Dust flux, Vostok ice core

Dust flux, Vostok ice core
Two dimensional phase space reconstruction of dust flux from the Vostok core over the period 186-4 ka using the time derivative method. Dust flux on the x-axis, rate of change is on the y-axis. From Gipp (2001).

Monday, February 14, 2011

Atlantis found and lost again

The internet is abuzz over what some claim to be the discovery of Atlantis, just off the west coast of Africa. This happens to be practically in my backyard.

Let's examine the evidence.

The overview map above comes from Google Earth, showing a region of the Atlantic Ocean west of Northwest Africa. The red boxes are my construction and show the location of the three images shown below. Image A is alleged to be Atlantis. The other two images show features which have not been commented upon by the discoverers of Atlantis.

Image A. The area shows the transition from the area dominated by ridge/shear structures associated with the mid-Atlantic ridge (to the west of this image) to the area where such features have been blanketed by sediments. Apart from a few seamounts, the eastern half of this area is relatively featureless--except for that impressive looking series of orthogonal "roads" near the middle of the image.

The simple story is that these "roads" are an artifact of the process by which Google compiled (and then presented) existing data to create their interactive map of the earth. Google doesn't (yet) have marine vessels surveying the oceans, so it must rely on existing data. There are two principal data types--blanket satellite coverage of the oceans, where gravitational anomalies measured over long intervals are used to infer seafloor topography, and depth soundings by marine survey ships.

The satellite altimetry, produces very low resolution images. Features normally have to be spectacular in order to have enough of a gravitational effect to be teased out from ocean surface data (subtracting out the effects of waves and tides). The resolution obtained by a ship survey is easily an order of magnitude more detailed.

Now suppose you have an area of seafloor which is covered by relatively small hills and valleys--features which are too small to be resolved by the satellite image. These small features are easily detected by a surveying ship.

Here is part of a seismic profile collected many years ago using NSRFC's V-fin (neither of which exist anymore). This shows the seafloor as well as layers of sediment below the seafloor. The v-shaped notches in the seafloor are pockmarks, which are formed by methane gas escaping from beneath the seafloor, and the two largest pockmarks in this image are no more than 10 m deep. They are considerably wider, as this image has a vertical exaggeration of about 20x. (I would have used this image in my MSc thesis, but it is not close to hand at this writing, so I don't have the exact numbers).

The point is that a surveying vessel is capable of resolving an object that may be 10 m high and a few hundred metres across, whereas the satellite gravimeter needs an object to have kilometre scale.

Google does not present contoured data. What they do is create a model, which is then illuminated, and the sense of topography is created by the resulting shadows.

If a large area that is only covered by low-resolution satellite data, and which appears to be featureless is traversed by a survey line which resolves many small features--and then that data is all illuminated by some computer program, then what you will see is a featureless plain traversed by a single dark line. That line is too thin for you to resolve any of the features it shows you, but the artificially generated shadows will trace out the track of the ship.

Now let's look at some other areas of the seafloor.

Here is area B. Notice the "road" emanating from the seamount just east of centre to another seamount just beyond the northern limit of the image. Another set of "roadways" cover the western portion of the image. Those ancients sure were busy.

Area C shows a major highway stretching from Portugal towards Atlantis. Another "road" swings to the north.

The northern Atlantic Ocean isn't the only place covered by roads. Let's take a look at southern Africa.

There are a lot of roads leading to Cape Town. I don't know what was there, but it seems to have been of some importance.

Here is a patch of the Indian Ocean, covering a mid-ocean ridge between India and Sudan. Note the high relief, but there are lines of higher-resolution surveys which show up against the lower-resolution background satellite data.

Marine surveys have two common forms. The accidental surveys, like in the image immediately above, occur when ships leave their equipment running as they leave port. You end up with a lot of lines emanating from a major port. The other common form is an orthogonal grid.

Sad to say, the above looks like a survey. If I had to guess, I would suggest an ODP site survey.

Here is a different survey. If memory serves, this is John Hughes Clarke's SEAMARC survey.

