Wikipedia:Featured picture candidates/Antipodes Islands

Topographical map of the Antipodes Islands[edit]

Voting period is over. Please don't add any new votes. Voting period ends on 20 Jul 2010 at 04:32:48 (UTC)

Reason: Highly EV vector map, created from public domain sources using purely open source software
Articles in which this image appears: Antipodes Islands, Bollons Island, New Zealand outlying islands, List of islands of New Zealand
FP category for this image: Maps
Creator: Matthewedwards : Chat

Support as nominator --Matthewedwards : Chat 04:32, 11 July 2010 (UTC)[reply]
Comment: It is difficult to tell where the separation is between Hut Cove and Anchorage Bay is on the existing map. One of the references, here, makes the separation clearer. Would it be possible to make this separation more visible? Spencer^{T♦Nominate!} 17:46, 11 July 2010 (UTC)[reply]
- And I don't know if it's possible to rotate the "South Bay" label so that it is level. The other bay labels are all straight, and use lines in cases where there may be confusion. Spencer^T♦C 17:53, 11 July 2010 (UTC)[reply]
  - No problem, done both. Matthewedwards : Chat 19:13, 11 July 2010 (UTC)[reply]
    - Support All of my concerns have been addressed. Clean, yet detailed, high quality and good enc. Good colour choices as well (icky map colours are a pet peeve). Spencer^T♦C 23:08, 12 July 2010 (UTC)[reply]
~~Support~~ Oppose per J Milburn. Nicely done, and (thank god) labelled. I hate Wikipedia maps a lot of the time, because you go and look at, say, a map of Africa, and if you don't have every country memorized already, not one map on Wikipedia will bother to tell you - you'll have to look at every damn country page to see which one is highlighted. Ugh! Anyway, enough off-topic ranting. Adam Cuerden (talk) 13:36, 12 July 2010 (UTC)[reply]
Support Excellent work: clear color scheme, good use of labels and a legend, good sourcing on the description page, and the global locater is a useful touch. Jujutacular ^T · C 19:45, 12 July 2010 (UTC)[reply]
Strong oppose. I hate to do this, but this just isn't up to the standard of the source maps- not nearly. The coastlines are far, far too smooth, and the level of detail is pretty minimal. Take, for example, the left side of South Bay (I've forgotten technical terms here, sorry). Even when viewed at low-res on the source maps, this is clearly a peninsula with only a narrow landbride across when compared to the size of the main bulk of the islet, as opposed to the lump on the bottom of the svg map. Also take a look at Bollons Island- two pointy outcrops clearly visible on the other maps stop this being the idealised crescent shape on the svg map. Further comparisons of the coastline reveal severe shape problems, without even going in to the more technical issues (I haven't looked in-depth for problems with the mountains themselves), or addressing the lack of detail when compared to the others, especially this one. Compare this nom, perhaps, to a current svg topographic map of FP status- File:Falkland Islands topographic map-en.svg. It's in a different league. J Milburn (talk) 00:52, 13 July 2010 (UTC)[reply]

I don't want to sound facetious, but perhaps you don't completely understand all the ins-and-outs of cartography. Firstly, I only used the LINZ map as a source for accurate heights of hills and mountains. Their help page tells us that their maps are at a scale of 1:25,000 (it's 1:50,000 for NZ, and 1:25,000 for the outlying islands). My map's scale is 1:372,000. All maps are scaled, and there is always some degree of "incorrectness". As LINZ says, "To show as much information as possible, maps at this scale are "generalised" (meaning that small twists and turns of features like roads and rivers are smoothed out). The best you can achieve from a map is the ability to calculate the length of the mapped (generalised) river. Different maps are likely to have different degrees of generalisation, so you may get different results for the same feature."

When most people talk about the scale of a map, and "scale: 1: xxx xxx" what they mean is that one centimeter on paper corresponds to xxx xxx cm on the ground; however, when a map creator talks about the scale, especially with digitally created maps, it can refer more to the accuracy of the spatial positioning of objects, and less about the linear scale. They are still tied together, though, and even though the accuracy is rarely shown on commercially available maps, each cartography institute or private cartographer will still operate by it.

As an example, the USGS's standards, set in 1947, on a map with a scale of 1:50 000, 90% of points tested must fall within a maximum of 0.508mm on the map in relation to their true positions, or 25.4m.[1] According to one Wikipedian who is a cartographer for the Portuguese Navy, the maximum error allowed there is 0.25 mm, which means that for a map scaled at 1:50 000 must be correct to within 12.5m. France is even more strict. Their maximum is 0.1mm, so a map at 1:50 000 must be true to within 5m on the ground.[2] I don't know what New Zealand's error margin is. They don't tell us on their website. Either way, both the Falklands Island map and this Atipodes Island map, and most other similar maps created by Wikipedians, are drawn with a maximum error level of 0.25mm. Both maps are created using NASA's Shuttle Radar Topography Mission [[digital elevation model], and the NGA's SRTM Water Body Data. The resolution for the data is three arc seconds, which works out to 93m. When a new map is created by copying a map or interpreting data published by another source, as in this case, cartographers often work under the rule that their map must be accurate to no less than three times that of the original data. So if you draw such a coastline from Landsat Enhanced Thematic Mapper Plus, which has a resolution of 14.25 m the accuracy of our map can be calculated as

14.25m\times 3{=}42.75m

. The scale can then be calculated as

{\frac {42\,,750\,mm}{0.25\,mm}}=171\,000

so the scale is 1:171,000.

