In this article we shall discuss the principles behind dendroclimatology. The reader may find it useful to re-read the article on dendrochronology before proceeding further.
How dendroclimatology works
You should recall from the article on dendrochronology that many species of trees produce annual growth rings. The dendrochronological method depends crucially on the fact that these rings will not be of constant width, but will be thicker when the tree in question has had a good year and thinner in bad years.
Now, from the point of view of a tree, the difference between a good year and a bad year is mainly going to be determined by the weather: by the amount of sun and the amount of rainfall in its growing season.
So by using dendrochronology we can put a date on a tree-ring, and by looking at the thickness of the ring we can find out about the climate in the year of its formation. This method is known as dendroclimatology.
Difficulties of the method
Some of the difficulties of dendroclimatology are, so to speak, inherited from dendrochronology: for example, a shortage of material limits the scope of dendoclimatology in time. Also, since annual rings don't form in the tropics, the use of dendroclimatology, like dendrochronology, is limited to temperate zones.
One difficulty with dendroclimatology is that although the weather is the most likely cause of a tree having a bad year, it is not the only possible cause: disease, fire, or other external causes can have the same effect. This difficulty is not insuperable: trees affected by such events will be the odd ones out, so analysis of a sufficiently large number of trees from different locations allows us to distinguish the climatic signal from the incidental noise.
Another difficulty is that we would expect both the temperature and the amount of rainfall to affect the growth of trees. We would ideally like a proxy for one or the other, whereas when we measure the width of a growth ring we are looking at the combined effects of both. One way around this is to study what are known as limiting stands. For example, if we look at trees near the snowline of a mountain, then because storms tend to break over peaks, those trees will get all the water they need; the limiting factor on growth will be temperature, which will therefore determine the width of the growth rings.
How do we know?
We can compare the growth rings in trees with records of temperatures going back in some cases hundreds of years.
Mostly, there is agreement between direct measurement and climatological data. However, there is a discrepancy between the most recent records and recent tree ring growth. As the reader will doubtless be aware, global temperatures have been rising over the last few decades, something we know by consulting actual thermometers; but the dendroclimatological proxy has not kept pace with the change in climate. This leaves a question-mark hanging over dendroclimatology. If it doesn't tell us about events which we know have happened, then what isn't it telling us about events that we don't know about?
So the same kind of data that tells us that dendroclimatology is somewhat reliable also tells us that it isn't completely reliable. Together with the other difficulties with dendroclimatology, you might think that climatologists should abandon it and rely on proxies which are based on something less capricious than the growth of trees. However, it does have its advantages: it provides us with data specific to particular locations on land which are not covered by other proxies.