John J. Battles, UC Berkeley
Meet: https://meet.google.com/niy-gtpk-sro
YouTube Stream: https://youtube.com/live/BLGqFvR01xM
Join group to receive calendar invite: https://groups.google.com/a/modelingtalks.org/g/talks
Abstract: Forests are a critical part of the biosphere’s life-support system, yet most forestry research is conducted by place-based scientists and managers. While many existing inventories deliver targeted, high-quality information, they too often rely on workflows that feature idiosyncratic design, bespoke analytics, and extensive human oversight. In this talk, I will share insights from the two-decade fire and fire-surrogate experiment at Blodgett Forest. A recent synthesis provided unique and relevant insights about management pathways that minimize the trade-offs between wildfire hazard, forest health, and carbon storage in a Sierra Nevada mixed conifer forest. I will also use it as an example to describe the data science challenges that reduce the informational return on investment -- uneven metadata, inconsistent protocols, and analysis workflows that do not generalize. More importantly, I will outline a plan that both supports place-based forest inventories and eases the extraction of “big insights” from small data.
Bio: John J. Battles is a Professor of Forest Ecology and the Rudy Grah Endowed Chair of Forestry and Sustainability at UC Berkeley. He is a field scientist engaged in long-term research of temperate forest ecosystems. His goal is to understand how and why forests change. Towards this end, his research seeks to understand the dynamic response of forest communities to disturbances and perturbations such as air pollution, invasive species, forest management, extreme drought, and fire. His recent work has focused on understanding the interactions among disturbances to assess their potential to reshape forests. The challenges of documenting these impacts have spurred an emerging effort to leverage the incredible investments of place-based forest inventory and monitoring. The goals are to gain big insights from these “small data” and to ensure that they provide empirically robust, benchmarks that support the next generation of forest research.