Abstract
Forest stock volume (FSV) is an important indicator for assessing the carbon sequestration potential of forests and is influenced by neighborhood environmental factors. However, most studies have disregarded the spatial neighborhood association of the modeled variables and its decay effect with increasing spatial distance in estimating the FSV. We propose an FSV estimation method that considers the neighborhood association decay (NAD) effect, that is, NAD-based FSV modelling, and constructed a framework for expressing and quantifying the NAD effect, specifically including the design of NAD models, determination of the optimal neighborhood size, and optimization of the NAD strategy. Finally, we estimated the FSV of the dominant tree species using UAV LiDAR and optical remote sensing data from Mengyin County, China, and evaluated the estimated results using field sample data. The results suggest that the proposed NAD-based model can effectively improve the accuracy of FSV estimation for each tree species (R(2) = 0.75 ~ 0.96) compared to the conventional pixel-based model. The analysis of the spatial distribution pattern of FSV in Mengyin County revealed high spatial heterogeneity of FSV (15.47-242.82 m(3)/ha), and a high potential for forest carbon sequestration was found with field surveys.