Talk:Sandy Island, New Caledonia/Archive 1

Maps
The maps show the fictitious island coinciding with a local rise of the continental shelf, this indicates that its actually a feature of the seabed and not a literal island.

The Landsdowne Bank is meant to have a feature called Nereus Reef in its very north, and the sea bed depth information shows the Sandy Island area marked is the shallowest water in that area.

The area marked appears to be the area that Nereus Reef may be.

The Chesterfield Islands and the New Caledonias north does have islands called Ilot de Sable, which would translate to Sandy Island. The name may have been confused by human error.

The Chesterfield Islands area does have mystery sand islands that can no longer be found since first reported.

Google maps ,etc, show blackness in the area of the island.. in fact they also show ocean inside the island's boundaries too, eg google earth has the 'historic' view, which shows ocean over half. 60.242.170.18 (talk) 10:49, 22 November 2012 (UTC)

This is complete nonsense. There is no "continental shelf" in the area in question. The water depth there is 1400 m -- there is no possibility that there ever was a "mystery sand island" there. RandomCritic (talk) 17:29, 22 November 2012 (UTC)


 * I’m looking at GTOPO30 bathymetry. There IS a seamount there and it appears to reach to or just below sea level in the same space as the outline of the purported island. I don’t know where this notion of deep water comes from. Indeed the bathymetry suggest an easy explanation for the island: vectors were digitized automatically from bathymetry data. I see the World Vector Shoreline and World Databank II both show the island. Strebe (talk) 19:58, 22 November 2012 (UTC)


 * I think it's coming from the RV Southern Surveyor ss2012_v06 "Tectonic framework for the easternmost Coral Sea and northern extent of the Lord Howe hotspot" research voyage that recently passed nearby. The program is here but I don't think they have published an cruise report yet.  Sean.hoyland  - talk 20:28, 22 November 2012 (UTC)


 * The more I muse over this, the more I am convinced we’re talking about at least two different locations and possibly three. The digital data I look at all agree the water is VERY shallow or there is even an island (at low tide?) where Google Maps says “Sandy Island”. These data are:
 * World Vector Shoreline (vector data source from NOAA)
 * World Databank II (vector data source from US State Department)
 * NASA Blue Marble imagery (satellite imagery)
 * GTOPO30 bathymetry (USGS 30 arcsecond DEM data)


 * Unless we imagine some massive pan-agency, long-running conspiracy in the US government concerning a region of no interest to the country, we have to consider each of these sources to be independent. Yet we have an expedition claiming to have visited the area and finding 1400m waters. 1400m! We have an interesting story from year 2000 about amateur radio enthusiasts wanting the area as a new entity on their DXCC list, and determining that the alleged Sandy Island does not exist.


 * How could a “scientific expedition” sail to the “same” place and find something so utterly different? Because they didn’t go to the same place, of course. How could they not have gone to the same place, given GPS, their modern navigational equipment, blah blah?


 * Not sure. I don’t think a mismatched datum could explain it; there isn’t deep water close enough to the seamount for a mere datum mismatch to explain it. What’s clear is that things are not adding up in this story.
 * Strebe (talk) 20:52, 22 November 2012 (UTC)


 * 3DGBR also shows shallow water there.  Until the research results are published, I think the "1400m" quotes should be taken as an invention of the press. AndrewBolt (talk)  —Preceding undated comment added 09:27, 23 November 2012 (UTC)


 * The sources cited, all of which are RS, are directly quoting the chief scientist for the voyage, Maria Seton, a geologist from the University of Sydney and Steven Micklethwaite, a structural geologist from the University of Western Australia, who was also onboard, so I don't think there is any policy based reason for Wikipedia to treat the figures as an invention of the press.  Sean.hoyland  - talk 11:21, 23 November 2012 (UTC)


 * I look at GE regularly in this area and the island outline has been there a only few weeks. Chesterfield Iss are atop the Bellona Plt. GE regularly get tiles misplaced, nothing new there. The outline is 25km long so I'd be well aware of such a large Is, and would have looked for it on GE many years ago. 1993 was the last time the Southern Surveyor was used for useful work. Projects that may have a commercial outcome are given priority on the CSIRO ship(s). Geology surveys may lead to seabed harvesting of minerals. 220.244.239.240 (talk) 02:24, 23 November 2012 (UTC)
 * "1993 was the last time the Southern Surveyor was used for useful work."...I don't think so. The list of planned and past RV Southern Surveyor voyages going back to 2003 are here. They are all useful from a scientific perspective and have a go at finding the ones "that may have a commercial outcome" or are related to "seabed harvesting of minerals".  Sean.hoyland  - talk 07:30, 23 November 2012 (UTC)

