Hill posted a masterful explanation for how the model of the bottom of the oceans are created and portrayed in Google Earth Ocean (GEO) in THIS thread, as well as a specific discussion of the Madeira Plain Survey which has led to (or, rather, resulted from) a barrage of "Atlantis" discoveries, addressed in THIS thread.

I've recently spent a fair amount of time looking closely at the seafloor model in Google Earth, with the purpose of helping users better understand what they are looking at. The only additional contribution I can make to Hill's two threads mentioned above is that I have found a collection of .kml files HERE which are the recorded tracks of research vessel sonar surveys. Many of these tracks correlate precisely with some features in the GEO seafloor model. I think it would be helpful for me to show you visually, using Google Earth, exactly how the seafloor model is assembled and displayed.

Before I get to specific visualizations, it is worthwhile to remember one important concept: We have better maps of Mars, the Moon, and other planets and moons, than we do of the surface of the seafloor on our own planet. I keep using the term "seafloor model" for the representation in Google Earth because that is what it is. It is not a collection of images of the seafloor, rather, it is satellite radar geodesy (a coarse model of the seafloor based on radar measurements of sea height) corrected by actual single- or multi-beam sonar recorded by research vessels (much finer detail, much higher quality depth information). Neither of these two sources are "images" in any sense of that word, they are both collections of mathematical data points. Researchers assemble these data into a mathematical model, an artificial surface, which is artificially color-coded to show depth, and artificially shaded to show slope. The researchers then either give, lease, or sell this model to Google.

On to the Google Earth visualizations:

The two images below are screenshots of a small area of the GEO seafloor model just East of Easter Island, about halfway between Peru and Tahiti in the South Pacific. Easter Island itself is visible on the left side of the screenshot. In the first screenshot, you can see the seafloor model has a generally "wrinkly" appearance, although some of the wrinkles appear shiny and smooth, and some of the wrinkles appear rough and coarse. Also noticeable are a few straight sections or lines which appear rough and coarse. At the top of the screenshot, I've added a line 100 miles long using the GE Ruler tool, to provide scale.



In the second screenshot, I've added the ship's track from the research vessel "Melville", owned and operated by the Scripps Institute of Oceanography, with a link HERE. The track (see the .kmz file below) is of a voyage between 3/23/93 and 4/26/93, and the Chief Scientist responsible for gathering the data is David Naar. I've added the .kmz of the track at the bottom of this post, in a file called GLOR06MV.kmz, zoomed in to this location, but it might be worth your while to zoom out and follow the ship's track to the East, where the track ends in Santiago, Chile. Sometimes while I was working on this, it was helpful for me to toggle the .kmz file off and on and off and on, to help me verify the correlation between the ship's track and what I was seeing on the GEO seafloor model.


You can clearly see, I hope, the correlation between the ship's track, and some of the "rough and coarse" areas in the GEO seafloor model. Other lines and areas of "rough and coarse" seafloor model are from other ship's tracks, which I have not included here. The smoother, almost shiny areas between the ship's tracks are the satellite's "best guess" of what the seafloor might look like. The "rough and coarse" pieces are the areas which have been accurately measured for depth by sonar. In essence, the satellite model of the seafloor is corrected by on-site measurements of the depth by sonar, which a very accurate measurement.

More to follow.

Continuing the discussion, I've got the same area as shown previously, but I've added three placemarks (the updated file is attached below as GLOR06MV2, you might want to download it to follow along).

The first placemark is an arrow (called "Large area of SRTM satellite sea height data") showing specifically the "smoother" area of the GEO seafloor model. This is the area of satellite radar sea height measurements, and it's not very accurate, which makes the GEO seafloor model seem smoother than the actual seafloor really is.

