Improved random forest algorithms for increasing the accuracy of forest aboveground biomass estimation using Sentinel-2 imagery

Xiaoli Zhang; Hanwen Shen; Tianbao Huang; Yong Wu; Binbing Guo; Zhi Liu; Hongbin Luo; Jing Tang; Hang Zhou; Leiguang Wang; Weiheng Xu; Guanglong Ou

Ecological Indicators (Feb 2024)

Improved random forest algorithms for increasing the accuracy of forest aboveground biomass estimation using Sentinel-2 imagery

Xiaoli Zhang,
Hanwen Shen,
Tianbao Huang,
Yong Wu,
Binbing Guo,
Zhi Liu,
Hongbin Luo,
Jing Tang,
Hang Zhou,
Leiguang Wang,
Weiheng Xu,
Guanglong Ou

Affiliations

Xiaoli Zhang: Key Laboratory for Forest Resources Conservation and Utilization in the Southwest Mountains of China, Ministry of Education, Southwest Forestry University, Kunming 650233, China
Hanwen Shen: Yunnan Institute of Forest Inventory and Planning, China
Tianbao Huang: Key Laboratory for Forest Resources Conservation and Utilization in the Southwest Mountains of China, Ministry of Education, Southwest Forestry University, Kunming 650233, China
Yong Wu: Key Laboratory for Forest Resources Conservation and Utilization in the Southwest Mountains of China, Ministry of Education, Southwest Forestry University, Kunming 650233, China
Binbing Guo: Key Laboratory for Forest Resources Conservation and Utilization in the Southwest Mountains of China, Ministry of Education, Southwest Forestry University, Kunming 650233, China
Zhi Liu: Key Laboratory for Forest Resources Conservation and Utilization in the Southwest Mountains of China, Ministry of Education, Southwest Forestry University, Kunming 650233, China
Hongbin Luo: Key Laboratory for Forest Resources Conservation and Utilization in the Southwest Mountains of China, Ministry of Education, Southwest Forestry University, Kunming 650233, China
Jing Tang: Key Laboratory for Forest Resources Conservation and Utilization in the Southwest Mountains of China, Ministry of Education, Southwest Forestry University, Kunming 650233, China
Hang Zhou: Key Laboratory for Forest Resources Conservation and Utilization in the Southwest Mountains of China, Ministry of Education, Southwest Forestry University, Kunming 650233, China
Leiguang Wang: Key Laboratory for Forest Resources Conservation and Utilization in the Southwest Mountains of China, Ministry of Education, Southwest Forestry University, Kunming 650233, China; Institute of Big Data and Artificial Intelligence, Southwest Forestry University, Kunming, China
Weiheng Xu: Key Laboratory for Forest Resources Conservation and Utilization in the Southwest Mountains of China, Ministry of Education, Southwest Forestry University, Kunming 650233, China; Institute of Big Data and Artificial Intelligence, Southwest Forestry University, Kunming, China
Guanglong Ou: Key Laboratory for Forest Resources Conservation and Utilization in the Southwest Mountains of China, Ministry of Education, Southwest Forestry University, Kunming 650233, China; Key Laboratory of National Forestry and Grassland Administration on Biodiversity Conservation in Southwest China, Southwest Forestry University, Kunming 650233, China; Corresponding author at: Southwest Forestry University, Kunming 650233, China.

Journal volume & issue: Vol. 159
p. 111752

Abstract

Read online

A simpler, unbiased, and comprehensive random forest (RF) model is needed to improve the accuracy of aboveground biomass (AGB) estimation. In this study, data were obtained from 128 sample plots of Pinus yunnanensis forest located in Chuxiong prefecture, Yunnan province, China. Sentinel-2 imagery data were applied to extract the important predictors of forest AGB, which were screened using the Boruta algorithm. We compared the fitting performance of two modified random forest models − regularized random forest (RRF) and quantile random forest (QRF) − with the random forest model. Moreover, we combined the smallest mean error of each quantile model as the best QRF (QRFb). The result showed: (1) Window sizes of 3 × 3 pixels and 5 × 5 pixels demonstrated greater sensitivity and suitability for estimating AGB than the 7 × 7 pixels window size. Enhanced vegetation indices derived from Red Edge 1 (B5) and Near-Infrared bands (B8A) were strongly correlated with AGB, indicating the heightened sensitivity of B5 and B8A bands to biomass and their potential in AGB estimation. (2) The RRF model outperformed both the standard RF and QRF in fitting performance, with an R2 of 0.56 and RMSE 57.14 Mg/ha. (3) The QRFb model exhibited the highest R2 of 0.88 and lowest RMSE of 29.56 Mg/ha, significantly reducing overestimation and underestimation issues. The modified RF regression supplies new insights into improving forest AGB estimation, which will be helpful for future research addressing carbon cycling.

Published in Ecological Indicators

ISSN: 1470-160X (Print); 1872-7034 (Online)
Publisher: Elsevier
Country of publisher: United Kingdom
LCC subjects: Science: Biology (General): Ecology
Website: https://www.sciencedirect.com/journal/ecological-indicators

About the journal

Abstract

Keywords