Predicting global terrestrial biomes with the LeNet convolutional neural network

18 April 2022

Accurately predicting global coverage of terrestrial biome is one of the earliest ecological concerns, and many empirical schemes have been proposed to characterize their relationship. Here, the authors demonstrate an accurate and practical method to construct empirical models for operational biome mapping via a convolutional neural network (CNN) approach.