State Key Laboratory of Subtropical Silviculture, Zhejiang A&F University, Hangzhou, Zhejiang 311300, China; Key Laboratory of Carbon Cycling in Forest Ecosystems and Carbon Sequestration of Zhejiang Province, Zhejiang A&F University, Hangzhou, Zhejiang 311300, China; College of Environment and Resource Sciences, Zhejiang A&F University, Hangzhou, Zhejiang 311300, China.
Jiangsu Key Laboratory for Numerical Simulation of Large Scale Complex Systems, School of Mathematical Sciences, Nanjing Normal University, Nanjing, Jiangsu 210046, China.
Sci Total Environ. 2023 Feb 20;860:160482. doi: 10.1016/j.scitotenv.2022.160482. Epub 2022 Dec 2.
Where and how many trees should be thinned in a pure managed forest to improve forest quality and increase ecological benefits are important forest questions. In this study we address such challenges by providing a novel framework for planning thinning operations through Unmanned Aerial Vehicle (UAV) remote sensing techniques, which can not only obtain forest attributes of its entire stand with spatial properties, but also optimize the selection of thinning areas, thinning intensities and cut-trees. This study helps to reduce the costs of time-consuming and laborious ground investigations. The framework was demonstrated by applying it into a subtropical Chinese fir plantation in southeastern China. Results showed that RGB images acquired by a low-cost UAV have great potential in depicting forest structure. The overall accuracy of the individual tree detection in the case study was 85.19 % ± 0.48 %. The overall accuracy and the intersection over union of the non-crown area extraction were 94.94 % and 82.65 %, respectively. For the two determined thinning areas, 19.5 % and 14.3 % crown density were required to thin in the primary and secondary regions, respectively. In addition, the top-down perspective of UAV remote sensing makes up for the limitations of the bottom-up perspective of traditional forestry. The framework can act as a basic model for forest managers to modify and expand for customizing detailed thinning guidelines.
在纯人工林管理中，应当在哪里以及砍伐多少树木以提高森林质量并增加生态效益是重要的林业问题。在这项研究中，我们通过使用无人机 (UAV) 遥感技术为规划间伐作业提供了一个新颖的框架来应对这些挑战，该技术不仅可以获取整个林分的森林属性及其空间特性，还可以优化间伐区、间伐强度和采伐树木的选择。该研究有助于降低耗时费力的地面调查成本。该框架通过在中国东南部的一个亚热带杉木人工林进行应用得到了验证。结果表明，低成本无人机获取的 RGB 图像在描绘森林结构方面具有很大的潜力。案例研究中单个树木检测的总体准确率为 85.19%±0.48%。非树冠区域提取的总体准确率和交并比分别为 94.94%和 82.65%。对于确定的两个间伐区，主伐区和次伐区的树冠密度分别需要降低 19.5%和 14.3%。此外，无人机遥感的自上而下视角弥补了传统林业自下而上视角的局限性。该框架可以作为森林经理人的基本模型进行修改和扩展，以制定定制的详细间伐指南。
