A little dendrochronology

The science of dendrochronology derives its name from the Greek words δένδρον (dendron, meaning tree limb) and χρόνος (khronos, meaning time). In a nutshell, the name means tree dating (Wikipedia). Many species of trees are known to develop annual growth rings, such as those in Figure 1. The outermost ring corresponds to the most recent year (the current year for a living tree, or the year when the tree died), and the innermost ring corresponds to the tree’s first year. Thus, by counting the number of rings, scientists can determine the age of a tree. Furthermore, the width of these rings are correlated to moisture, as the growth of a tree increases with higher rainfall. Therefore, by calibrating the ring widths with some environmental measurements, scientists can derive important information about the past climate, such as precipitation and drought.


Figure 1: An example of annual growth rings in a tree. Source: Wikipedia

Ring width measurements can be done without cutting open a tree. The equipment involved is called an increment borer, which helps scientists bore a small hole in the tree trunk and draw out a small core that contains a segment for each ring (Figure 2).  For a tutorial on how to use the borer, check out this video, and many others on YouTube. To increase accuracy, scientists usually need to take two cores from a tree, and sample from many trees at one site. After the cores are collected, the rings on each of them are measured, often automatically with a computer image processor. The result is a collection of ring width time series. Many factors affect the growth of a tree, some are due to the tree itself, and some are due to the environment. As a result, each ring width time series contains both endogenous and exogenous growth and thus needs to be standardized before use. Standardization involves three main steps (Cook and Kairiukstis, 1990, p. 104): (i) fitting a growth curve that represents the endogenous growth rate, (ii) for each year, dividing the observed ring width by the fitted value to get a dimensionless annual index, which is the ratio between the total growth and the endogenous growth, and (iii) using a statistical procedure to obtain a mean index time series across all trees at the site. The final result is called a tree ring chronology, which can then be used to derive information about past climate.

Dendrochronological drill hg.jpg

Figure 2: An increment borer and two cored tubes, with a ruler for size reference. Source: Wikipedia.

Tree ring data are available in the public domain on the International Tree Ring Data Bank (ITRD). An interactive user interface and a collection of standardized chronology can be found on the dendrobox project. Interested readers can find more details in Cook and Kairiukstis (1990). It is a classic textbook on dendrochronology, and is available in a the public domain here.


Cook, E. R. and Kairiukstis, L. A. (1990). Methods of dendrochronology. Applications in the
Environmental Sciences. Kluwer Academic Publishers.

Zang, C. (2015). Dendrobox – An interactive exploration tool for the International Tree Ring Data bank. Dendrochronologia, 33:31-33.

Photo credits

Figure 1: (Wikipedia) No machine-readable author provided, Arpingstone assumed (based on copyright claims). No machine-readable source provided, own work assumed (based on copyright claims). Public Domain, https://commons.wikimedia.org/w/index.php?curid=447580

Figure 2: (Wikipedia) By Hannes Grobe/AWI – Own work, CC BY-SA 2.5, https://commons.wikimedia.org/w/index.php?curid=1135047

Paleo-reconstruction of climate data

This is a new concept that I just heard on Thursday (5 May), and I have a strong feeling that I’m gonna embark on it.

The key idea is to reconstruct climate conditions of the past via proxies such as tree rings, speleothems (loosely speaking, those pointy structures that you often see in caves) and so on.

Why is this important to water resources management? In planning for the future, we always rely on past data. For example, to build flood control structures, engineers rely on past streamflow data. However, in most places, such data are available for only about less than a hundred years. If we want to predict something of a longer (e.g. multi-century) scale, that is not enough, mainly because extreme events are unlikely to be captured in such short period. So the rising trend in literature is to reconstruct hydrological data via proxies (as listed above). With this technique, extreme events in the past can be observed, and probability calculations can be done more accurately.

I’ll take a deeper look at some papers relating to this. I think this is quite fascinating.