Why You Can’t Measure the Coastline of Britain

Select video clips courtesy of Getty Images

Select video clips courtesy of the AP Archive

Special thanks to MapTiler / OpenStreetMap Contributors and GEOlayers 3

[D__Cain]

The paradox is even funnier to think of when you think that every single wave of water changes the coast

[decreasing_entropy]

"One of the most difficult parts for me about learning anything new is simply knowing what the right questions are to ask." This is absolutely accurate and very well put!

[bababababababa]

Fun Fact: Japan and Philippines both have longer coastlines than Russia somehow

[ironcito]

It's important to know _why_ you need to measure the coastline. For example, if you're doing military planning and you want to defend a coastline against seaborne invasions, you can use a rather long "ruler" because you can place patrol ships offshore, you can have radar coverage, etc, all of which "smooth" the coastline. On the other hand, if there's an oil spill and you want to know how much coastline is affected, you need a much smaller "ruler" because the oil will get into every nook and cranny. I suppose that's why those U.S. sources came up with different lengths: they had different objectives in mind.

[Numba]

I've heard of the coastline paradox before! It's kinda like the paradox of stepping half the distance between two points and never arriving. It's funny just how widely different measurements can turn out depending on the measuring method used lol. Thank you for this fun exploration of the conundrum!
Merry Christmas out there everybody! ✝️🎄

[haleysettembre]

The coastline paradox is one of my favorite paradoxes to wrap my mind around

[christinafidance]

I live on a boat on the Chesapeake Bay and I still find it absolutely astounding that I once heard if you could stretch out the entire coastline of the bay and it’s tributaries out into a straight line, it would stretch to California and back TWICE! That’s insane!!! It’s just so jagged with countless little points and peninsulas.

[who_stole_my_username]

I'm surprised there isn't a defined standard for this measurement. Personally I would use a 1 metre rule, as that is close to what you would use if you were to physically measure the coastline by walking along it.

[mattg]

I'm not sure which "ruler size" this should be applied as, but I think coast lines should be measured at a scale consistent with a person walking along the coast. Measure features that you walk along, not step over for example.

I also think there should be a secondary measurement ignoring fine detail, more like a shrink wrap method to determine the shortest path, like that a ship might take, which also seems useful.

[noodengrthree]

I read the adventure travel book Two Legs, Four Paws. It was about and by a woman and her dog who walked along the coast of the UK. She measured it off in miles, something like 4000 miles. The coastal paradox was driven home to me by comparing it to walking across the USA which is about 3/4 of that distance !

[ietsbram]

id be very interested in a video about the shortest coastlines, or the biggest countries that aren't landlocked but still have weirdly small coastlines like Bosnia and Herzegovina, Togo, Congo, etc

[shTree]

NOAAs numbers are due to the substantially higher quality of their data; they have a very complete coastal (topobathy) lidar dataset which they use for making charts.

[pawerybacki]

I think it’s relatively easy to agree that every coastline length should be reported with the measurement unit used (e.g. 1m) and angular precision (e.g. every 45° degrees of rotation). Alternatively, I’d suggest dividing a map into 1-meter pixels and counting a pixel as an extra meter of coastline if it contains both water and land. This kind of pixel map could be additionally standardized if we agreed on the precise coordinates of the center of a given country and centered the pixel map accordingly.

[KelvinNishikawa]

Conceptually, there's a simple method to normalize these lengths that we use in computer graphics: The signed distance field.

Rather than measure directly on the transition between land and water, divide the map into tiny grids, and assign each grid a distance to the nearest land/water transition. Then you can pick an arbitrary distance threshold and interpolate between the grids using marching squares or subdivision. You'll get a normalized border with much of the smaller details smoothed away. Using this method, coastlines would be much closer to the distance a plane would fly along. We actually use this method to define coastline EEZs, which are significantly smoother and less complicated than actual coastlines.

[Stop_being_younger_than_me]

Paul McCartney even included this in the album Let it Be. 'The Long and Winding Coast' however was changed to road though to avoid mass panic exposing so many people to the paradox.

[pepper]

this dude is the best at stretching out simple concepts to 14 mins

[geometryemperor]

I think this also applies to river lengths. This is what makes the longest river a hard thing to answer.

[snogard]

This channel is on a roll for making videos about a topic they already made a video for.

[sonnyhutchins]

scotland's coastline is utterly insane the amount of islands are up there is unbelievable

[shmoop]

For once It's nice to hear the creator behind RLL having a good time with a video and not worrying about existential geopolitical drama. The guy deserves a breather.