The gauge was installed in 1912 and has been read, by hand, almost every day since. It is a brass dial the size of a dinner plate, bolted to the harbour wall at a height nobody alive chose, and it does one thing: it tells you where the water is.
What the record actually shows
Nothing dramatic. That is the first thing to say, because a century of daily readings sounds like it ought to produce a headline and it does not. It produces a line. The line goes up. It goes up slowly enough that no single year looks like anything at all, and consistently enough that no ten years look like nothing.
The patience is the part that matters. A satellite can tell you the sea surface height across an ocean basin in an afternoon. It cannot tell you what this harbour was doing in 1913, and it never will be able to.
The method has not changed
Someone walks out, reads the dial, writes the number in a ledger. The ledgers are in a cupboard. There are forty-one of them.
This is not nostalgia. A measurement you can explain to a fourteen-year-old in one sentence is a measurement that can be checked, and a record that can be checked is worth more than a record that is merely precise.
Where the gaps are
There are gaps. The worst is 1943 to 1946, and there is a shorter one in 1978 when the dial itself was replaced and the new one was mounted eleven millimetres lower. Someone noticed, wrote it down, and every reading since has carried the correction.
That note, three lines in a ledger, is the single most valuable thing in the cupboard. Without it the whole series after 1978 would be quietly, invisibly wrong.
What happens next
The ledgers are being photographed. Not transcribed, photographed: the handwriting, the corrections, the coffee rings. Transcription can follow, and errors in a transcription can be found and fixed as long as the thing being transcribed still exists.