Last useful trip was about 96. Bios and geos using a Nansen bottle rosette to collect samples, is like using a micrometer as a G-clamp. The CDTO units were removed years ago. The mud grab is more their style! The poll tax on reef tourists has only ever been a commercial outcome ... not one cent has ever been put towards marine research, as was promised many years ago. This Coral Sea marine park will be the same thing. Keep an eye on the tonnage charges for bulk carriers in transit. It will not be announced publicly, since it will simply be a regulation. 220.244.239.240 (talk) 09:14, 23 November 2012 (UTC)

Link/pic
How about a link to Google Maps?

Or a screenshot (in case Google removes the island)?

CapnZapp (talk) 18:49, 22 November 2012 (UTC)


 * Google Maps is accessible via GeoHack from the coordinates at the top of the article so I'm not sure a link is worth it.  Sean.hoyland  - talk 18:56, 22 November 2012 (UTC)
 * Since Google will consider this an error and most likely remove this island any time soon, I fail to see the worth of a link. The location will probably just show blue water in the future. Thx CapnZapp (talk) 19:01, 22 November 2012 (UTC)

How could something show up in the GoogleEarth satellite photos if there is nothing there? If there were nothing there why put something in and then paint over it? Only a major military secret would get covered up so thoroughly. — Preceding unsigned comment added by 120.151.215.254 (talk) 23:05, 22 November 2012 (UTC)

Google Sandy
Here's an example of what we could use as illo. Feel free to tweak (make smaller) etc CapnZapp (talk) 18:59, 22 November 2012 (UTC) File:Non-existant Sandy Island Google Maps.png

Old maps
I have placed a link to Maps of Sandy Island Through History, showing how it has (or hasn't) appeared in maps since 1797. Whereisthygodnow (talk) 20:54, 22 November 2012 (UTC)
 * A very nice resource; thanks for linking it. I'll just take this opportunity to mention that this statement in the revision history is incorrect:
 * 16:03, November 22, 2012‎ 194.203.153.1 (talk)‎ . . (3,006 bytes) (-12)‎ . . (Removed hyperlink for 'Times Atlas of the World' as the Comprehensive Edition of this atlas does not contain reference to the island in question. Some other editions do contain the 'Ile de Sable'.)
 * I own the latest Comprehensive Edition of the atlas and I assure you "Ile de Sable" is there, on inset map 10 on page 4. Languagehat (talk) 15:42, 23 November 2012 (UTC)
 * My 1988 edition of 'Times Atlas of the World' shows two small islands labelled 'Sable'. Dsergeant (talk) 16:13, 23 November 2012 (UTC)

This series of old maps interestingly includes one produced following Cook's second voyage:

http://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~24064~870108:Chart-of-discoveries-made-in-the-So?sort=Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No&qvq=q:new%2Bcaledonia;sort:Pub_List_No_InitialSort%2CPub_Date%2CPub_List_No%2CSeries_No;lc:RUMSEY~8~1&mi=16&trs=32

This shows a "Sandy I." to the north of New Calendia. The chart also usefully shows a plot of Resolution's voyage, and the dates of each leg of the voyage. The 14 September 1774 date shown by "Sandy I." can be cross-referenced to "Captain Cook's Journal 1772-75", where he writes that "at three o'clock we passed a low sandy isle, lying on the outer edge of the reef, in latitude 19 25', and in the direction of N.E. from the north-westernmost land, six or seven leagues distant." KPOK (talk) 08:59, 24 November 2012 (UTC)

Just a big misunderstanding
The news reports that culminated in the creation of this article are farcical. Wherever the RV Southern Surveyor crew went or thought they went, it wasn’t the same place shown by Google Maps as “Sandy Island” at (19°13´S, 159°56´E). I have looked at vector data sources (World Vector Shoreline, World Databank II). I have looked at satellite imagery (NASA Blue Marble). I have looked at bathymetry data (SRTM30 Plus, with bathymetry consisting of the Smith & Sandwell global grid, the DLEO Ridge Multibeam Synthesis Project, the JAMSTEC Data Site for Research Cruises, and the NGDC Coastal Relief Model). ALL of them agree there is a 30 km N/S by 6 km E/W sea-level sand-bar at that location. Satellite imagery is completely independent of the other sources, and so are the bathymetry. Whether you want to call the thing an “island” or not is a matter of taste, but you’re not going to sail over it in a deep draft ship at any tide and you’re certainly not going to sound it at 1,400m. This is all just some goofy misunderstanding. Strebe (talk) 21:49, 23 November 2012 (UTC)


