Leaf Area Simulation Of Larix Olgensis Based On The Light Detection And Ranging In Montane Region Of Eastern Liaoning Province,China

Leaf area is an important parameter for the process of forest ecosystem and it affects the stand productivity and structure.In this study,we forcused on the Larix olgensis plannation which is the typical stand in montane region of eastern Liaoning province.Based on the characteristics measured in this area,254 branches’ leaf area, basal branch diameter, branch length and any more branch attributes were obtained by the branch analytical method. At branch level the models for leaf area and branch attributes were developed by unitary nonlinear regression and binary nonliner regression and multivariate nonlinear regression respectively.LiDAR is used to obtain the data of branch point cloud,and the branch attributes are extracted to vertify its accuracy.For the optimal estimation,we selected branches canopy leaf area modle and compared with the traditional methods to estimate leaf area.In the meantime,we analyzed the relationship between the crown position and leaf traits systematically.The main conclusions include:(1)The results showed that six branch attributes are significantly correlated to branch leaf area.The basal area is highly correlated to branch leaf area.The correlationships between branch leaf area and branch attributes are d2(0.827)>d2l(0.794)>d(0.785)>l(0.713)>DINC (0.497)>RDINC (0.419) successively.(2)In the research that build branch leaf area modle using branch attributes:the modle of y= 8.967/(1+50.901e-0.084d) was the optimal unitary non-linearity regress equation. It’s coefficient of determination (R2) was 0.719 and the precision was 86.34%(a=0.05).The coefficient of determination (R2) was improved significantly after we brought the variable of relative depth into crown (RDINC) into the equation. The modle of y= 0.002(d2)2.260e-1.701RDINC was the optimal binary non-linearity regress equation. It’s coefficient of determination (R2) was up to 0.796 and the precision was 88.57%(a=0.05).(3)The precision of base diameter value that extract base diameter from sample branch point cloud are 98.01% (a=0.05) and 97.16%(a=0.01).The precision of basal area using point cloud data in reverse accuracy test are 95.15%(a=0.05) and 93.06%(a=0.01)(4)The canopy leaf area is obtained by accumulating branch leaf area that take the base diameter value from point cloud into the optimal regress equation. The average relative error of this method is 14.61% compared with the true value. The average relative error of estimated canopy leaf area from diameter at breasr height and tree height is 43.46%. The average relative error of estimated canopy leaf area 197.88% conbined the model of leaf dry weight with the specific leaf area.(5)The leaf area is significantly affected by crown position. The leaf area in middle layers has significant difference among other canopies (p<0.05),and there is no significant difference between the upper and the lower layers.As canopy height increased,the leaf area increases first and then decreases.