Data and Code for Rapid Above-Ground Biomass Change Estimation Using Machine Learning for the Rohingya Refugee Camp & Surrounding Areas in Bangladesh

# Above-Ground-Biomass-Estimation
Last modified: 15 May 2025.

The codes have been implemented using Jupyter Hub and Google Earth Engine. Dewan Mohammad Enamul Haque, Dewan Ruksana Ruma, and Mashfiqur Shattique jointly updated the codes for multi-temporal above-ground biomass (AGB) estimation. 

Leveraging Sentinel 2A imagery, GEDI LiDAR biomass data, and ESA’s Biomass products, we estimated AGB for three key periods: 2017 (pre-refugee influx), 2019 (early forest restoration), and 2023 (restoration progress). We employed machine learning regression models, including Random Forest (RF), Support Vector Machine (SVM), and Gradient Boosting (XGBoost) regression. 

The original version of the code is available at https://spatialthoughts.com/2024/02/07/agb-regression-gee/ and https://eo4society.esa.int/resources/copernicus-rus-training-materials/

The GitHub link contains further instructions.