 * I'm still waiting for User:Strebe to provide the data that will conclusively prove that this is "just a big misunderstanding". How much longer should I expect to have to wait? RandomCritic (talk) 17:41, 13 January 2013 (UTC)


 * I know it’s a shock, but the world isn’t arranged around your expectations. Hopefully you’re holding your breath while you wait. Strebe (talk) 22:25, 13 January 2013 (UTC)
 * Strebe is working hard to find it.--Milowent • hasspoken 04:18, 14 January 2013 (UTC)


 * That is not satellite imagery, as satellites do not see beyond shallow water. (and ASTER uses thermal which doesn't see into water at all)  You see shaded relief model bathymetry data, likely processed using interpolation, etc.  I looked at the source DTED0 (military) data for this location and there are like 4 pixels... that's a lot of Kriging, eh?  Anyway, I've ordered a full Landsat product from the USGS to use in this article; this takes a few days.


 * In the meantime, this is the browse image of what the satellite imagery of the area truly looks like... not just half like on Google Earth historic imagery. There is nothing there, at least near or above the surface.  (there are other images on GloVis from Landsat, ASTER, and EO-1 ALI all showing this.)


 * File:Glovis.usgs.gov - Sandy Island browse image, Landsat + water boundaries.png (for some reason the thumbnails are broke, click view full size if interested)


 * Разрывные (talk) 13:22, 24 November 2012 (UTC)


 * The image you give doesn’t yield any useful information. It is consistent with shallow or deep water either one. I spent quite awhile browsing various data sources in USGS’s Global Visualization Viewer. It’s surprising how sparse data is for the area, and it’s also interesting to note that the area is almost always cloudy. However, the MODIS imagery clearly shows our Sandy Island, but it also clearly shows that it is submerged sometimes. Here is a typical image:
 * [[File:MODIS-MCD43A4-02-2004.png]].


 * The red circle is Sandy Island. The large mass to the right is the shallow sea north of New Caledonia, whose mainland is not visible in the image. Be careful of the orientation: You are looking at the right edge of a sinusoidal projection, so north-south is tipped very much to a diagonal. Sandy Island itself is oriented almost perfectly north-south, but shows up at a 45° angle on the image because of the projection. On GLOVIS, you can cycle through the months and years of MODIS combined data, and you will see that the island is sometimes visible and sometimes not. When it is not visible, you can assume it is submerged enough that it does not yield any reflectivity data for the satellite. Strebe (talk) 22:51, 24 November 2012 (UTC)


 * Interesting, though I think you're making an error of assumption with the low-resolution browse image and the nearby reefs/atolls (of whose name(s) I don't know offhand) to the west of the proposed location of Sandy Island. Can you show a georeferenced WGS84 lat/lon for the projected feature in question?  You're right that MODIS data has a distorted sinusoidal projection on GloVis-- Landsat does not, though.


 * You're right that my image doesn't show Sandy Island. That's entirely my point... and I'm aware that the low-resolution browse image on GloVis isn't particularly impressive, which is why I ordered and am awaiting the full level 1 product.


 * The image of Johnston Atoll I made was when the sun was at a decidedly non-nadir angle (non-optimal for depth) and the illumination of the atoll looks somewhat different, but no matter the conditions of the sea it is always visible to some extent. Theoretically, I'll grant you something at the very deepest extent of what it is visible underwater from the surface could have its visibility vary with the position of the sun but the combined Landsat, ASTER, and EO-1 coverage of this area had those optimal (nadir) solar exposures among them.  So how could this be missing from every Landsat, ASTER, and EO-1 image?


 * MODIS on GloVis is not that great. The actual daily reflectance granules available through NASA are much higher resolution and are georeferenced though the NASA interface is much more awkward than GloVis.  If you really think that shows Sandy Island, you should grab the granule id and download the full HDF (or whatever multispectral format it is).


