From Forests to Farms: Quantifying Ecosystem Services in Managed Landscapes using Remote Sensing and Machine Learning

This defense will be presented in person. Join us in Bren Hall 1424 or watch online using this link and passcode farms

ABSTRACT

Ecosystem services underpin human well-being, yet they remain difficult to measure in a consistent and scalable way. This challenge is especially pronounced in working lands, where interactions both within these systems and across broader landscapes can alter these services in unpredictable ways. To address this, my dissertation develops and applies remote sensing and machine learning approaches to better quantify these services. In one vein of research, I leverage multi-sensor fusion to understand how carbon is stored in Chilean forest plantations, and how this changes over time. In a separate vein of research, I use dense time series derived from Doppler weather radar to monitor bat activity across agricultural regions at fine spatial and temporal resolutions. I pair this data with historic agriculture and pesticide use data to estimate the value of bat-mediated pest control services in California's Central Valley. Together, these studies demonstrate how data-driven approaches can advance our understanding of ecosystem services in working lands and support conservation and sustainable land management.

BIO

Brian is a landscape ecologist whose work bridges ecology, remote sensing, and data science to understand how human and natural systems interact. His research leverages machine learning and multisensor ground-based and satellite data to quantify ecosystem services across working landscapes—with projects ranging from tracking carbon storage in managed forests to valuing bat-mediated pest control in agriculture. Guided by a commitment to sustainability and applied solutions, Brian integrates causal inference, deep learning, and large ecological datasets to generate insights that highlight the value of ecosystem services and support more just, sustainable, and resilient landscapes.