Dendrochronology Alignment Calculator

What Is the Dendrochronology Alignment Calculator?

The Dendrochronology Alignment Calculator works out two of the core statistics dendrochronologists use to test whether an undated tree-ring series crossdates against a reference chronology: Gleichläufigkeit (GLK), and the Baillie and Pilcher t-value. According to the Laboratory of Tree-Ring Research's overview of crossdating, crossdating is the basic principle underlying all of dendrochronology: matching a pattern of wide and narrow rings from a new sample against a master chronology built up, year by year, from overlapping living and historic trees.

Figure out which mode fits your data before entering it: single alignment mode scores one candidate overlap position between two series, while the best-fit offset search slides a shorter floating sample across a longer master chronology and ranks every tested position by statistical strength, mirroring how dedicated crossdating software works through the same problem.

Gleichläufigkeit (GLK): Measuring Year-to-Year Agreement

GLK, also called the percentage of parallel variation, is a straightforward, non-parametric measure: for every year-to-year interval, check whether both series moved in the same direction, both wider or both narrower than the year before, and divide the count of agreeing intervals by the total number of comparable intervals. A 15-year overlap with 11 intervals moving in the same direction out of 14 comparable intervals gives a GLK of just under 78.6%. Work out GLK alongside the t-value rather than relying on either alone, since research on the Gleichläufigkeitskoeffizient published in Archaeometry notes that GLK is more commonly applied in European dendrochronology while t-value dominates practice in the United States, and the two measures can disagree at a given position.

The t-value: Correlation-Based Significance Testing

The t-value converts the Pearson correlation coefficient between two ring-width series into a significance statistic that accounts for how many overlapping years the comparison is based on: t equals r times the square root of n minus 2, divided by the square root of 1 minus r squared, where r is the correlation and n is the number of overlapping years. Baillie and Pilcher originally proposed 3.5 as an arbitrary preset threshold for an acceptable date; a t-value that high has roughly a 1-in-600 chance of occurring by coincidence for series 100 years or longer. That said, come back to the number itself with some caution, since many labs now treat 4.0 as the working baseline for oak specifically, having found that higher values are often needed to produce genuinely secure crossdates in practice. A later empirically-determined study of the Baillie and Pilcher t-statistic's statistical significance for British Isles oak revisited the original 1973 threshold directly, testing it against a much larger body of real chronologies than was available when the figure was first proposed, and broadly supports treating 3.5 as a minimum floor rather than a comfortable target.

Given that overlap length changes how meaningful any single t-value actually is, always report the number of overlapping years alongside a t-value rather than the statistic on its own. A t-value of 4.2 calculated from a 25-year overlap carries meaningfully less weight than the same 4.2 calculated from a 100-year overlap, even though the raw number looks identical in both cases.

Searching for the Best-Fit Offset

An undated, "floating" tree-ring sample does not arrive with a known position against the master chronology, so dendrochronologists test many candidate offsets and look into which position produces the strongest statistical match. Pull out the t-value and GLK at every possible offset, then rank the results, since a single high t-value in isolation means far less than a high t-value that stands out clearly from every neighboring offset. On top of that, a genuinely secure crossdate should show one offset scoring well above all the others; two offsets with similarly strong t-values sitting close together is a real warning sign that the match is ambiguous rather than settled, exactly the kind of pattern a full search across offsets is built to surface. This sliding-window approach follows the same underlying logic used by dedicated professional software; the CROSSDATE program overview from the Laboratory of Tree-Ring Research describes its interactive graphic tools as built specifically to help dendrochronologists test a sample against many candidate positions in a master chronology before settling on one.

Narrow down a long list of tested offsets to a shortlist worth checking by eye, since visually inspecting every possible position by hand across a long master chronology is impractical; a ranked statistical shortlist is what makes that visual confirmation step feasible on chronologies spanning hundreds or thousands of years.

Accuracy and Limitations

The correlation, t-value, and GLK arithmetic here is exact given the ring-width index values supplied. This calculator does not perform ring-width standardization or detrending itself; it assumes you are entering already-standardized index values, the kind produced by curve-fitting software such as ARSTAN or the dplR Dendrochronology Program Library in R, rather than raw ring-width measurements, since raw widths carry an age-related growth trend that would bias a direct correlation. A statistical match, however strong, is also not a confirmed date on its own; the field's own practice, as described in dendrochronology laboratory documentation on crossdating by skeleton plotting, always follows a promising statistical hit with visual pattern confirmation before a date is formally accepted.

The Most Common Dendrochronology Dating Mistake

The mistake I see most often is treating a tree-ring date as automatically equal to the date an artifact or structure was actually made or used. A dendrochronological date tells you when a ring formed, and at best when the tree was felled if bark or the bark edge survives; it says nothing on its own about how long the wood then sat in storage, was reused from an earlier structure, or was carved from heartwood laid down decades before the tree died. With that in mind, always check whether a sample retains its outermost, bark-edge ring before treating a felling date as secure, and treat any sample missing that edge as giving only a "terminus post quem," a date after which the tree was felled, not a precise felling year. This distinction is central to what dendroarchaeology describes as the gap between a tree-ring date and the actual construction or use event; timber reuse, prolonged storage before use, and heartwood cut well before a tree's eventual felling all turn up most often in exactly the cases where nobody checked for the bark edge before publishing a date as final. Once a wood sample's date is established, our Radiocarbon Calibration Calculator relies on exactly this kind of dendrochronological data to build its calibration curve, and the Pollen Count Percentage Calculator can help place the same context within its broader environmental sequence.