The second placemark is an arrow (called "Ship's track from another voyage") showing a nice linear "rough and coarse" feature of the GEO seafloor model. This is a place where a ship accurately measured the depth of the ocean along the line of it's transit. (I may or may not have that actual track in another .kmz, it doesn't matter for this discussion, you can see the same thing along the "Melville" track further to the East). It appears "rough and coarse" relative to the SRTM satellite geodesy, because it's much more accurate. I guess an analogy would be of a very person with very bad eyesight looking at a newspaper, seeing mostly a soft fuzzy blur, then putting his or her prescription glasses on. The newsprint would snap into crisp, sharp detail. That's what we're seeing with this linear feature, except the analogy is not a visual one, remember, we're not actually looking at the seafloor, we're looking at an artificial model of sonar and radar data.

The third placemark is an arrow (called "Inflection Point") of the Melville's track. It is one example of many along this ship's track which I looked for to confirm that I had a one-to-one correspondence between the .kml of the ship's track and the features of the GEO seafloor model. I looked for areas with mostly SRTM satellite geodesy, crossed by an angular ship's track. It's easy to assume a line is a line is a line, and that any linear ship's track which matched up with the GEO seafloor model is a correspondence. But in reality, there really is only one great circle line between, say, San Diego and Hawaii, so a linear ship's track which followed linear features in the GEO seafloor model may or may not be a correspondence. But if I could find (and this is what I looked for) an instance of a linear feature changing direction, with no ambiguous or confused overlay from other ship's tracks, then I knew I had an exclusive correspondence. I call this an "Inflection Point", where a ship changes direction, and the GEO seafloor model matches this change in direction.

My next post for this series is the following collection of ship's tracks I extracted from the huge collection from the Marine Geoscience Data Systems website HERE. I've called it "Dead Bang Evidence", because in each and every one of these ship's tracks, I've found inflection points or grids, or some other irrefutable correspondence between the ship's tracks and the GEO seafloor model.

The file is a collection of ship tracks in a folder for each separate ship. I'd suggest turning on one ship at a time, or even one voyage from one ship, and examining one track at a time by clicking the track off and on and off and on. Within each ship folder is a collection of tracks, and if you double-click on the track name, GE will take you to the area which includes that track. You then need to zoom in to an area of interest along that track, again, I'd suggest you look at grids and changes in direction, but also the places where the tracks encounter continental shelves. I've found an eye height of between 250-300 miles (look in the lower right corner of your GE viewer) works best, although sometimes a gridded area can best be seen a little higher, up to 400 or 500 miles. It works the other way, too: if you "turn on" a ship's folder, and with the folder open, you click on a track of interest somewhere on the GE globe, a balloon will pop up showing you the metadata for that track and the track itself will be highlighted in the list of tracks in the open folder. You can then toggle that track off and on.

Check out the gridded areas, look at angles in a straight ship track where it makes a turn, and also look near seaports. You can see sonar ship tracks radiating away from most seaports, with Perth, Australia, and Cape Town, South Africa representing excellent examples of this. The radial features there are not actually present on the sea floor, they are artifacts of the sonar ship tracks.

I learned a few other things about the GEO seafloor model: there are many very well studied areas in the oceans, which correspond to the Niagara Falls, Eiffel Tower, Giza Plateau etc. on land. Oceanographers like to study ridges and rift zones, and any active underwater volcanic structures, which usually correspond with ridges. So many sea ridges are well surveyed. Obviously any place with military significance is well surveyed. The seafloor between Greenland, Iceland, and the United Kingdom is one of the most surveyed ocean seafloor areas on the planet, as a hangover from the Cold War in the 1980's. Hawaii is mapped with extreme precision, not only because of any military significance, but also because of the active volcanic processes and other geological interests (think tectonics). Any area with potential subsurface oil fields is also well surveyed: check out the Northern Gulf of Mexico south of Texas and Louisiana, but also off Indonesia.

Some areas simply have bad data, for whatever reason. East of Florida is a large area that is probably much flatter than the ocean bottom really is. There are large gridded areas on the margins of the continental shelf all along the Eastern seaboard of the United States, as well as off the West coast of Ireland. I suspect both of these have to do with poor processing of the SRTM data, but I don't know that for a fact.

