One of the fundamental scales of measurement in ecology is the individual, because individual organisms are a core biological unit that respond directly to environmental conditions and biotic interactions. Ideal data for large-scale ecology would provide information on every individual over large geographic extents, but individual-level data is difficult and expensive to collect. We are working to bridge this gap by combining newly available remote sensing and field data with innovative machine learning approaches to generate estimates of individual-level data on the location, size, traits, and species identity of individual trees at scales of 10,000+ ha using data from the National Ecological Observatory Network (NEON).
S Marconi, SJ Graves, D Gong, MS Nia, M Le Bras, BJ Dorr, P Fontana, J Gearhart, C Greenberg, DJ Harris, SA Kumar, A Nishant, J Prarabdh, SU Rege, SA Bohlman, EP White, DZ Wang. A data science challenge for converting airborne remote sensing data into ecological information. PeerJ, 2019.
Daniel M Perkins, Andrea Perna, Rita Adrian, Pedro Cermeño, Ursula Gaedke, Maria Huete-Ortega, Ethan P White, Gabriel Yvon-Durocher. Energetic equivalence underpins the size structure of tree and phytoplankton communities. Nat Comm, 2019.
Ben G Weinstein, Sergio Marconi, Stephanie Bohlman, Alina Zare, Ethan White. Individual tree-crown detection in RGB imagery using semi-supervised deep learning neural networks. Remote Sens, 2019.
Molly F Jenkins, Ethan P White, Allen H Hurlbert. The proportion of core species in a community varies with spatial scale and environmental heterogeneity. PeerJ, 2018.