Work

ICESat-2 data denoising and forest canopy height estimation using Machine Learning

Year: 2024

Type: article

Abstract: Supervised classification methods can distinguish between noise and signal in ICESat-2 ATLAS data across various feature perspectives and autonomously optimize parameters. Nevertheless, model generalization remains a significant limitation for practical applications. This study focused on developing a universal denoising model for ICESat-2 using machine learning algorithms and analyzing its spatial transferability under various forest and terrain conditions. A photon-denoising feature parameter system was developed based on the analysis of the three-dimensional distribution of photons in forested regions. This system reduces the parameters dependent on absolute physical quantities and increases those that are less influenced by terrain and forest features to enhance the model's transferability. Subsequently, automated machine learning algorithms (AutoML) were used for model selection and parameter optimization across six non-parametric regression models. We evaluated the accuracies of the local, global, and transfer models in estimating canopy height across four representative forested areas in China. Results show that the algorithm can effectively distinguish between signal and noise photons. The estimated canopy heights from signal photons were highly consistent with heights obtained using airborne laser scanning (ALS), exhibiting a Coefficient of Determination (R²) of 0.8 and Root Mean Square Errors (RMSE) of 3.75 m. Notably, the accuracy of canopy height estimation by the global model has increased by an average of 21% compared to ATL08. Furthermore, the model exhibited significant spatial transfer capabilities, with the accuracies of the transfer model exceeding those of ATL08 by margins ranging from 4% to 41%. This study marks a significant advancement in photon-denoising methodologies, providing a robust and transferable solution for large-scale environmental data analysis.

Source: International Journal of Applied Earth Observation and Geoinformation

DOI: 10.1016/j.jag.2024.104263

Authors: Dan Kong, Yong Pang

Institutions: Institute of Forest Resource Information Techniques, Chinese Academy of Forestry

Language: English

Topic: Remote Sensing and LiDAR Applications

Subfield: Environmental Engineering

Field: Environmental Science

Domain: Physical Sciences

SDGs: Life on land, Climate action

Open Access status: gold

Estimated APC paid: $2,820