The Seismic Fingerprint of Tree Sway

Changing climate, especially the increase in frequency and intensity of extreme events such as heat waves and droughts, places many forests under significant pressure. However, we lack methods to efficiently track stress responses of trees across large scales. Real-time monitoring of physiological and structural stress indicators of trees, for instance via sap flow, stomatal conductance, or photosynthetic activity are often expensive, require high maintenance, and are therefore not efficient on a larger spatio-temporal scale.

We propose to investigate whether the stress responses of trees can be approximated as a function of the seismic power generated by tree sway - referred to as the tree’s seismic fingerprint. These wind-induced sway signals are intrinsically linked to the material properties of leaves, branches, and trunks, which are influenced by changes in cell water content and corresponding turgor pressure. Seismic measurements offer scalability and low maintenance, making them viable for extensive long-term monitoring. Moreover, the data’s high temporal resolution provides detailed and characteristic sway frequency information that could be linked to tree individuals, species or traits.

Using complementary observations from ground-based seismometers and tree-attached accelerometers collected at the ECOSENSE site in the Black Forest, we successfully isolated and analysed the seismic fingerprint of tree sway through frequency analyses and signal correlations. We further integrated these sway data with direct tree traits and meteorological time series using machine learning techniques. We present the first results of this innovative approach, marking a significant step towards understanding the intricate relationship between tree motion and their immediate surrounding ecosystem.