OOS 52-4 - Forest insect outbreak monitoring and ecology across scales with harmonized field, drone, and satellite data

Bark beetles are a natural component of conifer forests in the western US that kill trees in a contagious process governed by both broad- and local-scale conditions. Broad-scale hotter droughts and past forest management can set the stage for increased host tree susceptibility to beetle attack, while individual host tree selection by individual beetles is dictated largely by local forest structure and composition. Interactions between these scales can initiate positive feedbacks and lead to mass tree mortality events like those observed in California’s Sierra Nevada mountain range between 2012 and 2016. Thus, measuring conditions at fine resolutions across a broad extent is key to refining our understanding of bark beetle ecology and identifying opportunities for management intervention. Collecting and analyzing fine resolution, broad extent data is typically labor-intensive and costly, but new tools (e.g., drones, lidar, high-resolution space-based sensors) and analytical approaches (e.g., machine vision-based photogrammetry, cloud-based workflows) can democratize access to this critical observational domain.

The Open Forest Observatory (OFO; https://openforestobservatory.org/) is a new, NSF-funded collaborative effort building cyberinfrastructure and accessible training for generating large forest inventories using drones and artificial intelligence. Our tools will allow ecologists to geolocate, measure, and identify species of individual trees across 10’s of hectares of conifer forest per day of field data collection using relatively low-cost, easy-to-fly drones and containerized workflows that can be run locally or on cloud computing resources (e.g., NSF’s Jetstream2 or Amazon’s AWS). To increase generalizability of the OFO, we will initiate the network with 150 drone-derived forest inventories parameterized for a variety of conifer forest types in the western US, and will provide a means for new users to contribute their own inventories to the network.

Vibrant Planet (https://www.vibrantplanet.net/) is a public benefit corporation with a mission to increase ecosystem and community resilience to climate change, particularly in disturbance-prone areas of the western US. We combine geospatial data science and deep ecology domain expertise to help decision makers collaborate and prioritize management action. As part of this effort, we are harmonizing diverse data streams to locate, measure, and monitor the health of individual trees across the entire western US in an automated, standardized way to help understand the drivers and consequences of disturbances including bark beetle outbreaks.