 * I believe what you will find is the coordinates for the formation you indicate as Sandy Island are actually to the west of it, and the projection and low resolution in the browse image made it difficult to tell. You can export the multispectral image to a GeoTIFF and get an exact lat/lon using ArcGIS Explorer Desktop or similar program.  The multispectral data would also provide possible evidence for variable visibility.  Visible spectrum vs. SWIR would differentiate surface vs. underwater features.


 * (Compare the MODIS browse images you see on GloVis to the processed ones here ... you should be able to get much better detail and resolution.)


 * On the other hand, I certainly could be wrong, in which case I don't want to dissuade evidence of the island's existence from coming to light. If you really think you're on to something, by all means pursue it further.  You could end up debating a cute Australian scientist on CNN or something.  Or at least make for a very interesting Wikipedia article.  Разрывные (talk) 00:16, 25 November 2012 (UTC)


 * I think you're making an error of assumption with the low-resolution browse image and the nearby reefs/atolls (of whose name(s) I don't know offhand) to the west of the proposed location of Sandy Island. Can you show a georeferenced WGS84 lat/lon for the projected feature in question? No, I am not going to do that. Just go to GLOVIS to the tile in question, move your mouse pointer around, and let the GLOVIS application tell you where your mouse pointer is. The low resolution is a red herring; the sand bar is so huge that neither the resolution nor georeferencing matter. It’s the right place, and you can convince yourself that the coordinates returned by the GLOVIS application are correct by comparing against known points of the reefs north of New Caledonia.


 * So how could this be missing from every Landsat, ASTER, and EO-1 image? It’s not. Or, rather, most of the time the area isn’t even available, and the few times that it is, the images are too cloudy to be useful. I can see something of interest in one or two of them, but in general they’re just not useful data in this case. In order for anything to be visible, the images need to be relatively free of clouds AND the tides have to be low. If you examine the bathymetric data surrounding the purported island, you will find the depths drop off very slowly, which accounts for the lack of any well-defined “sighting” in those few images that show open water.


 * The data are sparse for two reasons: Constant cloudiness interferes with most remote sensing, and ships can’t go there, so there is no bathymetric data. IT’S TOO SHALLOW. Meanwhile all around the edges of the alleged island, the water from real (not interpolated, no kriging) bathymetry data are shallow and slope downward. For remote sensing, the MODIS data are reliable, and sure enough they show shallow water or an island there. The thing is clearly a seamount that breaches or nearly breaches the surface. All the data show this. It‘s nothing like 1,400 m depths.


 * So. You can suppose two hundred years of maps and charts from innumerable surveys and completely independent data sources and technologies are wrong… or you can suppose a single research vessel with a crew of unknown navigational prowess combined with lay news reporting have resulted in a big misunderstanding. I know where I’m putting my money. Strebe (talk) 01:03, 25 November 2012 (UTC)


 * I guess it's not entirely accurate to say that satellites do not see beyond shallow water given that seafloor topography can be predicted from the variations in the gravity field derived from the satellite altimetry data. We probably have an article about it somewhere...  Sean.hoyland  - talk 13:44, 24 November 2012 (UTC)


 * By "see" I mean detect and image electromagnetic radiation, as it seemed there was some confusion the sonar-based bathymetric data products being shown under "Satellite" view on Google Earth or Maps were satellite imagery -- the previous poster had cited satellite imagery as evidence for the island's existence, and I wanted to clarify that it was sonar data, not satellite. Even though it's shown on "Satellite" view in Google Earth... it's not satellite.


 * I'm not trying to be pedantic about this either, because the what the actual satellite imagery shows is no evidence of the island at all. It's kind of frustrating and counter-intuitive-- you click on "Satellite" view to see satellite imagery, and for Sandy Island, you don't get any.


 * I think what you might be referring to is the GRACE satellites which work off measuring changes in distance between the twin satellites. Unfortunately, the resolution of that particular data seems much too low to shed light either way on Sandy Island.  SAR doesn't seem to be much of a replacement for sonar, either.


 * At any rate, visible spectrum imagery will indeed show formations on or near the surface of the water; in this case if Sandy Island were anywhere near the surface we would see it.


 * Here is an example of imagery I processed for Johnston Atoll, which well illustrates surrounding underwater formations. Compare that actual image to what you get in Google Maps searching for Johnston Atoll... or rather, what you don't get.


