Fusing tree ring and forest inventory data for ecological forecasting of tree growth responses to climate change

Authors: Kelly Heilman*, Kelly Heilman, R. Justin DeRose, Utah State University , John D. Shaw, US Forest Service, Andrew Finley, Michigan State University, Michael Dietze, Boston University, Jacob Aragon, University of Arizona, Andrew Grey, University of Arizona, Alex Arizpe, Gregor Mendal Institute, Vienna, Stefan Klesse, Swiss Federal Institute for Forest, Snow and Landscape Research WSL, Margaret Evans, University of Arizona
Topics: Global Change, Quantitative Methods
Keywords: tree rings, forest inventory, ecological forecasting, uncertainty components
Session Type: Virtual Paper
Day: 4/9/2021
Start / End Time: 11:10 AM / 12:25 PM
Room: Virtual 30
Presentation File: No File Uploaded


Forest responses to future climate are highly uncertain, but critical to understand if we want to manage the forest-climate feedback. To improve ecological forecasts of forest response, we harness the strengths of two large ecological datasets: 1) spatially extensive, unbiased, and representative forest inventory data and 2) a new collection of tree-ring time series data sourced from that forest inventory, which adds annually resolved growth responses. We fuse tree ring and forest inventory data using a Bayesian state space model. This model allows us to quantify the effects of climate, tree size, stand density, site quality, and all two-way interactions on tree growth, and their associated uncertainties. We forecast growth responses to an ensemble of projected changes in climate, and separate uncertainty surrounding these forecasts into different components: uncertainty about tree diameter at the start of the projection time period (initial conditions), uncertainty about future climate conditions (climate drivers), uncertainty about climate and other factors affecting tree growth (parameters), and residual unexplained ring width variability (process error). Combining national forest inventories and tree ring data is a promising frontier in dendroecology which allows us to forecast forest responses at large scales, quantify uncertainties, and improve our understanding of forest ecosystems and the terrestrial carbon cycle, as we move into an era of climate rep air.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login