Anyway, I preserved the original resolution of the DEMs, which is 93m. So I calculate

{\frac {93\,,000\,mm}{0.25\,mm}}=372\,000

and my scale is 1:372,000.

You're asking me to change the resolution of the source material (the DEMs). I can't do that. This is the one of the best resolutions that we can work with at Wikipedia. The LINZ maps are protected by crown copyright, so we can't use them here. I would love to make them more accurate, but we can only work with the material that is available to us. You said that the Falklands Island map is in a different league - it was created in exactly the same way. Yes, the path of the coastline was simplified, and nodes were deleted, but the same shape remains. There are so many nodes in the original path that the coastline looks just as smooth. The path is only simplified to provide a significant reduction in the number of nodes to produce a file small enough for accessibility on the Internet, but it still preserves the original path of the coastline. I downloaded the material from NASA last night, and projected one of the segments. There were 20,904 nodes in the original vectorized coastline. In the exact same area of the promoted image, there are only 678 nodes. The number of nodes has been significantly reduced, but what remains is still true to the original lines. In this map, there are 542 nodes in the original, unmodified coastline. There are 90 in the finalized version, and they are still true to the original lines. The level of detail is no more minimal than for the Falklands map. The area and natural topography of the Antipodes is less than the Falklands', which is why it may appear to someone who doesn't work with maps that this is minimal and in a lesser league to the map of the Falklands. There are no "problems" with the coastline, or with the topography of the land. And by the way, if you overlaid a 1:50,000 or 1:25,000 map with a precision of 0.25mm of the Falklands over our map, you will see that our map's coastline is significantly different.

I don't point out technical faults with photographs because I don't understand them. Matthewedwards : Chat 19:07, 14 July 2010 (UTC)[reply]

Oppose per J Milburn, the coastline definitely needs to be more detailed, this shows us it is more detailed. — raeky ^{(talk | edits)} 01:57, 14 July 2010 (UTC)[reply]
Oppose -- It is a good work, but not featurable imo. As noted above, the coastline lacks detail (the technical jargon is too generalized). I don't like the triangular black symbols to mark the elevation points; why not white and equilateral? Why is the longitude of the central meridian between brackets? Finally, this is not a topographical map but a hypsographical map, since only the relief is represented (with hypsometric tints). -- Alvesgaspar (talk) 16:11, 14 July 2010 (UTC)[reply]

The black triangles were taken from File:Maps template-en.svg, which is based off File:Maps template-fr.svg. Both use the same triangles, and maps featured here and at Commons use the same triangles. I removed the parentheses. I'm not sure why they were there but I think it was to do with when I was making them all fonts I put them in there to forget not to change the style. And yeah, you could call it hypsographical, but all our other maps on Wikipedia are called "topographical", I added the streams but it's an uninhabitable island with no other distinguishable features such as roads, tracks, etc, and so calling it hypsographical didn't seem entirely correct to me either. And frankly, I'm sick of being accused of putting incorrect information into images. It's the same as being told I'm putting incorrect text information into articles: Vandalism. Matthewedwards : Chat 19:27, 14 July 2010 (UTC)[reply]

Strong oppose A low technical standard. To be FP, it needs more gradient lines, and something flashy. Something like a unique, artistic presentation of the concept. Gut Monk (talk) 00:16, 15 July 2010 (UTC)[reply]
- I would agree here, using publicly available GIS Shape files that are low-res, is definitely not "high technical standard" that would warrant a FP status. Although I highly admire your work and effort to provide nice maps for articles, and don't want to discredit your work, because, trust me, we greatly appreciate it. But try to see it from our point of view, there is a pretty significant detail loss in the coast line with that dataset. — raeky ^{(talk | edits)} 00:23, 15 July 2010 (UTC)[reply]
- Oh, I get it.. you mean something like this.. I think that a ladybird and a robin-red-breast should be flashy enough. ;) Matthewedwards : Chat 23:46, 15 July 2010 (UTC)[reply]
  - I love it. J Milburn (talk) 23:48, 15 July 2010 (UTC)[reply]
    - Well, at least the bird and bug folk will support it! :p Matthewedwards : Chat 23:35, 16 July 2010 (UTC) [reply]
- This reasoning is absolutely bogus for a map. Cartographers should and do strive for consistency. It makes the maps easier to read and more accurate. Asking for originality for the sake of originality undermines EV and likely degrades the IQ. Cowtowner (talk) 07:41, 18 July 2010 (UTC)[reply]
Support I find the nominators arguments to be convincing and in line with the logic (which is sometimes used here at FPC): larger subjects require more resolution, while smaller ones may have less. Cowtowner (talk) 17:28, 17 July 2010 (UTC)[reply]
- Look at the maps linked on the image page- an image showing far, far more information is possible and desirable. J Milburn (talk) 14:15, 18 July 2010 (UTC)[reply]
  - I have no doubt that there is; the same could be said about the Falklands image. More detail will always be available. For the size of the islands, I think this amount of generalization is acceptable. Cowtowner (talk) 14:45, 18 July 2010 (UTC)[reply]

Not promoted --Makeemlighter (talk) 23:22, 20 July 2010 (UTC)[reply]