the glide, and the terrain would transition seamlessly (in "real time"), just as our visual system is able to do. Again, the artifacts you mentioned disappear when not gliding (or moving in any way), and would do so even when gliding if the computer had a really powerful graphics card. The image I took was taken when I was completely still, so if the formation is an artifact, it would have to be a "static artifact."

But there is no reason for a static artifact to exist at the location I mapped, other than as an ad hoc argument. Such an anomaly would be a form of random, rather than systematic error. But if it indeed is form of random error, then the randomness of the error has a systematic commonality in that if you swing around at any angle and height, the formation retains its essential geometry. Furthermore, the "artifact" increases its clarity and distinctiveness when viewed at roughly "eye level," just as the 3-dimensional geometry of an object seen far below or above eye level, in everyday experience and on dry land, is difficult to make out. In other words, the supposed artifact changes in appearance in precisely the way one would expect a "real artifact" to do so.

I argue that there is no basis, save for claiming the contents of the images contain an arbitrary and random anomaly, to assert that the images (only the first two images, the third one, I agree, is suspect) do not represent, more or less, whatever geological feature that is actually there.

Since what I am claiming is that the image reveals the existence of a geological anomaly whereas you are claiming that the image reveals the existence of an imaging anomaly, we seem to have some common ground, namely that if the image is "real," then it is an anomaly, and that if it is not, then it is an anomaly as well, but of a different kind.

DaveC426913 Feb29-12 09:06 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Quote:

Quote by jay.yoon314 (Post 3791903)
I do agree with the fact that there are rendering artifacts, and that calculation and extrapolation of altitudes occurs. But this point does not apply to the photos I took, because they were taken in a static position in which all of the rendering, calculation, and extrapolation had fully adjusted to the camera's particular viewing angle and altitude, whereas the example you describe is dynamic gliding from Toronto to Bolton.

No. You misunderstand. The cliff is quite static, whether you glide over it or stop. It has nothing to do with the gliding. I am not talking about a movement artifact. As far as Google Earth is concerned, there is a cliff there, 200 feet high and many miles long. It is there no matter what I do or don't do.

I should never have mentioned the gliding thing. It is confusing you.

Quote:

Quote by jay.yoon314 (Post 3791903)
Again, the artifacts you mentioned disappear when not gliding (or moving in any way),

No, they do not.

Quote:

Quote by jay.yoon314 (Post 3791903)
The image I took was taken when I was completely still, so if the formation is an artifact, it would have to be a "static artifact."

Precisely.

jay.yoon314 Feb29-12 11:50 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Quote:

Quote by DaveC426913 (Post 3791906)
No. You misunderstand. The cliff is quite static, whether you glide over it or stop. It has nothing to do with the gliding. I am not talking about a movement artifact. As far as Google Earth is concerned, there is a cliff there, 200 feet high and many miles long. It is there no matter what I do or don't do.

I should never have mentioned the gliding thing. It is confusing you.


No, they do not.


Precisely.

Could you please upload a picture of your particular view that clearly shows what you are describing? My friend and I have tried, for almost an hour, to find what you've described, and we haven't been able to do so.

Mech_Engineer Mar1-12 12:15 AM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
The fact that you're trying to use Google Earth as an authoritative resource concerning underwater topology is hopeless. The ocean floor (and the entire planet for that matter) is modeled in there as a low-resolution polygonal approximation, and what you're seeing is exactly as Dave has described: a graphical glitch in the software.

jay.yoon314 Mar1-12 03:39 AM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Quote:

Quote by Mech_Engineer (Post 3792249)
The fact that you're trying to use Google Earth as an authoritative resource concerning underwater topology is hopeless. The ocean floor (and the entire planet for that matter) is modeled in there as a low-resolution polygonal approximation, and what you're seeing is exactly as Dave has described: a graphical glitch in the software.

I forgot to mention that Google is not the primary source that collected the data. The actual images were taken by the U.S. Navy and NOAA, and three other institutions. The data was collected independently by each institution, and so if there's an error, not only would all of these institutions have collected the wrong data, they would have had to have made the same error in the same place in the same way.

The resolution is determined by the data that was collected by multiple sources (not Google Earth). The accuracy of this data determines the maximum resolution.

If NOAA or the U.S. Navy generated these images from their own data and using their own software, which they probably have (it's far from likely that Google Earth was the first software of its type, it's just the first one's that been offered free to the public), and those images show the same formation, the probability that it's just a bug diminishes to almost zero, literally.

If that were the case, would you bet against the U.S. Navy?

Borek Mar1-12 04:43 AM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
There are two kinds of artifacts in this thread. One is a Google artifact:

http://www.informationweek.com/news/...ogle/214502197

I am afraid second is a thinking artifact - Jay, you are so sure of being right you don't listen to arguments.

Dave: are you sure the cliff is still there? I have seen many such things corrected in Google Earth over the years.

Edit: compare this search. Zillions of artifacts observed.

DaveC426913 Mar1-12 08:40 AM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Quote:

Quote by jay.yoon314 (Post 3792214)
Could you please upload a picture of your particular view that clearly shows what you are describing? My friend and I have tried, for almost an hour, to find what you've described, and we haven't been able to do so.

Quote:

Quote by Borek (Post 3792468)
Dave: are you sure the cliff is still there? I have seen many such things corrected in Google Earth over the years.

The moment I went to download Google Earth, I realized that of course it's going to be gone. They update their data all the time. It was foolish of me to suggest the data would not change over a year or two.

DaveC426913 Mar1-12 08:43 AM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
Quote:

Quote by jay.yoon314 (Post 3792407)
I forgot to mention that Google is not the primary source that collected the data. The actual images were taken by the U.S. Navy and NOAA, and three other institutions. The data was collected independently by each institution, and so if there's an error, not only would all of these institutions have collected the wrong data, they would have had to have made the same error in the same place in the same way.

The resolution is determined by the data that was collected by multiple sources (not Google Earth). The accuracy of this data determines the maximum resolution.

If NOAA or the U.S. Navy generated these images from their own data and using their own software, which they probably have (it's far from likely that Google Earth was the first software of its type, it's just the first one's that been offered free to the public), and those images show the same formation, the probability that it's just a bug diminishes to almost zero, literally.

If that were the case, would you bet against the U.S. Navy?

You are ascribing too much to this.

Google Earth simply renders polygons based on however detailed or sparse the data points are. If data points are of distance X apart, yet you choose to zoom in such that X fills your screen, it will render a polygon as wide as your screen. That's what you are seeing.

As for differing data sources, well, that certainly would explain why there are discrepancies, wouldn't it? As anyone who uses Google is familiar with, even the best resolved areas have discrepancies in their data, showing discontinuities. True, the ones on land are of a different type but the principle remains - two datasets don't stitch together perfectly and you get a road or building or city block that is obviously misaligned with its neighbour. You know perfectly wll this happens all over the place (though they are getting better).

For some reason you forgive these errors as insignificant artifacts because you understand them, yet when it comes to the ocean bottom, you short-circuit this conceptual caveat and you assume that the map is the territory.

jay.yoon314 Mar1-12 05:19 PM

Re: Atlantis, the lost continent: Please debunk/interpret these photos
 
3 Attachment(s)
I'll post some more photos, from now on, on a more regular and frequent basis.

The first photo attached was taken just off the eastern coast of Nicaragua in the Caribbean Sea, in very shallow water, in a place called Cayos Miskitos. Second photo, not much to really say. I'll find some better ones. The third photo is just a closer view of the first one.


All times are GMT -5. The time now is 03:05 AM.

Powered by vBulletin Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
© 2014 Physics Forums