Using machine learning on tree-ring data to determine the geographical provenance of historical construction timbers

Aug 4, 2026

Upload by Himbo

Dendroclimatology enables annual-resolution reconstruction of past climate, while wood from historical buildings can extend records back thousands of years. Uncertain timber origins, however, can alter the climate sensitivity of tree-ring samples. This study compares ring width and density from 143 living larch trees across seven elevational sites (1400–2200 m asl) and 99 historical ring-series to train and test supervised classification models for the European Alps.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages14

Reading Time35 min

AccessFree to download

Abstract

Introduction

Machine learning in tree-ring research

Linking climate and tree-ring proxies

Data and methods

Study sites and elevational transect

Tree-ring series and feature parametrization

Model training and evaluation

Supervised classification algorithms

Ring-width vs ring-density datasets

Results and implications

Best-performing model and site-specific performance

Applications to provenance studies

DOI: 10.1002/ecs2.4453 ARTICL E Emerging T e chnologies Using machine learning on tree-ring data to determine the geographical provenance of historical construction timbers Eileen Kuhl 1 | Christian Zang 2 | Jan Esper 1,3 | Dana F. C. Riechelmann 4 | UlfBüntgen 3,5,6,7 | Martin Briesch 8 | Frederick Reinig 1 | Philipp Römer 1 | Oliver Konter 1 | Martin Schmidhalter 9 | Claudia Hartl 10 1Department of Geography, Johannes Gutenberg University, Mainz, Germany 2Department of Forestry, University of Applied Science Weihenstephan-Triesdorf, Freising, Germany 3Global Change Research Centre (CzechGlobe), Brno, Czech Republic 4Institute for Geosciences, Johannes Gutenberg University, Mainz, Germany 5Department of Geography, University of Cambridge, Cambridge, UK 6Swiss Federal Research Institute (WSL), Birmensdorf, Switzerland 7Department of Geography, Masaryk University, Brno, Czech Republic 8Department of Information Systems and Business Administration, Johannes Gutenberg University, Mainz, Germany 9DENDROSUISSE-Labor für Dendrochronologie, Brig, Switzerland 10Nature Rings-Environmental Research and Education, Mainz, Germany Correspondence Eileen Kuhl Email: [eikuhl@uni-mainz.de](eikuhl@uni-mainz.de) Funding information Bavarian Climate Research Network (BayKliF); European Research Council, Grant/Award Number: 882727; German Research Foundation, Grant/Award Numbers: ES 161/12-1, HA 8048/1-1; Gutenberg Research College; SustES, Grant/Award Number: CZ.02.1.01/0 .0/0 .0/16_ 019/0000797 Handling Editor: Julia A. Jones Abstract Dendroclimatology offers the unique opportunity to reconstruct past climate at annual resolution and wood from historical buildings can be used to extend such information back in time up to several millennia. However, the varying and often unclear origin of timbers affects the climate sensitivity of individual tree-ring samples. Here, we compare tree-ring width and density of 143 living larch (Larix decidua Mill.) trees at seven sites along an elevational transect from 1400 to 2200 m asl and 99 historical tree-ring series to parametrize state-of-the-art classification models for the European Alps. To achieve geographical provenance of the historical series, nine different supervised machine learning algorithms are trained and tested in their capability to solve our classification problem. Based on this assessment, we consider a tree-ring density-based and a tree-ring width-based dataset for model building. For each of these datasets, a general not species-related model and a larch-specific model including the cyclic larch budmoth influence are built. From the nine tested machine learning algorithms, Extreme Gradient Boosting showed the best performance. The density-based models outperform the ring-width models with the larch-specific density model reaching the highest skill (f1 score = 0.8) . The performance metrics reveal that the larch-specific density model also performs best within individual sites and particularly in sites above 2000 m asl, which show the highest temperature sensitivities. The application of the specific density model for larch allows the historical series to be assigned with high confidence to a particular elevation within the valley. The procedure can be applied to other provenance studies using multiple tree growth characteristics. The novel approach of building machine learning models based on tree-ring density features allows to omit a common period between reference and historical data for finding the provenance of relict wood and will therefore help to improve millennium-length climate reconstructions. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2023 The Authors. Ecosphere published by Wiley Periodicals LLC on behalf of The Ecological Society of America. Ecosphere. 2023;14:e4453 . [https:

FAQs

Why is timber origin important for tree-ring based climate reconstructions?

Different and often unclear origins of construction timbers can change the climate sensitivity of individual tree-ring samples, reducing reliability of climate signals.

What data and geographic coverage does the study use?

The approach uses ring width and ring density from 143 living larch trees across seven sites along an elevational transect from 1400 to 2200 m asl, together with 99 historical tree-ring series for model training.

Which machine learning model performs best and what does it imply?

Extreme Gradient Boosting yields the best performance; density-based models outperform ring-width models, and the larch-specific density model reaches the highest skill (F1 score = 0.8). This supports high-confidence elevation assignment of historical series and can improve long-term climate reconstructions.

Aug 4, 2026

Upload by Himbo