Tree rings are dated by matching their width pattern against a reference built from thousands of trees, and the match is a statistic, not a guess. One oak on one estate matches the reference perfectly to this year and then carries on, forty rings further, with a pattern the reference does not have yet.
I am an arboricultural consultant and I core trees for a living, mostly for age and condition.
How a core is dated.
You take an increment core, a pencil of wood from bark to centre, and you measure every ring's width to a hundredth of a millimetre. The widths make a sequence. Software slides that sequence along a regional reference chronology, which is built from thousands of dated trees, and finds the position where the widths agree, and reports how well. A good match is unambiguous: it is a statistical fit that would not occur by chance, and it tells you the year of every ring.
The outermost ring, under the bark, is the current year. That is the check. If the outer ring dates to this year, the whole sequence is right.
There is an oak on an estate I work for, a big one in parkland, and I cored it in 2020 for a condition report.
The core dates perfectly. Every ring from the centre out matches the reference, and the ring that should be 2020 is 2020.
And then the core carries on. There are forty more rings outside it, under the bark.
Forty rings is forty years. A tree does not have rings for years it has not lived.
What has been checked, because this is a claim no one in my trade would accept on one core.
The core. Five more from the same tree, at different heights and aspects, in 2020, 2022 and this year. All five have the extra rings, and the number of extra rings on the cores from this year is forty, the same as in 2020. The tree has not added to them. It has added three ordinary rings since 2020, dated to 2021, 2022 and 2023, and they are inside the forty, not outside them.
So the forty rings sit outside the current year, and the tree goes on putting its real rings in underneath them.
The measurement. Two laboratories, blind. Both date the inner sequence to 2020 with a strong fit and both report the outer forty as unmatched.
The wood. It is oak. The outer forty rings are normal oak wood, with normal vessels and normal density, and they are not a fungus or a growth or anything a pathologist has a name for. A pathologist has looked.
The part I did in 2023 that nobody asked me to.
Unmatched rings are just widths with no date. But a width pattern can be tested against itself, and against the climate. So I sent the outer forty widths to a colleague who does climate reconstruction, without telling her what they were, and asked her what years they looked like.
Her answer was that the sequence has the signature of a run of very dry summers, three of them together, about a decade in, then a long stretch of ordinary years, and then, towards the end, four narrow rings in a row of a kind she has only seen in the reference around the great droughts.
She asked which tree, because she wanted to add it to the reference, and I said it was one I could not use.
What is done about it.
The condition report says the tree is mature, sound, and of an age consistent with the parkland's planting, which is all true.
The extra rings are in my files, in the two laboratory reports, and nowhere else.
What I do.
I survey the estate's trees on the cycle, and that oak gets a look like the others.
The one thing I have changed is that I do not core it again.
A core is a small wound and a healthy oak seals it in a season, and I have taken six from that tree and it has sealed every one.
The last, in March, was the one that stopped me. A core comes out of the borer in one piece and you can read it in the field with a lens. I read it standing by the tree. The outer forty rings were there, and the last of them, right under the bark, was only the start of a ring: the early wood of a spring that had just begun, in a year forty years from now, and the sap on the borer was wet.