南方典型红壤侵蚀区马尾松林立木生物量无人机遥感估测

Remote Sensing Estimation of Standing Tree Biomass in Pinus massoniana Forest in Typical Red Soil Erosion Area in Southern China

以南方典型红壤侵蚀区长汀县河田镇为例,结合无人机与激光雷达产生的点云数据优势,通过局部最大值和分水岭算法反演单木树高(H)和冠层半径(R_(c)),拟合以H和R_(c)为变量组合的异速生长方程,得到以新冠层参数为底的马尾松立木生物量模型。结果表明:提取树高的决定系数(R^(2))和均方根误差(RMSE)分别为0.93和0.49 m;计算冠层半径的R2和RMSE分别为0.88和0.64 m;估算立木生物量的R^(2)和RMSE分别为0.89和3.37 kg。本研究通过无人机遥感影像定量参数并构建的异速生长方程中,以组合(H+R_(c))为底的异速生长方程估测马尾松林立木生物量的精度较高,可以有效估测马尾松林立木生物量,可为南方典型红壤侵蚀区马尾松林立木生物量准确估测提供参考。

Taking the typical red soil erosion area in the southern Hetian Town,Changting County as an example,and combining the advantages of point cloud data generated by UAV and Li DAR,while later inverting the single wood tree height(H) and canopy radius(R_(c)) by local maximum and watershed algorithms,fitting the anisotropic growth equation with the combination of H and R_(c) as variables,and obtaining the the model of standing wood biomass of Pinus massoniana based on new canopy parameters.The evaluation accuracy results showed that the coefficient of determination(R^(2)) and root mean square error(RMSE) of extracted tree height(H) were 0.93and 0.49 m,respectively;the R^(2) and RMSE of calculated canopy radius(R_(c)) were 0.88 and 0.64 m,respectively;the R^(2) and RMSE of estimated standing wood biomass were 0.89 and 3.37 kg,respectively.In summary,this paper quantifies by UAV remote sensing image and constructed anisotropic growth equations,the anisotropic growth equation with the combination(H+R_(c)) as the base has a high accuracy in estimating the standing wood biomass of P.massoniana forest,and can effectively estimate the standing wood biomass of P.massoniana forest.This study can provide methodological reference for the accurate estimation of standing wood biomass in P.massoniana forest in southern typical red soil erosion areas.