please empty your brain below

What an amazing achievement. Thanks for bringing this to a wider audience.
About a decade ago I went on a cruise across the Pacific and teamed up with an American, who had been on their equivalent of University Challenge, for the daily afternoon quiz. Same question but re USA states. He swore blind the answer was California and it took a heck of a lot of arguing to change to Alaska!

Good idea today, I would certainly have said Cornwall top, probably Devon second, but Essex?
And to prove the point of how difficult this is to measure, google suggests both Essex and Cornwall have the longest coastline.
Fantastic and informative post! I knew the final would be contested between Cornwall and Essex, but really wasn't sure which would come out on top.

I wonder where Islands come into this though - does the circumference of the Isles of Scilly count as part of Cornwall's total?; and similarly what part of all of the numerous (in some cases large - e.g. Foulness) islands in Essex (which in most cases are attached, at least at low tide, to the mainland by a strood)?

I'd previously seen claims that Essex had a longer coastline than Cornwall, which I wasn't quite convinced by, but possibly if you included all sides of Foulness, not just those facing the sea, you might make it up.
When do estuaries stop being coastline and start being riverbank? Presumably long before they stop being tidal otherwise we’d be looking at the Middlesex coastline.
The Ordnance Survey’s blogpost confirms they didn’t include the Isles of Scilly.
Please be very careful when describing Cornwall as an English County. As natives of the area will be quick to inform you, Cornwall is a Duchy, and historically a quite different place from England.

Take a look at the Queen's flotilla on the Thames during her Jubilee. The flags of England, Scotland, Wales, Northern Ireland and CORNWALL were shown flying. No county flags were included.
That’s an interesting point Philip.
DG did a post recently about the Crow Stone at Chaikwell and the London Stone on the opposite side of the Thames Estuary, would this be where the sea ends and the river begins in today’s calculation?
If the OS High water mark is used, as DG describes, the coastline would include the tidal Thames as far up as Teddington Lock. On that definition, Hammerton's ferry between Ham House and Twickenham would count as a sea voyage!
My cycle "computer" which measures wheel rotations, always shows longer distances than the map, even though I calibrate it carefully by wheeling the bike in a straight line with a marker on the tyre to count rotations. If it were just calibration error it would sometimes be more, sometimes less. I think this is because it's taking a reading about every two metres so includes my many minor meanders.
Reminds me of a question about which country has the most islands; almost impossible to measure.
From a walking perspective, Cornwall's coastline also goes up and down an awful lot. The SW coast path has never really appealed because I think I'd find the thought of every down being followed by an up, coupled with the detours round each estuary, rather dispiriting.
david - Not almost impossible, but completely impossible, because it keeps changing. Some islands become peninsulas at low tide, others disappear under the waves at high tide.
Just take the coastline without entering any bays or inlets. The bays rather depending on their shape. My huge Port Phillip Bay in Australia would not be included with its very narrow inlet.
According to the Ordnance Survey the UK county with the largest coastline is Na h-Eileanan Siar (Western Isles) in Scotland with a boundary of 3726 km.

Cornwall and Essex are relative tiddlers.
Thankyou DG for that Quintin Lake link - new to me and a fascinating timesink!
Fascinating... but I'm surprised Yorkshire doesn't seem to figure on any of the lists!
Interesting about the Western Isles (which used to be called the Outer Hebrides). But that moves the discussion away from traditional counties, where it started. The Island of Lewis and Harris, bizarrely, is shared between two different traditional counties.
A cycle computer will always give a bigger reading from the front wheel than the back wheel.
Yay for Cornwall! Kernow bys vyken!
My county of birth
Essex might not have the longest coastline, but it certainly feels like it does when walking it! Some interesting islands too like Foulness and Mersea, not sure if you included these.

I certainly enjoyed walking the coast of Cornwall more than I did Essex. However most of the coast of Essex was also really lovely to walk - just not as nice as Cornwall. A lot of estuaries though.

However I was really surprised how remote much of Essex feels. For example the Dengie peninsula and Foulness island.

The classification of coast the Ordnance survey use is where it is tidal it's part of the coast. So I think technically Middlesex (as was) has a coastline because the Thames is tidal (or at least, semi-tidal) as far as Teddington.
DG, how did you measure the Google Maps coastlines? Automatically in some way, or by tracing the coastline from the images by eye?

A related (but more mundane) quirk of length measurement when using any grid approximation is the diagonal (or shortcut) paradox. A bitmap image is made of a grid of rectangular pixels, so a diagonal line can only be represented as a series of horizontal and vertical steps. With a high enough resolution, or a blurry enough view, the eye is tricked into seeing a smooth line. But however small the pixels, the total distance of the horizontals plus the total distance of the verticals for any zig-zag along the diagonal remains the same, even though intuitively the total ought to converge towards the length of the diagonal that it steps along. For example, a line 400 along and 300 up will have length 700 however many square pixels represent it as a staircase; but Pythagoras shows that the length of the diagonal is 500, and that is what our eye perceives.

The grid-length illusion gives a discrepancy no worse than √2:1 for any straight segment, rather than raising the more formidable spectre of infinity. Then again, when measuring fractal coastlines in the physical world, perhaps quantum uncertainty can also spare us infinity, diverting us eventually into a different kind of ineffable.
Devon, I think, does hold the distinction of being the only county with two completely separate stretches of coastline. ("Completely" to allow for exceptions like exclaves -- if any still exist in the UK -- little offshore islands, and such.)
Andrew S - having been lucky enough to walk the whole Cornish coast I can highly recommend it; the stunning scenery for almost the whole route makes up for the physical challenges. On your point about climbing and estuaries the geology means that the north coast tends to have more ups and downs while the south coast has more estuaries that you have to divert inland to get round.
I too wonder which, if any, of the Essex islands were included. According to Ian Yearsley's book 'Islands of Essex', the county has "something like 30-40 islands in all ... this equates to more islands than all the counties in the rest of eastern England can manage put together".
Thanks for linking to my project. Interesting observations in this post. My distances are calculated on flat map 1:25,000 OS map plots. I found GPS trackers or the iphone pace counter generally increased that distance by 10-15%. In Essex I included the islands of Mersea (21km), Foulness (29 km), Canvey (25 km) & Two Tree (4 km) in my total So the walkable coast of mainland Essex would be 447 km.










TridentScan | Privacy Policy