The Name Outlived the Island

Every clause in that sentence is accurate, which is the only reason it's worth your time.

The destination was Oodaaq — a scrap of gravel off the northern tip of Greenland, sighted in 1978 during the first modern survey of the region, added to Greenland's national topographic map, and widely credited as the northernmost land on Earth. Morten Rasch of the University of Copenhagen and his colleagues meant to stand on it. Checking their coordinates afterwards against the Technical University of Denmark's, they found the GPS had put them 780 metres northwest of where they believed they were. What they had actually walked onto was a mound of mud and pebbles roughly sixty metres by thirty, three metres high at its peak, that nobody had catalogued.

So they proposed a name. Qeqertaq Avannarleq — Greenlandic for "the northernmost island."

Rasch said at the time that it would probably be short-lived; one good storm and gone. He was being conservative. In September 2022, the Danish researcher René Forsberg reported that Qeqertaq Avannarleq had never been an island at all. It was a grounded iceberg with dirt on top. Soil, pebbles and glacial mud, sitting on ice, sitting on water. The northernmost land on Earth was weather.

At which point the story stops being an anecdote about a bad GPS.

Because Oodaaq had done the same thing. A few years after it went on the map it stopped being findable, and the conclusion was that it, too, had been an iceberg. Same for the ones in between — 83-42, spotted in 1998; ATOW1996, named for the expedition that found it. A rotating cast of gravel banks, each briefly announced as the top of the world, each unavailable for a second visit.

Then somebody went and looked underneath. A bathymetric survey in 2022 checked what the gravel banks north of Kaffeklubben Island were actually resting on and found that none of them are connected to the seafloor. All of it is gravel riding on sea ice. Not one of the record-breakers was attached to the planet.

Which leaves Kaffeklubben Island — sighted by Robert Peary in 1900, landed on by Lauge Koch in 1921, and named by him after the coffee club at the Mineralogical Museum in Copenhagen. It has been the northernmost land on Earth this entire time. Forty-four years of people beating the record while standing on floating ice, and the correct answer had been sitting on the map since 1921, and it is named after a coffee break.

The instinct here is to file this under science was wrong, which gets it precisely backwards. Nothing in the story is an embarrassment except the four decades nobody checked, and the only reason we can say that is that somebody checked. The survey that demolished every claim came from the same enterprise that made every claim — and, more pointedly, from the same country. Denmark charted these banks. Denmark put Oodaaq on Greenland's map. Denmark named the newest one. Danish institutions are what dragged a sonar across the top of the world to establish that none of it was land. A body auditing a claim it owns is a rarer thing than self-correction in the abstract, and it is worth more.

Worth more, and cheap — because the record itself is worth nothing. Nobody lives there. Nothing ships through it. No border moved when a gravel bank changed status. Going to that much trouble over a piece of trivia is either absurd or the most honest thing anyone did that year. But it would be pushing my luck to call the whole question idle. In the Arctic, what counts as land arrives with lawyers attached, because territorial waters are measured outward from it. Canada and Denmark needed a formal treaty, signed in June 2022, to settle who owned Hans Island — a bare outcrop in the Nares Strait a few hundred kilometres south, argued over for half a century and resolved by drawing a line roughly down the middle. Nothing in the Kaffeklubben survey touched any of that. It's only that the question it answered — is this land, or is this weather — is not free everywhere it gets asked.

What I can't stop turning over is the category itself. The line between land and not land here is real, and it is not a matter of degree: the gravel banks are sitting on sea ice, Kaffeklubben is sitting on rock, and one of those is lithosphere while the other is a phase of water. The survey drew a hard boundary and drew it correctly. Which is exactly what makes the next part uncomfortable — because the boundary holds and the permanence doesn't. Kaffeklubben is glacial debris on a rock the ice will eventually get around to, on a landmass that is moving, on a crust that recycles. Nothing there is being kept. It is only being kept longer than anyone is around to notice, and geology is the one discipline that says so out loud. Permanence isn't a property of anything. It's a rounding error we apply to whatever outlasts the person looking at it.

And the name is still there. Qeqertaq Avannarleq, "the northernmost island" — a phrase now attached to an iceberg that has, five years on, very likely melted or drifted off or both. The word turned out to be the durable part. Which is either the whole problem with maps or the nicest thing you can say about them: we build language faster than the world will hold still for it, and then we go back with sonar to find out what we were actually standing on, and we write that down too.