Skip to main content

Macroscale resilience

Across the Western U.S., an estimated 6.3 billion dead trees stand as evidence of increasingly severe and frequent beetle outbreaks, wildfires, and droughts, all intensified by decades of regional warming. Yet, a critical question remains: Are these forests transitioning to new ecological states, or are they showing signs of resilience?

Despite advances in resilience theory, measuring forest recovery, especially in response to compounding disturbances, remains a challenge. To bridge this gap, our research combines field-based forest inventories, uncrewed aerial systems (UAS, i.e., “drones”), the National Ecological Observatory Network’s (NEON) Airborne Observation Platform (AOP), and Landsat data. Our goal is to track plant functional types over time at a landscape scale, providing key insights into forest recovery.

By reconstructing forest dynamics at a spatial scale relevant to land management, this research delivers a first-of-its-kind temporal analysis of ecosystem resilience. We are also committed to developing an open, reproducible methodology that can address ecological questions at regional to continental scales, shaping future forest management strategies.

 

Macroscale disturbance legacies across the Western U.S. from 1984-2016, including (a) total number of years of severe anomalous summer drought (Palmer Drought Severity Index < -2), (b) the maximum extent of tree mortality by mountain pine beetle and spruce beetle, (c) burned area by wildfire, and (d) areas within the Western U.S. that experienced combinations of > 10 years of severe drought, bark beetle outbreak, and wildfire.

 

Example state variable history, with a range of initial state conditions prior to disturbance (dark blue), followed by a disturbance that initially changes the state by amount Δsi for some duration Δti (red) with abruptness Δsi / Δti. Following this initial state perturbation, the state may recover (green) over some period of time Δtr, or failure to recover resulting in a persistent state change (yellow). Note that following the end of the current data record, there will be increasing prediction uncertainty (light blue) that informs recovery probability estimates over a range of time scales. One dimension of state space is shown on the y-axis for clarity.