Estimation Of Carbon Storage In Open Space Collaborative Forest Based On Laser Point Cloud

Forest is the largest terrestrial ecosystem on earth,is an important natural resource to maintain the ecological balance of the earth and ensure the global material and energy cycle,and plays an irreplaceable role in maintaining the earth’s ecosystem,social and economic development and human life.Forest carbon storage reflects the carbon sequestration capacity of forest and is an important index to evaluate the ecological function of forest.Traditional forest resource survey requires a lot of manual work,long cycle,high labor intensity and low investigation efficiency,which is difficult to meet the needs of modern forest monitoring.Li DAR technology can directly obtain the three-dimensional structure information of forest through active detection,which has great advantages in forest vertical structure detection.Li DAR technology has a broad application prospect in obtaining various forest structural parameters,estimating forest biomass and carbon storage.Aiming at the national "dual carbon" strategy,this paper focuses on the difficult problem of accurate estimation of large-scale forest carbon stocks,and uses the advantages of Li DAR technology in obtaining forest structure parameters to study the estimation method of largescale forest carbon stocks based on single wood scale,so as to solve the key problems of forest carbon stock estimation such as single wood segmentation and single wood parameter estimation.The specific research contents are as follows:(1)Aiming at the problem of poor segmentation effect of airborne Li DAR point cloud single tree,it is proposed to use point cloud density characteristics to improve the accuracy of single tree top detection.The recall rate of single tree detection is higher than 0.75,and the accuracy rate is maintained above 0.90.Then,the single tree segmentation method based on Angle analysis is adaptive improved to improve the integrity of single tree segmentation.The DICE coefficient of single wood segmentation was higher than 0.75 and greater than 90% of the total.;(2)In view of the difficulty in constructing growth models of individual trees of different species,the parameters of DBH,crown width and tree height of individual trees were extracted from the ground and forest point clouds of unmanned aerial vehicles,and the relationship between the parameters was analyzed.A fitting model was constructed to estimate DBH based on the tree height and crown width of individual trees.The coefficient of determination of each model was greater than 0.6,which laid the foundation for estimating forest carbon stocks in a large range.(3)Aiming at the bottleneck of difficult accurate acquisition of large-scale forest carbon storage,a technical method of high-precision mapping of large-scale forest carbon storage was proposed by using forest point cloud single-tree segmentation and single-tree growth model,combined with single-tree biomass estimation model and average carbon content coefficient.The feasibility of accurate estimation of forest carbon stocks based on single wood scale was proved through the verification of the sample plots.The root-mean-square error of the results of large-scale estimation of forest carbon stocks based on single wood scale was 2.12 tons per hectare.