Dear All,
On consideration, I doubt very much whether this is as easy a question as it seems. It can be very difficult to measure "remoteness", because of inaccuracies (bigger than you think) in the relevant maps, differences of datums and several other factors. The other interesting point is that measuring remoteness (as opposed to distance froma specific coastline) is very sensitive to minor inaccuracies. If I may give an example, we tried to locate the Pole of Inaccessibility on BAS Misc 3, and in those days did it using a semi-manual technique, manually fitting circles of known radius within the coast. When we repeated the exercise, with more accurate coastline data and using an automatic technique, we found that a) we had been wrong by several hundred kilometres and b) there were TWO widely separated points (one of which was close to our original determination) that were almost equally close to the coast - the difference was actually quite small.
SO, I would actually be surprised if there was a definitive answer!
The technique I would advise would be to look at it the other way on. Decide which coasts you were measuring from (After all, in one sense Tristan isn't at all remote - it is next to Inaccessible and Nightingale Islands!). Then, create a map of distance from these coasts. That is easy - most GIS will do it for you. However, it is harder because you want to determine great circle distance - I leave doing that as an exercise for the student (that means it is too hard for me!) Finally, simply look see where the various candidate islands plot on that map.
Good luck!
Paul
|