 * Visible green also matters. Consider this true color image of Bikini Atoll, compared to the false color IR version that replaces red and green color channel data with two SWIR bands (keeping visible blue).  By selecting which bands to use you can easily differentiate surface and near-surface features, and blue/green can be used to discern relative depth to some extent.


 * My point is that satellite (or aerial) imagery does have limited water visibility (in blue/green bands) and that in no case does it serve as evidence to support the theory of Sandy Island being a surface or near-surface formation such as an island, sand bar, atoll, or reef, as those would be visible in satellite imagery. Unless of course, Ben Linus moved the island or something. Разрывные (talk) 23:23, 24 November 2012 (UTC)


 * Sorry, to clarify, for interest, I didn't mean GRACE, I meant the method used [ here] and described in more detail here. Here is a white paper from 2001, "Bathymetry from Space: White paper in support of a high-resolution, ocean altimeter mission", and the data itself (including a .kmz file) is available from Scripps here.  Sean.hoyland  - talk 16:55, 25 November 2012 (UTC)


 * Interesting -- that link shows correlation between the SRTM data and their gravity anomaly data for the location. Разрывные (talk) 20:37, 27 November 2012 (UTC)


 * as it seemed there was some confusion the sonar-based bathymetric data products being shown under "Satellite" view on Google Earth or Maps were satellite imagery -- the previous poster had cited satellite imagery as evidence for the island's existence. No, the confusion is your reading Google Earth into it. I stated the source is NASA Blue Marble imagery. The NASA imagery clearly shows shallow water in the shape of the “island”. Here it is.
 * [[File:NASA_Blue_Marble_Sandy_Island.PNG]]


 * I have changed the gamma to make it bloody obvious. The size is exactly as shown in all the maps. The uniform areas are deep water. The stuff to the west is the Chesterfields. It’s not deep water. Clearly. Strebe (talk) 01:05, 25 November 2012 (UTC)


 * I haven't verified the image above, but assuming it to be in good faith, it still doesn't show what Google Earth suggests exists there: a large, continuous, sand-covered island occupying the entirety of the shallow-water zone. At best one might expect to find a few segments of reef protruding over the surface. On consulting the first-generation news reports about the Southern Surveyor's observations, it doesn't actually say that they sounded to a depth of 1400 m, but merely that their charts showed that depth there -- so if there's a big mistake, it's by the compilers of the nautical chart, not by the Southern Surveyor crew. (Apparently this got confused in later reports.) So I see no difficulty in believing that the Southern Surveyor sailed to the correct spot, found nothing above water, and moved on. And even if there is shallow water in the spot, it's still not part of a "continental shelf", as the earlier draft of this article claimed. It also remains unexplained when and why Google smudged over the area with black pixels. RandomCritic (talk) 17:03, 25 November 2012 (UTC)


 * I haven't verified the image above, but assuming it to be in good faith. Sorry. I don’t have time to waste on lazy, not-even-armchair-criticasts who can’t even be bothered to look up some data from the convenience of their keyboard before blathering their ugliness. And to the broader audience, no, the error is that the crew wasn’t where they say they were and weren’t even close enough to tell what exists at Sandy Island. Here is a video which includes a diagram of the ship’s track vis a vis the island: . And note the crew member who states, “The ocean floor actually didn’t ever get shallower than 1,300 below the wave base…” That’s not a statement of what the charts say; that’s a statement of their own measurement. There isn’t anything wrong with the charts. There isn’t anything wrong with the maps. There isn’t anything wrong with Google Earth. The only thing wrong here is the grand, exciting narrative that… isn’t. It’s not the island that’s phantom. It’s the story. Strebe (talk) 19:11, 25 November 2012 (UTC)


 * For someone who has no "time to waste", User:Strebe spent a great deal of time on this. The fact is that I made an effort to find out where User:Strebe got that image from and was unable to do so. "NASA Blue Marble" was not a terribly helpful hint; I could not find any images at the resolution shown, probably because I'm not familiar with the site and their navigation tools are underwhelming. That's why I wrote "assuming good faith". As for the video, if one listens to the audio, one will hear them say that the supposed location of "Sandy Island" was not far off their track (as shown), and so they altered their course to sail through it. There's no inconsistency. User:Strebe's conclusion that the story "is phantom" is still not supported. And there still remain large questions about the imagery problems, which User:Strebe has no answer for. Why is the area blacked out on Google Earth? Perhaps instead of limiting himself to Talk messages on Wikipedia Mr. Strebe could approach some news organizations with his theory and see if they are interested in exploring it. As it is, nobody other than Strebe has challenged the veracity of the Southern Surveyor crew's account. RandomCritic (talk) 20:28, 25 November 2012 (UTC)


