INS 7-6 - Probabilistic urban tree inventories with mixed-pixel imaging spectroscopy

Tuesday, August 16, 2022

3:30 PM – 5:00 PM EDT

Location: 520A

Presenting Author(s)

JO

Jonathan P. Ocón, M.A. Geography

Instructor / Ph.D. Candidate
UCLA, United States

Co-Author(s)

ES

E. Natasha Stavros, n/a

Director
Analytics Hub, Earth Lab, CIRES, University of Colorado, United States
JR

Justin Robertson

Los Angeles County Department of Public Health
SS

Steven Steinberg

Los Angeles County Information Services Department
TG

Thomas Gillespie

UCLA Geography

Los Angeles County manages at least 800,000 public trees, with inventory costs using traditional methods exceeding $5 million (Los Angeles County Department of Public Health). The ability to provide probablistic urban tree inventories using spaceborne imaging spectroscopy will optimize the planning and management of the County's urban forest without the need to comission costly airborne acquisitions. Spaceborne imaging spectroscopy data would also have an advantage over airborne acquisitions in monitoring seasonal changes using regular repeat observations, but there remains the issue of mixed-pixels in complex environments such as urban areas. There are a number of existing methods to unmix imaging spectroscopy data, however, due to coarse spatial resolution (> 30 m) of spaceborne missions and the complexity of urban environments, there is a need to develop a model to identify probabilistic species presence and condition within mixed-pixels.