GEOG3839: Introduction to Dendrochronology

Annual growth rings tell us much more than just a tree’s age. They also provide clues that help us understand how our environment has changed in the past, and provide insights into how key processes in atmosphere, biosphere and geological systems operate over long timescales.

Trees are among the oldest living things on our planet. Many trees can live for several centuries and a few exceptional specimens have survived for more than 5,000 years. In part because of their great age, the annual growth rings from trees provide us with an incredibly powerful and adaptable tool to study Earth’s history.

The science of dendrochronology uses information encoded into the annual growth rings of trees to address issues related to climate change, hazards, ecology and natural history. Because tree vitality is strongly influenced by local environmental conditions, major events such as a change in climate, insect attack or severe flood often create a distinct ‘fingerprint’ in the tree’s rings. By studying these signatures in the rings, we can develop an annual record of past environmental events extending back several centuries or millennia.

By the end of the course, students will understand how to collect, analyze, and interpret tree-ring data as chronological and environmental evidence. They will be able to explain both the key concepts underlying dendrochronology and discuss how tree rings are used to address contemporary issues in natural history, resource management and Earth Systems Science.

Last details on paper submission and course grades

posted Dec 5, 2019, 11:29 AM by Scott St. George

Today in lecture I'll pass on the final few details about the submission of your final project, but I thought I should repeat myself on the blog to avoid any possible oversights on my part.

Deadline Although a few students have already shared their final paper with me (thank you!), the deadline for its submission is next Tuesday, December 10. Please remember we will not meet in person next Tuesday, so all submissions must be sent to me via email ( So people have a bit more time to prepare their report, I would ask everyone to please send your report to me no later than 5PM (Central time).

Report format First, papers should be sent to me as a single PDF (combining all text, figures and references into one file). Second, please name your report in the following style: yourfullname.geog3839.pdf.

Evaluation and final grades I expect to grade papers during the week of December 16 and will return comments on your paper to your UMN email no later than December 19 (and likely will send them back a day or two earlier than that). In addition, I will also send you a summary of grades for each course activity and report your final grade. If you see I've made a mistake, either in recording your grade or in the sum, please let me know as soon as you can so I correct the error.

Class 22: Finale

posted Dec 4, 2019, 6:20 PM by Scott St. George

Tomorrow will be our final class meeting, so my main priority is that we take a step back and consider the value of the information we can recover from tree rings. As part of that discussion, we will conduct a brief in-class exercise that revolves around four recent scientific studies that included tree rings as part of their main line of argument. 

I'll describe the exercise in detail in tomorrow's class. For now, I'll only post PDFs of the four articles, as well as link to a news feature connected to each one:

Topic one: The effects of climate change on drought

Topic two: Unveiling the scope of early European trade routes

Topic three: Tracing the ecological impacts of the Second World War

Topic four: Finding the 'smoking gun' behind 6th century cooling

Class 21: Paleoflood hydrology

posted Dec 4, 2019, 6:06 PM by Scott St. George

In our penultimate class meeting, we discussed the ways in which trees and tree rings are able to preserve evidence of past floods. In addition to posting the lecture slides, I've also uploaded a short overview paper on the same topic that provided the main framework for our discussion.

No class meeting tomorrow (but drafts are still due)

posted Nov 25, 2019, 1:57 PM by Scott St. George

Several people have contacted me to ask whether it would be possible to submit draft research papers via email rather than in person. Because of the winter storm watch for Tuesday evening and its potential effect on holiday travel (, I have decided the following:
  • Our Tuesday class meeting is cancelled.
  • Everyone will submit electronic versions of their first drafts (as Word documents or PDFs) to me via email (to In addition to the drafts, people are welcome to also send along specific questions to me about their drafts and chosen research topics. 
  • Electronic drafts will be due the same time as outlined in our original schedule — the start of Tuesday’s class session (2:30PM).
  • I will return targeted individual feedback (and answer any questions you might also have) to everyone via email by end of day on Wednesday, November 27.
  • The deadline for the final paper will remain the same: December 10.
I’m sorry not to get a chance to see everyone before the holiday but hope this schedule change will give everyone the chance to get home before the storm hits. Happy Thanksgiving, and I’ll look forward to receiving your drafts in my inbox tomorrow afternoon.

Class 20: Dendroarcheology

posted Nov 25, 2019, 1:55 PM by Scott St. George

Archeological applications extend back to the very earliest use of dendrochronology, and archeological wood remains an irreplaceable resource for chronology extension as well as uses that link tree rings and human activity. In addition to the slides from lecture, I've posted the article (published in the somewhat obscure journal 'The Newsletter of the Wetland Archeology Project') that connected the Roskilde viking longship wreck to central Ireland.

Class 19: High-resolution radiocarbon dating

posted Nov 20, 2019, 10:29 AM by Scott St. George   [ updated Nov 20, 2019, 10:31 AM ]

This lecture was shorter than usual because of the two graduate presentations, but I hope it captured the reasons why the relatively new frontier area of high-resolution radiocarbon dating (when paired with dendrochronology) is so exciting. And in addition to the lecture slides, I want to share a few articles, one old and two new, that represent important milestones in the refinement of radiocarbon dating.

The first paper, by Minze Stuiver and Hans Suess in 1966, reported the relationship between radiocarbon ages (computed by measuring C14) and 'true' sample ages as represented by tree rings. The main point reported by the authors was that radiocarbon ages turned out not to be exactly the same as calendar dates, and were sometimes older or younger than the 'true' date by a few decades or centuries. For this reason, it was necessary thereafter to 'calibrate' or correct radiocarbon dates to estimate the real age of the material under analysis.