More "roads" on the Nova Scotia continental slope. I actually helped make some of these, and there were no Atlanteans on the boat with me.

Besides, I thought Atlantis looked like this.

More on the Japanese discovery here.


  1. My biggest question in regards to this is why there has been no one the has had try to go discover these underwater locations. I have seen these photos for years. Please say if you know any research about this!

    1. Um, next time read the text that goes with the pretty pictures. The author didn't put in "sarcasm" html tags, but it is still pretty obvious he is saying the lines are simply a data artifact from two different types of sensing. One really large scale provides the background for most of each image, and the "lines" that look like roads are simply much higher resolution scans that scale down to what seems to be a line in each larger picture.

      In short, these lines were in fact *caused* by someone who went and scanned these locations; they didn't go there because of the lines, but by taking a higher resolution sensor, and then allowing their data to be assimilated by Google Earth, the very act of exploring the area created them.

      It is as if you are looking at a photo of footprints at Tranquility Base on the moon, and asking when we are going to send men to the moon to check out obvious signs of extra-planetary intelligence.

  2. l think atlantis is in greenland near Nuuk under water.

  3. Atlantis was discovered long time ago , several Political guys don't wont common people knows , but we've Google maps haven't we? Look Madeira island left side in Satellite view, the circle of Atlantis:

    Link :http://www.madeira-web.com/pagesuk/map.html

  4. Appreciate the analysis but if you every read the actual citation for Atlantis in Plato's dialogues (re Timaeus and Critias) you will find any rebuttal unnecessary as Plato invented Atlantis as a joke. Albeit a rather nerdy one (in Grecian terms). The story was passed down in this way:

    The boys are at a 3 day conference. The previous day the discussion entailed what would constitute a perfect government. At the time none could recall one ever having existed so they discussed various requirements for one. The discussion managed to jog Critias' memory later that evening.

    The next day Critias relates the story of a most perfect government:
    Critias heard the story from his grandfather who was told it by Critias' great grandfather who heard it from his good friend Solon. Solon himself was told the story while visiting Egypt by an Egyptian priest living in Sais on the Nile Delta. Here Solon was told that Greeks didn't know old from old like the Egyptians did. The priest then relates how the story of Atlantis had been passed down to him from a time memorial 12000 years past. The priest then spins the story to even further absurdity. Read for yourself here:


  5. Google altered their seafloor rendering some time ago, making this article moot.

  6. https://www.google.com.br/maps/preview/@27.4150608,34.6246581,72792m/data=!3m1!1e3

  7. City and a pyramid https://www.google.com.br/maps/preview/@51.3173279,-18.7509199,102503m/data=!3m1!1e3

  8. https://www.google.com.br/maps/preview/@11.3582111,45.1509337,160791m/data=!3m1!1e3

  9. https://www.google.com.br/maps/preview/@12.8302996,52.21334,159908m/data=!3m1!1e3

  10. https://www.google.com.br/maps/preview/@9.4283314,51.9066195,161788m/data=!3m1!1e3

  11. https://www.google.com.br/maps/preview/@7.1578771,53.7492188,162725m/data=!3m1!1e3

  12. City? https://www.google.com.br/maps/preview/@22.2710746,61.4863404,75884m/data=!3m1!1e3

  13. City? https://www.google.com.br/maps/preview/@-27.6577503,73.6411214,145264m/data=!3m1!1e3

  14. https://www.google.com.br/maps/preview/@-24.6564045,13.5957192,149050m/data=!3m1!1e3

  15. https://www.google.com.br/maps/preview/@22.4682793,-18.7659847,151554m/data=!3m1!1e3

  16. https://www.google.com.br/maps/preview/@27.4243218,-18.5462581,145573m/data=!3m1!1e3

  17. https://www.google.com.br/maps/preview/@29.4600512,-18.3155452,142797m/data=!3m1!1e3

  18. Okay, you have made some kind of point. Kindly explain.

  19. What am I supposed to be seeing here, I do not see any roads. I must be blind! I guess all these darker lines are what they speak of as roads.