Speaker
Description
Forest simulation input typically relies on data derived from conventional field inventories. However, field inventory surveys are often characterized by labor intensity, time constraints, and inherent scalability limitations. In contrast, recent advancements in remote sensing and digital technologies have significantly improved the ability to collect data in forested areas. In this context, integrating forest simulations with high-resolution data could lead to a methodological shift in simulating a range of management scenarios.
High-resolution point cloud data provides accurate information about current forest conditions and tree characteristics, which can be incorporated as inputs for simulations. We use the spherical camera Insta360 Pro 2 to acquire 360-degree video data in forest areas. Our photogrammetric pipeline generates a precise reconstruction of the 3D structure of the forest directly from the captured videos. This 3D reconstruction enables an accurate assessment of each tree's characteristics within a specified forest stand, thereby eliminating the need for traditional inventory methods.
The single-tree features (such as diameter at breast height and tree position) extracted from such high-resolution reconstructions can serve as a foundational input for forest growth simulations. This advancement has the potential to improve the precision of the simulations by providing detailed data for any given forest area, allowing us to simulate tree growth, competition, and mortality on a per-patch basis. Consequently, we could gain deeper insights into forest dynamics under various intervention scenarios and climate change conditions, strengthening the basis for adaptive management. Finally, by marking trees of interest virtually, we could enhance planning activities in forest operations.
| Keywords | 3D reconstruction; videogrammetry |
|---|
