Each growing season trees produce a new layer of wood under the bark; which varies in width from year to year depending on the climatic conditions (wet = wider, dry = narrower) – and this proportional width will be shared by all trees of the same species within that climatic zone (hence this technique is region-specific).  Scientists can count and measure annual ring widths to a very accurate degree, these measurements are analysed statistically to produce a regional reference pattern or sequence.  Any piece of wood from the present day backwards will usually overlap its tree ring pattern with an older piece – e.g. a tree felled in autumn 1945 may contain rings in its early (inner) pattern which will match to the outer (later) pattern of a tree felled in spring 1870.  That tree may match wood from 1790 and so on, into the prehistoric past. Many years of painstaking research has now compiled detailed sequences for many parts of the world, in Britain covering the period up to 5,500 years ago.

 

Patterns of seasonal variation are mostly unique and recognisable, so will provide a comparison for wooden artefacts, furniture, structural timbers in historic buildings, bridges and ships, or finds of wood from excavations.  In well-preserved samples, the date of felling can be refined almost to the week or day.

 

The drawbacks to dendrochronology are straightforward: it requires a substantial piece of wood with no less than fifty years’ worth of rings, so that the pattern can be established beyond doubt.  Some regions and periods (eg. the British Iron Age) have poor sequences and are therefore more problematic for this technique.  With the exception of waterlogged sites such as deeply-stratified deposits in towns such as London and York, wood from excavations is very often not well-preserved enough to use for dendro dating. But it may still be used for radiocarbon.  The role of dendro sequences in calibrating radiocarbon chronology is the other major contribution of this technique (see Radiocarbon).