Never mind. Apparently I didn’t see previous discussion on Somalia.
It looks like Chile’s is close to the border, ~10 miles.
Before the independence of Bangladesh, Pakistan might have fit this criteria.
The Indian Ocean and Pacific departments and territories, although small, are probably enough to shift it a bit. These include Mayotte and Reunion in the Indian Ocean (which are Overseas Departments), Kerguelen and Crozet in the Southern Ocean, and New Caledonia and French Polynesia in the Pacific, among others.
If you ignore the land requirement, Japan would be an example. Its geographic center lies outside of the country, in the Sea of Japan.
Depends on what you mean by “outside the country,” since the point is within Japan’s territorial waters.
Malaysia’s centroid is at sea, but it also looks like it may be outside the country’s territorial waters (barely) because there are some Indonesian islands between the provinces on the Malay Peninsula and those on Borneo.
There are quite a few French islands but due to their size and how they are spread out it looks like it would pull the centroid south but very slightly.
.svg){kind=link}
French Overseas Territories all together total 119,396 km2 (about 70% is in French Guiana), but almost all of them, totaling around 116,345 km2, are in the Southern Hemisphere (and all but St. Pierre and Miquelon are south of France). Metropolitan France has an area of 543,940 km2. So somewhat surprisingly these southern territories collectively are about 21% the size of metropolitan France and pull its centroid south enough to be outside the country.
The statistics seem correct (did not re-calculate the centroid, though), but politically they do not call a place like French Guiana an “overseas territory”, as it is a straight-up department of France, part of the EU, and so forth, and as such should unquestionably be included in the calculation of the geographic center. I have heard the collective term “overseas departments/territories” (DOM/TOM) as well as “overseas regions”.
Yes, French Guiana and Reunion have the same political status in France as Alaska and Hawaii do in the US.
The center of the weighted distance between met. France and Guyana is about 1000 km from met. France.
That’s in the ocean. The other stuff is almost noise.
Evidently it’s not, based on the calculated centroid that’s been linked to. If you have a better calculation, please post it.
France is definitely an interesting case. The CIA World Factbook notes that French Guiana, Guadeloupe, Martinique, Mayotte, and Reunion are considered parts of France proper, but there are still separate listings for Clipperton Island, French Polynesia, the French Southern and Antarctic Lands (which includes assorted uninhabited islands in the Indian Ocean–but not Reunion–as well as the French Antarctic claim to “Adelie Land” which is in abeyance under the Antarctic Treaty), New Caledonia, Saint Barthelemy, the French half of Saint Martin (but apparently Saint Martin is part of the EU), Saint Pierre and Miquelon, and Wallis and Futuna.
The Netherlands is another complicated one; the Caribbean islands of Bonaire, Sint Eustatius, and Saba are part of The Netherlands itself (and would therefore need to be counted when computing a centroid for The Netherlands), while Aruba, Curacao, and Sint Maarten (the Dutch half of Saint Martin) are all “self-governing countries” within the Kingdom of the Netherlands (and would therefore presumably get their own centroids).
In general, I think if it’s listed on this page, a territory probably shouldn’t be used when computing the geographical center of the country upon which it is a dependent.
Nitpick : The OP did specify that the countries be converted to planes (Euclidean)but different cartography methods do it differently.
If we did this for the US, we would likely include Alaska and Hawaii but exclude the territories. Therefore, places like Queen Maud Land or even Svalbard don’t affect Norway’s centroid, but French Guiana is analogous to AK or HI in that it’s an integral part of France, and so it applies.
The other parts of France that apply are Martinique, Guadeloupe, Mayotte, and Réunion. Corsica has a very minor pull.
The OP’s question did not define the “geographic center” as the centroid; he asked for the point with the minimum mean squared distance. This is an unconventional choice that would accentuate the effect of the distant disconnected areas.
That sounds like the usual convention, as long as we remember to work over a sphere or ellipsoid to calculate the centroid (has anyone checked to see if that web page list is accurate?). One could try a different approach to defining the geographic center, but what would it be, and why would it be any better? It’s no more than an amusing geometric question anyway.
This thread makes me very happy. Nothing to add, just wanted to say that.
If you think it makes more sense, we can use orthodromic distance instead. But I think this is unlikely to change the answer to my question, possibly except for a few rare cases where the country’s landmasses are located about a hemisphere apart or more. (So France, maybe, and also Norway if we decide to count dependent territories.)
That’s debatable IMO without inserting some kind of qualifier like ‘almost’. For one obvious thing the overseas departments have their own political designation, département et régions d’outre-mer (DROM). HI and AK have no special political designation, but are simply states (and ‘Lower 48’ is a geographical not political term). There are some differences in local autonomy and inclusion in French statistics of DROM’s v metropolitan departments. They are more similar to US states than US territories in the important respect that they are represented in legislative and presidential elections, but not quite as indistinguishable from metropolitan departments as US states are from one another.
Not that it’s a big deal, I just think there’s more ‘play’ in the question of whether DROM’s should be included in France’s geographic centroid, or US territories should be included in its, than just saying DROM’s are exactly like US states.
Earlier in thread it was implied (by saying Denmark’s centroid would be well out at sea) that Greenland is the same as metropolitan Denmark and that’s definitely not true. Although again Greenland is proportionally represent in the Danish parliament unlike US territories it is a ‘constituent country’ of the Kingdom of Denmark not a province and has lots of local political autonomy that subdivisions of metropolitan Denmark don’t. Also, complete independence is an active mainstream issue.