Past climate from tree-rings: can Liebig’s law help?
Lead supervisors: Dr Tim Osborn
Location: School of Environmental Sciences, University of East Anglia
Duration: 8 weeks
Suitable undergraduate degrees: Any science or quantitative subject, Physical Geography
Project background
Tree-ring records are a key indicator of climate change during recent centuries. Of course, an individual tree’s growth is determined by a variety of factors, so samples are taken from many trees in a region. The standard approach is to date each individual annual growth ring and then to average the measurements (e.g. ring width) across all the trees that have a growth ring in a particular calendar year, under the assumption that average tree growth will be more closely linked to that year’s weather.
Stine (2019) recently proposed an alternative approach that may yield better reconstructions of past climate by applying Liebig’s “law of the minimum” to the data. Instead of averaging across all the measured trees, we discard the X% of trees that grew most poorly in any one year and use measurements only from the remaining trees (those that grew best that year). The argument is that these better-growing trees are still affected by the weather that year (which might be warmer, cooler, wetter or drier than normal) but not affected by other local stresses such as an insect infestation. The optimal choice for X is unknown and may depend on the characteristics of a region and the species of tree being measured.
This placement project will apply this approach to two very different case studies. The first case study is one of the longest tree-ring records in existence: more than 7000 years from Yamal in Arctic Russia. Here tree growth is sensitive to summer temperatures. The second case study is closer to home: 1000 years of oak tree growth from East Anglia, which is affected by spring and summer droughts.
For each case study, the tree-ring measurement data and the instrumental climate data will be compiled. Many alternative tree-ring records will then be constructed from these data, using both the standard approach and the Liebig approach for different values of X% from 10% to 90%. These will be statistically compared with the instrumental data and with other records to evaluate whether the Liebig approach can yield a record that has an even stronger link with the instrumental climate data.
There will be regular opportunities to discuss the work as it progresses with other staff and students in the Climatic Research Unit and in our climate datasets group. There will be time set aside for writing up the results of the data analysis including placing them into the context of existing literature.
The analysis can be carried out using spreadsheet software such as Excel, though there will also be time and support for the student to try coding the analysis with R or Python if they preferred.
Although all the necessary tree-ring data are already measured and available, there might be an opportunity to assist in fieldwork to collect additional oak data in East Anglia during the summer – however, arrangements to allow this to happen are not yet confirmed.
Stine A.R. (2019) Global demonstration of local Liebig’s law behavior for tree‐ring reconstructions of climate. https://doi.org/10.1029/2018PA003449