Authors: Fang Fang*, West Virginia University, Brenden McNeil, West Virginia University, Gregory Dahle, West Virginia University
Topics: Remote Sensing, Urban and Regional Planning, Environment
Keywords: Urban trees, Remote sensing, WordView, Multi-temporal image
Session Type: Poster
Start / End Time: 9:55 AM / 11:35 AM
Room: Lincoln 2, Marriott, Exhibition Level
Presentation File: No File Uploaded
To maintain sustainable urban environments, timely and accurate information on tree stress is crucial that affecting urban forest managers’ decisions such as tree pruning, removing stressed trees or planting new trees. In this study, we evaluated the potential of WorldView-3 images to discriminate tree health conditions at individual tree level. We digitized 2000+ trees and further extracted six sets of vegetation indices (VI) from WorldView-3 images captured on June 11th, July 30th and August 30th , 2017. All the VIs are significantly different among health condition classes, and we found that NDVI1 from the July image, where NDVI1 is defined as the normalized ratio of the red and first near-infrared bands, has the most potential to discriminate trees in three health conditions: good, fair and poor. It is also noticed that the variability due to phenology from June to August is relatively larger than the variability caused by health conditions. We suggest that this green down phenological patterns should be considered when using images from different months to discriminate tree health condition. We also used multi-year images from 2014 to 2017 to analyze the pattern of VI in terms of tree health conditions. We suggested that the anniversary date image can be used to track tree health conditions.