Close-range laser scanning in forests: towards physically based semantics across scales.

Area of Science:

• Forestry
• Remote Sensing
• Ecology

Background:

• Laser scanning provides dense, accurate 3D vegetation data from terrestrial, mobile, and aerial platforms.
• Transforming 3D laser scan data into ecological information is complex, unlike simpler empirical models used in airborne/space-borne approaches.
• Deriving complex variables like leaf area index requires understanding measurement physics and point cloud semantic labeling.

Purpose of the Study:

• To explore the potential of close-range laser scanning for revolutionizing 3D vegetation structure assessment.
• To demonstrate how physical information from laser scanning can be integrated with 3D radiative transfer models for improved data analysis.
• To advocate for a physically based approach to ensure robustness and transferability of vegetation structure analysis methods.

Main Methods:

• Utilizing dense and accurate 3D point cloud data from modern laser scanning systems.
• Developing quantified structural models for semantic labeling of point clouds, including stem and branch architecture.
• Combining physical information of the laser scanning process with three-dimensional radiative transfer models.

Main Results:

• Simpler variables like diameter at breast height are readily derivable.
• More complex variables necessitate a deep understanding of measurement physics and semantic labeling.
• A combined physically based approach using laser scanning data and radiative transfer models promises robust and transferable information retrieval.

Conclusions:

• Close-range laser scanning data holds significant potential for detailed vegetation structure quantification.
• Integrating physical scanning properties with radiative transfer models is key to unlocking advanced information retrieval.
• Collaboration with users is essential to refine structural concepts and variables for optimal use of laser scanning data.