It is a little weird to get your head around unless you're used to thinking in terms of the scientific idea of measurement, where
every measurement you make has some associated error. Think about using a yardstick to measure the length of the kitchen counter: say you get 83 3/8, 83 1/8, and 83 inches on three tries. They differ because 1) you had to reposition the yardstick twice and 2) there's perhaps 1/16 inch error in your reading of the durn thing any one time. If you dad came over, he might get 83 1/2 inches with your yardstick, and 83 5/16 with his steel tape! Which is "correct?" None of the five, most probably. The "true" length might be 83.263 inches, plus or minus 0.002 inch, but there's still some error there - the +/- 0.002 piece.
Even if you use some kind of super-accurate laser measurer, there will still be some uncertainty in the length. It might be less than a millionth of an inch, but it's still there.
Now back to Lake Suigetsu. Some poor grad student counted 38,000 or so little black/white pairs of lines in those cores. There are potential errors there - losing count ( compare counting votes in Florida??), missing bits of core, years when no diatoms grew, so no white layer formed.... etc. The paper makes a estimate as to the size of this sort of error. Pretty small, but there, and estimate-able.
Then the researchers sent 250+ little bits of bug wings, leaves, and such, each marked with the count of the varve it was picked from, off to Germany for measurement of the carbon-14 it still contained. Each bit had to be handled and prepared for measurement. The measurement had to be actually run on each. Everything had to be recorded. Every step has its own error associated with it - small, except for misidentification of a sample or two or dyslexia interfering, maybe, but still present.
Then we come to putting the two strands together. One of our clocks is the "count the varves" clock, and it might say that the varve we'll call Louie is 15,456 years old - 15,456 pairs below the present lake bottom. And let's say that the estimated error is +/- 444 years. The other clock is the carbon-14 clock. Louie might be 16,001 years old by this measure, which was figured to a first approximation by 1) measuring the 14C in the bug wing 2)
assuming constant 14C input into the atmosphere and the bug back then and 3) using the known decay rate of 14C to calculate how long it took to lose 14C from the amount initially in the bug wing down to the amount measured today.
Now, when Kitagawa and van der Plicht actually compared these two clocks, an amazing thing happened: they agreed pretty well! (They actually intended in the first place to use the varve clock to correct the 14C clock, since the knew from other studies that the assumption in 2) above wasn't exactly true - 14C
doesn't form at a constant rate. Their whole goal was really to provide a measure of how its formation varied in the past. But for the purpose of illustration, bear with me a moment more.) At times in the past when the Earth's magnetic field was weaker than now, more cosmic rays got in to hit nitrogen in the atmosphere, and more carbon-14 was formed. In the Ice Age, when the oceans were colder than now, more carbon dioxide ( and thus more carbon-14) got dissolved in the deep water and thus less was in the air for our Louie bug to pick up.
The paper simply uses the varve count to correct the "fast" and "slow" times in the 14C clock caused by this sort of thing. Impressive as the agreement before this correction is, it gets even much more impressive when you go to the footnotes. The clocks of tree rings in Germany and Finland agree. The uranium-thorium dates of corals in New Guinea and in the Bahamas agree. Dates from ice cores in Greenland, Antarctica, Peru, and Kenya agree. It's just a little tough to make that many independent measurements, each made with independent yardsticks, all line up to give the same numbers (each with errors, still) unless the really are all measuring something real.
Am I making sense?
™ Version 4.2
