AI-FORCAST

Forests play a critical role in mitigating the impact of climate change, acting as carbon sinks that absorb and store atmospheric carbon dioxide. In alignment with initiatives of the Kyoto Protocol, countries with positive carbon balance can sell their carbon credits to other nations. However, the basis for this is an accurate assessment of forest carbon stocks, which enables informed decision-making and the formulation of targeted conservation and restoration plans.

Traditional forest carbon assessment methods have often relied on labor-intensive field measurements and complex modeling techniques, which can be time-consuming and resource-intensive. AI-FORCAST investigates a promising alternative by applying remote sensing technologies and machine learning algorithms: Satellite imagery, LiDAR data, and ground-based sensor networks yield data in which deep learning methods estimate high-resolution spatial information on forest structure and composition. From this, carbon stock can be estimated more rapidly and accurately at a much larger scale. Particularly, a promising solution is to integrate remote sensing datasets with field-based sample forest inventories, enabling the mapping of extensive forest areas with greater accuracy and coverage.