Carbon Stock Estimation and Biodiversity Assessment in Vietnam Forests using Remotely Sensed Data and Deep Learning Neural Networks
Research Objectives
In this research we intend to develop a new carbon stock estimation technique using satellite and drone imagery data integrated with machine learning and deep learning models. The aim is to create a robust methodology that can provide accurate and up-to-date information on carbon stock quantity over the forests in Vietnam. This approach will help in better understanding the carbon sequestration potential of forests, monitoring deforestation and forest degradation, and ultimately supporting this country to benefit most from the fast-growing carbon credit market. In the context of carbon credit, the targeted new level of accuracy of carbon stock estimation by this project will help in ensuring the credibility and effectiveness of carbon offset projects and verify the legitimacy of emission reduction claimed by projects. In addition, it will help enhance the integrity of the carbon credit market, enabling investors to make informed decisions, and driving sustainable investment in emission reduction initiatives.
Timeline
Year 1:
Month 1-3 Project Planning, team formation.
Month 4-7 Obtaining low resolution satellite data (Sentinel 1,2 and Int. Space Station) and data
preprocessing.
Month 7-10 Training the CNN and other machine learning models with the obtained data.
Month 11-12 Models testing and validation for the best combination of the input data.
Year 2:
Month 1-3 Obtain drone data over the selected three national forests (Lidar and Multispectral)
Obtain high resolution satellite data (MAXAR)
Month 4-7 Drone and satellite data preprocessing
Month 8-12 Training the CNN models with MAXAR high resolution satellite data and drone.
Patent filing.
Year 3:
Month 1-3 Models testing and validation.
Month 3-5 Finalize the models and select the best performing one.
Month 6-9 Prepare project reports, research papers, and presentations detailing methodology,
findings, and implications for carbon stock estimation.
Month 10-12 Conduct project evaluation and identify the next phase.
Patent filing.