Comparing remote sensing-based forest biomass mapping approaches using new forest inventory plots in contrasting forests in northeastern and southwestern China

TL;DR

This study compares remote sensing methods for mapping forest biomass in China, demonstrating that machine learning models trained with GEDI LiDAR data and optical/sar data produce more accurate high-resolution maps than existing global products, with faster computation.

Contribution

It introduces local models combining GEDI, Sentinel-1, ALOS-2, and Sentinel-2 data for high-resolution forest biomass mapping, outperforming existing GEDI-based products in accuracy and speed.

Findings

LightGBM outperforms Random Forest in accuracy and speed.

High-resolution maps show increased error with slope.

Models trained on local data generalize well to nearby regions.

Abstract

Large-scale high spatial resolution aboveground biomass (AGB) maps play a crucial role in determining forest carbon stocks and how they are changing, which is instrumental in understanding the global carbon cycle, and implementing policy to mitigate climate change. The advent of the new space-borne LiDAR sensor, NASA's GEDI instrument, provides unparalleled possibilities for the accurate and unbiased estimation of forest AGB at high resolution, particularly in dense and tall forests, where Synthetic Aperture Radar (SAR) and passive optical data exhibit saturation. However, GEDI is a sampling instrument, collecting dispersed footprints, and its data must be combined with that from other continuous cover satellites to create high-resolution maps, using local machine learning methods. In this study, we developed local models to estimate forest AGB from GEDI L2A data, as the models used to create GEDI L4 AGB data incorporated minimal field data from China. We then applied LightGBM and random forest regression to generate wall-to-wall AGB maps at 25 m resolution, using extensive GEDI footprints as well as Sentinel-1 data, ALOS-2 PALSAR-2 and Sentinel-2 optical data. Through a 5-fold cross-validation, LightGBM demonstrated a slightly better performance than Random Forest across two contrasting regions. However, in both regions, the computation speed of LightGBM is substantially faster than that of the random forest model, requiring roughly one-third of the time to compute on the same hardware. Through the validation against field data, the 25 m resolution AGB maps generated using the local models developed in this study exhibited higher accuracy compared to the GEDI L4B AGB data. We found in both regions an increase in error as slope increased. The trained models were tested on nearby but different regions and exhibited good performance.