 * I could not find any images at the resolution shown, probably because I'm not familiar with the site and their navigation tools are underwhelming. Doesn’t that suggest you should go concentrate on something you do know and understand instead of raising sophomoric objections? I already answered the imagery problems, but you’re too busy imagining you have something to say, to listen. The hole is there because you CAN’T TAKE SOUNDINGS IF THE SHIP CAN’T GO THERE. Meanwhile the island doesn’t show up on satellite imagery except possibly at low tide… if even then. The SRTM30 Plus bathymetry data shows constant 1 m elevations across the entire 150 km² region of the “island”. Those are interpolation (kriging) artifacts from the soundings they DO have around it… which, by the way, clearly shows a seamount reaching or nearly reaching the surface. Strebe (talk) 20:48, 25 November 2012 (UTC)
 * "Doesn’t that suggest you should go concentrate on something you do know and understand...?" No, it suggests that if someone really wants a source he or she cites to be verified, he or she provides appropriate citations and/or links. If those aren't provided, then the source remains valueless for the purposes of Wikipedia. "The hole is there because you CAN’T TAKE SOUNDINGS IF THE SHIP CAN’T GO THERE." If that statement is in response to my questioning Google Earth's smudging of its satellite image with black pixels, then it is utterly absurd. The satellite image is of course not based on soundings. What Google thinks it is doing is beyond my ken, but it does underscore the fact that Google's map products are so altered and redacted, in response to various pressures, as to be scientifically useless. As User:Strebe has ultimately admitted that there is "no island there in the obvious sense", all of this sound and fury seems to be misplaced.
 * Ultimately, I think that the following conclusions can be drawn with good probability:
 * In the absence of other, contradictory first-person data from the region, the Southern Surveyor crew's observations are controlling. They were there; we weren't. Until something else shows they were wrong, we have to take their word.
 * There is no good reason to assume that the Southern Surveyor did not visit the area where "Sandy Island" was supposed to be. In the era of GPS navigation, a navigational error seems to be the least likely possibility; if there was a problem with their GPS, they would have noticed it by the time they reached their destination.
 * There may be a local elevation of the sea floor in the region, in which case Micklethwaite's claim that the floor never rose above a depth of 1300m must be incorrect.
 * The Southern Surveyor's observations aside, the satellite imagery does not clearly show any part of the elevation breaking the surface. Therefore Google Maps' representation of this part of the world as a single large, integral, oval island is certainly incorrect.
 * Google Earth isn't doing anything to enhance their credibility by altering their image ex post facto.RandomCritic (talk) 13:59, 28 November 2012 (UTC)


 * “If those aren't provided, then the source remains valueless for the purposes of Wikipedia.” In case you missed Sean.hoyland ’s posting immediately below, this entire conversation is valueless for the purposes of Wikipedia. ‘As User:Strebe has ultimately admitted that there is "no island there in the obvious sense", all of this sound and fury seems to be misplaced.’ The sound and the fury is over your uninformed, officious pronouncements. It has nothing to do with the status of the island. Your new set of assertions has not improved; you contradict yourself any number of ways. If you want to discuss it, take it to one of our own Talk pages. It doesn’t belong here. Strebe (talk) 16:38, 29 November 2012 (UTC)


 * Remember, even if everything you say were entirely correct, everyone here is required to not care because we aren't reliable sources. The talk page is for improving the article contents based on published reliable sources. The science ref desk, or a blog, is a better place for this kind of thing.  Sean.hoyland  - talk 22:22, 25 November 2012 (UTC)


 * That’s not quite true. What this article reports is that the island does not exist. That’s never been demonstrated to the flimsiest of scientific standards. All you have is a research vessel from a field unrelated to marine bathymetry that claimed to the news media to have sailed there and found nothing. The news media, loving a good story, dutifully reported it. There is no scientific paper. There is no peer review. There is nothing but some breathless announcement to the media. If you listen to the breathless announcement, it dawns on you the crew isn’t terribly knowledgeable about what they’re saying. They talk about the “World Vector Coastline”, but it’s “World Vector Shoreline”. They speculate that all these alleged independent sources must all go back to this “World Vector Coastline”, but they seem unaware that the island has appeared on maps for over 200 years. Yet this article reports their breathless announcement as fact. Meanwhile we have two hundred years of charts and maps that DO show the island, including modern data sources independent of anything from two hundred years ago. We have scientists around the world scratching their heads trying to figure out what’s going on (see this thread of oceanographers amongst many others). It’s clear there’s no island there in the obvious sense, but it’s also clear it’s not deep water. At the very least this article needs to quit just parroting unreliable sources as fact and instead report what the news media is saying, leaving the “factness” open to question. Strebe (talk) 02:33, 26 November 2012 (UTC)


