Aboveground Biomass Estimation in Tropical Forests: Insights from SAR Data—A Systematic Review

Journal of the Indian Society of Remote Sensing, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 17, 2025

Language: Английский

---

SCE-BiLSTM: A Hybrid Deep Learning Model for Regional Forest Biomass Estimation with Spatial-Channel Attention and Extreme Learning

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: March 17, 2025

Aboveground biomass (AGB) is crucial for ecosystem monitoring, forest surveys, and management. Accurate efficient AGB estimation remains challenging, as large-scale machine learning methods often sacrifice accuracy, while deep models enhance precision but struggle with efficiency generalization. To address this, we propose an advanced framework (SCE-BiLSTM) regional inversion, integrating spatial (SAM) channel attention mechanisms (CAM) to improve feature extraction. An extreme (ELM) enhances by randomly weights thresholds. Using 11 remote sensing features from Luoyang forests GEDI L4A data, the model outperforms CNN-BiLSTM, reducing MAE 3.59 Mg/ha, RMSE 6.46 increasing R² 0.9052, runtime reduced 19 seconds. Validation in Yellow River region shows strong generalization, achieving of 11.48 14.72 0.8335. A time-series analysis 2015 2023 reveals temporal variations, highlighting influencing factors. These results demonstrate framework’s potential accurate, scalable assessments, providing valuable insights sustainable

Language: Английский

---

Forest Aboveground Biomass Estimation in Küre Mountains National Park Using Multifrequency SAR and Multispectral Optical Data with Machine-Learning Regression Models

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(6), P. 1063 - 1063

Published: March 18, 2025

Aboveground biomass (AGB) is crucial in forest ecosystems and intricately linked to the carbon cycle global climate change dynamics. This study investigates efficacy of synthetic aperture radar (SAR) data from X, C, L bands, combined with Sentinel-2 optical imagery, vegetation indices, gray-level co-occurrence matrix (GLCM) texture metrics, topographical variables estimating AGB Küre Mountains National Park, Türkiye. Four machine-learning regression models were employed: partial least squares (PLS), absolute shrinkage selection operator (LASSO), multivariate linear, ridge regression. Among these, PLS (PLSR) model demonstrated highest accuracy estimation, achieving an R2 0.74, a mean error (MAE) 28.22 t/ha, root square (RMSE) 30.77 t/ha. An analysis across twelve revealed that integrating ALOS-2 PALSAR-2 SAOCOM L-band satellite data, particularly HV HH polarizations significantly enhances precision reliability estimations.

Language: Английский

---

Forest Biomass Assessment Using Multisource Earth Observation Data: Techniques, Data Sets and Applications

Journal of the Indian Society of Remote Sensing, Journal Year: 2024, Volume and Issue: 52(4), P. 703 - 709

Published: April 1, 2024

Language: Английский

---