A collaborative team of researchers led by Ben Weinstein from the University of Florida, Oregon, US, used machine learning to generate highly detailed maps of more than 100 million individual trees from 24 locations across the US, publishing their findings on July 16 .e in the open access journal PLOS biology. These maps provide information about individual tree species and conditions, which can greatly contribute to conservation efforts and other ecological projects.
Ecologists have long collected data on tree species to better understand a forest’s unique ecosystem. Historically, this has been done by surveying small plots and extrapolating these findings, although this cannot account for the variability across the entire forest. Other methods can cover broader areas but often struggle to categorize individual trees.
To generate large and highly detailed forest maps, the researchers trained a type of machine learning algorithm called a deep neural network using images of the canopy and other sensor data taken by plane. This training data covered 40,000 individual trees and, like all data used in this study, was provided by the National Ecological Observatory Network.
The deep neural network was able to classify the most common tree species with an accuracy of 75 to 85 percent. In addition, the algorithm could also provide other important analyses, such as reporting which trees are alive or dead.
The researchers found that the deep neural network had the highest accuracy in areas with more open space in the canopy and performed best at categorizing conifer species, such as pines, cedars and redwoods. The network also performed best in areas with lower species diversity. Understanding the strengths of the algorithm can be useful for applying these methods in different forest ecosystems.
The researchers also uploaded their models’ predictions to Google Earth Engine so their findings can inform other ecological research. The researchers add: “The diversity of overlapping data sets will promote richer areas of understanding for forest ecology and ecosystem functioning.”
The authors add: “Our goal is to provide researchers with the first large-scale maps of tree species diversity from ecosystems in the United States. These canopy maps can be updated with new data collected at any location. By working together Working with researchers at NEON sites we can make better and better predictions over time.”