 * I have (finally) added the highest quality Landsat image of the area from GloVis I could find to the page which has much less cloud cover than the Digital Globe imagery. (apologies for the downsize and crop, but it couldn't be helped, the original was 15801x14101 px)  From the reference Chesterfield Islands to the west, the greatest visible depth is approximately 35m.  This does not "disprove" Sandy Island per se, but it does provide evidence there is nothing in this extent to approximately 35m in depth.  Note that Landsat imagery of this area goes back as far as 1978; the older imagery showed the same results.  It is odd that multiple sources of bathymetric data indicate otherwise.


 * I am not sure if I should mention this, but ... on a sidenote, the Landsat imagery also includes thermal band data. I see some evidence of a thermal anomaly in the "Sandy Island" area that seems consistent with that of the Chesterfield Islands.  But it is difficult to interpret and make any conclusions from it, so I'm not sure it's worth uploading.  I've never seen an article with satellite thermal imagery.  But, I can upload if someone wants to see it.  It's not exactly hunting gubernatorial candidates in "Predator", though.  Разрывные (talk) 03:26, 27 November 2012 (UTC)
 * THEMIS imagery from Mars Odyssey is used routinely...probably not what you meant. :)  Sean.hoyland  - talk 06:11, 27 November 2012 (UTC)


 * Sandy Island (Alleged Location) 2002-01-10, Landsat 7 ETM+, Thermal Fusion.png Strebe, I played around with the thermal imaging data from the Landsat granule some more, producing far more visually compelling results than I obtained initially, and have uploaded something you might want to look at. The thermal imagery (minus cloud cover) seems to correlate with the bathymetric data in showing there's topology approaching the surface in a suggestively island-shaped fashion in the alleged location.  However, in the visual spectrum bands, it quite clearly is *not* a surface feature within ~35m or so.  I am hesitant to include this image in the main article because it seems a bit too much like "original research" and I do not see any mention of thermal imagery regarding Sandy Island anywhere else.  But yes, there is clearly something there.  Just underwater.  Разрывные (talk) 06:01, 27 November 2012 (UTC)


 * Thanks, Разрывные. Very interesting. The discussion over at GMT-Help has fizzled for now on an inconclusive note. One thing that did come out of it, though, is that some data sets have been “masked” by vector data during processing so that they do not show what they purport to show, accounting for some of the anomalous “positives”. The most puzzling pieces of data remaining are the historical maps, robust bathymetry data reaching nearly to the surface before degenerating into interpolation artifacts, and the MODIS reflectivity data, also robustly showing something just below the surface. We should move this discussion elsewhere, by the way, if we continue it. Strebe (talk) 16:48, 29 November 2012 (UTC)


 * This talk page is hilarious.  I believe the ham radio operators from 2000, there's no island in the "obvious sense," meaning reality.--Milowent • hasspoken  04:36, 26 November 2012 (UTC)

Ile de Sable
If D'Entrecasteaux's discovery from 1792 refers to the real existing Ile de Sable to the NW of New Caledonia, http://upload.wikimedia.org/wikipedia/commons/b/ba/New_Caledonia_and_Vanuatu_bathymetric_and_topographic_map-fr.svg when was the Sandy Island in question discovered or sighted? I can't understand why just this information was erased, which is missing now in the article ... Best regards, Agisymba (talk) 22:08, 23 November 2012 (UTC)

For the answer to your question, please see the "Old Maps" discussion above. kind regards KPOK (talk) 21:36, 24 November 2012 (UTC)

