Modeling neighborhood association decay effects improves forest stock volume estimation using UAV lidar and optical...

Area of Science:

• Forestry
• Remote Sensing
• Geospatial Analysis

Background:

• Forest stock volume (FSV) is a key metric for forest carbon sequestration potential.
• Existing FSV estimation methods often overlook spatial neighborhood effects and distance decay.

Purpose of the Study:

• To develop and validate a novel FSV estimation method that accounts for neighborhood association decay (NAD).
• To quantify the NAD effect and optimize its integration into FSV modeling.

Main Methods:

• Developed a framework for expressing and quantifying the NAD effect in FSV estimation.
• Designed NAD models, determined optimal neighborhood sizes, and optimized NAD strategies.
• Utilized UAV LiDAR and optical remote sensing data for FSV estimation in Mengyin County, China.

Main Results:

• The NAD-based FSV model significantly improved estimation accuracy (R² = 0.75–0.96) compared to conventional pixel-based models.
• High spatial heterogeneity in FSV was observed across Mengyin County (15.47–242.82 m³/ha).
• The study identified significant forest carbon sequestration potential.

Conclusions:

• The NAD-based modeling approach effectively enhances the accuracy of forest stock volume estimation.
• Understanding spatial neighborhood decay is critical for precise forest resource assessment.
• The findings support improved forest management and carbon accounting strategies.