The other two papers lay out the newly-discovered 'spikes' in C14 production in the years 774 and 993 CE. The first paper described the discovery of these spikes in Japanese cedar trees, while the second used the spikes as markers to demonstrate that tree-ring chronologies from dozens of locations worldwide all had the correct dates (back to the 8th century AD) because they show the C14 spike in exactly the right year. As I said at the end of class, I think that result is a beautiful demonstration of the power of cross-dating and the care and effort made by so many people in so many labs around the world to correctly date each of their individual tree ring specimens.

Work due on November 19

posted Nov 18, 2019, 9:54 AM by Scott St. George

Hi everyone,

As reminder, tomorrow students are expected to submit or bring with them two pieces of class-related work. 

First, I will collect hardcopy versions of your response to Exercise 4 (Signal processing). Please remember to include your name on your submission, but if you don't have a stapler handy, I will bring mine to class.

Second, the first part of tomorrow's class will be reserved to discuss progress on our final papers. As part of that discussion, I asked everyone to please bring in their preliminary list of sources that will provide the foundation for your research project. That should include citations for 10 to 15 articles related to your subject. Please bring two printed copies -- one to share with a partner and a second to give to me (I will provide feedback on your provisional list through one-on-one discussion in class.

Besides that, we will have two presentations from graduate students enrolled in 5839 and I'll provide a short mini-lecture on the exciting frontier of high-resolution radiocarbon dating in dendrochronology. See you all tomorrow (in 220 Blegen).

Class 18: Signal processing II

posted Nov 7, 2019, 11:26 AM by Scott St. George   [ updated Nov 7, 2019, 11:29 AM ]

In today's lab session, we'll continue our experiments to estimate and remove the (confounding or trivial) effects of tree age and size on ring-width and produce composite tree-ring chronologies. And because at least a few people bumped into technical problems with the (user-hostile) research tool ARSTAN, I've modified the exercise so everyone has the opportunity to review our decisions and final products.

First, let's return to the final few steps of the exercise from last class. In session 1, I asked you to apply two different methods -- horizontal or arithmetic means and negative exponential curves -- to standardize or detrend tree-ring width measurements from northwestern New Mexico. And by this point, I hope everyone has had the opportunity to (1) download tree-ring measurements from the International Tree-Ring Data Bank, (2) get a copy of ARSTAN from the Lamont-Doherty Tree-RIng Laboratory and (3) try (or help another person) to use ARSTAN to produce chronologies from our New Mexico collections.

If you were able to cajole or wrestle ARSTAN into working properly, then you should have produced two separate chronologies from NM556. If ARSTAN refused to cooperate, then you may access those two chronologies on our shared Google Drive (a direct link is here). 



Question 1
Within the 'Tree-ring data' folder, I've uploaded plots that illustrate the application of the two different detrending methods (horizontal mean and negative exponential) to the same single tree-ring width measurement sequence from a single core within the NM556 collection. In the first example, explain briefly (one sentence) how applying the (arithmetic) mean of the measurement series changes or alters the detrended version (hint: the change is very modest). Second, examine the same data detrended using a growth curve estimated from a negative exponential function. How does the detrended series differ from the original (raw) data measured under the microscope?

Question 2
Now we can move one level 'up' to consider the effects of detrending choice on the final tree-ring chronology. In the same 'Tree-ring data' folder, I've created a Google Sheet with both versions of the NM556 chronology (the first produced by detrending the original measurements with a horizontal line, the other being constructed using negative exponential curves). 

First, download a copy of this spreadsheet to your computer or personal Google account. This step is important so everyone has the chance to answer the question. Then, please create a line graph that shows both chronologies as they change through time.In a sentence or two, describe how these two chronologies differ from each other. Based on your observations so far, which chronology would be likely to provide a better estimate of the overall growth trends of trees at this location, with the effects of age and size being removed? Please include a copy of your graph in your response.

Question 3
In the informal terminology of ARSTAN, a 'dirty dog' refers to the situation where the growth curve you've chosen is such a poor fit to your observed tree-ring measurements that the program cannot carry out the detrending. I've uploaded a screen capture of one such messy puppy that I encountered after trying to apply negative exponential curve to a ring-width series from the Abouselman Springs record. In one or two sentences, explain why the growth curve (chosen by me) is not physically realistic and therefore cannot be used to detrend this set of measurements.

Question 4
Finally, I've used ARSTAN to produce tree-ring width chronologies for nine collections from northwestern New Mexico. Please download a local copy of 'Tree-ring width chronologies, northwestern New Mexico'. Then I'd like you to make two graphs comparing this suite of chronologies. The first should show year-by-year changes in tree-ring width across all nine chronologies for the period 1900 to 2011. And the second should present the same records over a longer period: 1500 to 2011. Again, please include both plots in your response. In a short paragraph (5 to 7 sentences), please highlight examples when most or all chronologies exhibited either above average or below average growth, either for a single year or an extended period.

Please submit a hardcopy version of your report, answering these four questions, at the beginning of class on Tuesday, November 19. It is not necessary to include color graphics - black and white is fine.

Tuesday November 5 in 167 Social Sciences

posted Nov 5, 2019, 11:51 AM by Scott St. George

Hi everyone. A quick reminder that, to accommodate last week's discussion of the final project, we will meeting this week (today and Thursday) in 167 Social Sciences. See you there!

Class 16: Dendrogeomorphology

posted Nov 5, 2019, 10:44 AM by Scott St. George

I'm sorry to be late posting the visual aids from last Thursday's lecture on dendrogeomorphology. I've also included a copy of the classic 1978 paper by John Shroder Jr. that first presented the process-event-response framework that is now so common in this subdiscipline.

1-10 of 31