Regarding the expedition of the Southern Surveyor
This is the official University of Sydney biographical page for Dr. Maria Seton nee Sdrolias, with a nice photograph. Yes, that is how her name is specified, with maiden name, and please, no comments about perceived or imagined gender bias on my part, as I am female, and a fan of cartography as well as earth sciences. (Sorry, I don't intend to sound hostile, just wanted to nip anything in the bud; you know how it is...). My opinion again, but it seems plausible that that area of the world is so little trafficked that there could be confusion about location, as someone here said earlier. There are still some remarkably remote parts of the world, e.g. the island which is associated with the ccTLD of *.tk is in the South Pacific and takes nearly a week to travel to, even now. --FeralOink (talk) 04:06, 24 November 2012 (UTC)
 * 1) Is it plausible that a post-doctoral fellow would be in charge of a research expedition? Perhaps it is. Maria Seton completed her PhD at the University of Sydney in 2005, over seven years ago.
 * 2) Perhaps it will be of use to those who may know, and have questions, to peruse the following two blog entries, including photographs of the expedition, and a somewhat melodramatic daily journal written by a few of the students (a very normal tendency, as they were no doubt excited, no criticism intended): Onboard Southern Surveyor with Dr Maria Seton, time stamp 12 Nov 2012, although it covers the chronology from mid-October through 3 Nov 2012, and Undiscovering Sandy Island onboard Southern Surveyor, time stamp 23 Nov 2012, and mostly a collection of news articles.
 * 3) There seem to be plenty of screen shots from the news media of how Sandy Island appeared on Google Earth, so it shouldn't be a problem preserving that. The image from the BBC was quite good.
 * 4) Finally, from what I could tell, Sandy Island seems like a partially submerged narrow N-S aligned lump of dirt there, also known as a sandbar. I am not qualified to make any sort of informed commentary though. Also, there is the matter of chronology, and seeming confusion over depth and location from presumably informed sources, which is odd.

Requested move 1

 * The following discussion is an archived discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. Editors desiring to contest the closing decision should consider a move review. No further edits should be made to this section. 

The result of the move request was: not moved, and the way is now clear for the RM forshadowed below. Andrewa (talk) 12:27, 1 December 2012 (UTC)

Sandy Island (New Caledonia) → Sandy Island (phantom island) – Sandy Island has now widely been reported as being a phantom island. 94.192.38.84 (talk) 23:22, 23 November 2012 (UTC)


 * I oppose this move. Sandy Island’s status as a “phantom island” is likely to evaporate when more information comes out. The sources are merely hours old, not rigorous, and it’s quite clear from bathymetry and satellite data that there is, in fact, an island-sized sandbar there. Meanwhile Sandy Island will always be in New Caledonia. The present title is correct. Strebe (talk) 00:41, 24 November 2012 (UTC)


 * I also oppose the move. I would suggest waiting at least another 24 hours, for the same reasons that Strebe mentioned. Please see my comments, in the prior section, #6 of the talk page, for my rationale for opposing the move. I believe we have seven days. --FeralOink (talk) 04:00, 24 November 2012 (UTC)

I also oppose this move, albeit on different grounds (see "Old Maps" above): looking at the series of maps referred to in the External links, it is clear from the chart used to illustrate Captain Cook's voyage in Resolution, when cross-referenced to his journal of 14 September 1774, that there is at least one "sandy isle" which is almost certainly a real island, albeit one just to the north of New Caledonia. There should at least be some reference to the reality of a "Sandy Isle" in the area of New Caledonia, if only to stop the press "rediscovering" the "lost island" at a later date!


 * By the way, this page does need to get moved. It should be named “Sandy Island, New Caledonia”, not “Sandy Island (New Caledonia)”. The parenthetical form is to disambiguate (WP:NCDAB) a title’s nouns that are not proper. I will make this proposal once this current one runs its course. Strebe (talk) 21:28, 26 November 2012 (UTC)
 * The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page or in a move review. No further edits should be made to this section.

Musings
Interestingly, the series of maps does not show a "Sandy In" or "Sandy Is" much further to the west of New Caledonia until a German map of 1881 and a British map of 1885 respectively. KPOK (talk) 22:15, 24 November 2012 (UTC)

Problem for naming is that there is another, real, island called Sandy Island off the north coast of New Caledonia. Sandy Island (New Caledonia) is still ambiguous, although only one of them (the one that has been in the news this week) has an article. Kevin McE (talk) 20:40, 25 November 2012 (UTC)
 * There is no need to include "phantom" in the name, based on longstanding precedent. See Agloe, New York and Argleton.--Milowent • hasspoken  04:43, 26 November 2012 (UTC)