Also, there's a minimum depth at which the ship tracks "die off". I suspect this has to do with the fact that at shallow depths of the continental shelves (less than 1,000 feet or so) there probably isn't much difference between the satellite sea height and actual depth, plus the surface of the sea bottom on the continental shelves is by and large very flat, anyway, due to siltation and sedimentation.

As a final example of how little we know about the seafloor on planet Earth, look at the image below. It's taken off the Mendocino Coast in Northern California. On the right hand side is extremely fine detail bathymetry characteristic of most of the west coast of the U.S. This is probably as close as we can get to actually "seeing" the seafloor, although the "resolution" is several hundred feet, at least. And on the left side of the image is a large area of what is obviously poor quality data. Near the center of the image is a smooth linear feature running slightly diagonally SSW-SNE. Using the Ruler tool, this is about 4 miles wide by about 40 miles long. Essentially, there's at least 160 square miles of seafloor which is a complete blank, at least in the Google and Scripps Institute of Oceanography bathymetry, sitting less than 70 miles offshore. How many thousands of square miles of the rest of the Earth's seafloor is complete terra incognita? No wonder the maps which the ancient mariners used said, "Here thar be Sea Monsters".



Anyway, there's lots to be discovered in Google Earth Ocean. I had fun working on this little project, and learned quite a bit about bathymetry and the Google Earth Ocean seafloor model in the process.

Good luck, and have fun!

Here is a screenshot of the GEO seafloor model off the coast of Cape Town, South Africa. At the top is a 100 mile line I added using the Ruler tool, for scale. About 50 or 60 miles West of Cape Town is a sea canyon which extends Northward for perhaps 80 miles.

All of the other lines radiating away from Cape Town are artifacts, errors in the model which correspond to ship tracks from the research vessels which left (or arrived at) Cape Town, recording sonar data as they traveled which is replicated in the GEO seafloor model.

The Madeira Abyssal Plain contains a grid of ship tracks which Hill found the source for, and explained it in THIS thread.

In less than a half hour's search, I've found more than a dozen similar grids. These are common artifacts of the way that sonar data is taken, and reflect nothing more mysterious or supernatural than that. These are in fact human creations, but they are not physical objects sitting on the bottom of the ocean. They are human creations only in the sense that they are cyber-structures, features which exist in the seafloor model database only. They don't exist in any real way, if you could somehow get down to the bottom of the ocean in these places, you would see nothing at all other than a flat expanse of seafloor.

It's important to remember that when you look at the bottom of the ocean in Google Earth, you're not actually looking at the bottom of the ocean. What you're looking at is a very sophisticated model of human construction, with human errors in the way the model was put together.

As an example of the difference between an image of the seafloor and a model of the seafloor, I've taken a screenshot of the North Pacific, just South of the Kamchatka Peninsula and the Aleutian Archipelago. I picked this area because of the high latitude, the seamounts, and the "shadows" in the model adjacent to the seamounts.

If you notice, North is up, and the latitude at the center of this place is about 48 degrees North (look at the bottom of the screenshot below). The "shadows" near the seamounts all well to the South or Southeast of the seamounts that are in this area. It looks as though the Sun was illuminating this scene from somewhere to the Northwest. It should be readily apparent that it's not possible for the Sun to cast shadows towards the Southeast anywhere near 48 degrees North latitude. If you browse around the globe, you'll see that everywhere the "shadows" are uniformly the same, with the Sun to the Northwest of every place, and also as though the Sun was 45 degrees up in the sky at every place.

What this is, is a tool model builders use to indicate slope. By creating artificial shadows, called a "hillshade", a Digital Elevation Model can show slope, or at least trick your mind into thinking you're seeing slope. The steeper the slope, the larger and darker the shadow. Where there are no shadows applied to the model, your mind is tricked into seeing flatness.

Don't be confused by a sophisticated model into thinking you're seeing the seafloor. You're not; you're seeing a model of the seafloor.