| Standsare complexbiological systemsthat are characterized by independentspatial structures and distributions, internal mechanismsandexternalenvironments. Scientifically and reasonably analyzing and predicting stand structuresand growth is very important in terms of providing guidelines for forestryproduction and scientific research.At the same time, because forest ecosystemschange slowly with characteristics of long time rotations and un-inversion, using traditional methods to study forest growth and management is often time-, manpowerand money-consuming. As forestry information technologiesdevelop, computer based simulation and prediction of forest structures and growth using visualization technology provides a novel approach to study forests, intuitively and really reveal forest growing process and spatial structures, and predict forest growth and yield. This will overcome the gaps that currently exists.In this study, Huang Feng-Qiaofarm of Hunan was selected as a pilot study and Chinese fir plantations were investigated. Forest stand structures were studied using stand structure adaptive simulation algorithm. Based onstand structures, a comprehensive competitionmodel was developed. By combiningthe model with tree mortality model, the partioning of forest biomass into individual tree growth was then explored to realize the growth simulation of individual trees. Finally, astand structureand growth visualization and simulationsystem was developed. The significant findings included:1)Developed a new forest inventory and survey method using the Total station and made it possible to layout square sample plots, set up reference points based on different terrains and topographic features, obtain tree locations, and re-produce the spatial configuration of measured trees.The results showed that this method is efficient and can be applied toforest inventory and survey of anyspecies.2) Stands are often characterized by self-similarity in terms of minimum spatial structure units. The simulation algorithm of stand spatialdistribution, pattern and structure was developed basedontheself-similarity theory and its combination with modified IFSfractal algorithm andangularscale model.This algorithm uses the idea that individualsform thewhole stand. It is assumed that eachminimumspatial structure unit is similar to the whoel stand. During the simulation, it is required that eachminimumspatial structure unit can reveal the real stand structure so that the simulated stand matches its truth.3) The competition among trees is important factor that impacts tree growing. The mechanism of the competition is the interaction between each other of plants. Currently the tree competitioniswidelyintroduced into the simulation of stand growth. This requires notonlyan integratedindex of competitionbut also a model that can be used to partition theoverall stand growthto thegrowths of individual trees. However, thedevelopment of most competition basedmodelsneedcollecting a large amount of field data. The work of field data collection is complex and diffifult.On the other hand, mostofthemodelsonly take into acount the competition from individual aspects, such as theneighboringcompetitionorlightcompetition.The development of the competition based stand growth models took into account the partitioning of stand growth and thus requires the models function in terms of both the overall stand growth and the growth of individual trees. Both the neighboringcompetitionand thelightcompetition have to be considered. Based on the requirements, this study used the data of tallied trees for Chinese fir plantation in Huang Feng-Qiaofarm of Hunan from 2008 to 2011 and explored and combined Open Comparison Model fO, Simple Competition Index Model fH, and Competition Pressure Index Model fCSI, which led to a new comprehensive Competition Index Model fCi=1/(afOi+bf Hi+cf CSIi).Using this competition index model, a stand growth model was obtained by partitioning the overall stand increment into the growths of each individual trees:?( )′==n1iciciffTotaliGG. Moreover, using regression analysis resulted in three parameters of the competition index model: a= 5.416081713,b= 1.037485025,c=0.157524494, with standard error of0.17912842 and R2 value of 0.966704535. The model was validated using 10 sample plotsand the results showed that this modelfit the data for Chinese fir plantations best, then the Liriodendron plantations, and the worst results were obtained for the mixed stands. This implied that this model worked for pure plantations well.4) By studyingstand structure, growth, and visualization and using human-computer interaction technology, a “tree-stand” LOD model was developed to realize stand growth visualization under different structures.Then, this study used the Direct X 3D technology and developed acomputer software in which stand structureadaptive model, process-based stand growthmodel, GIS data and so on were included. This software provided the potential for visualization of tree and stand growth. This software was applied to simulation and visualization of stand structures and growth under different stand ages for Chinese fir plantations in Huang Feng-qiao, Hunan and provided a reference for forest management.This study resulted in the followinginnovations:1) Based on characteristicsofdifferentterrainsand spatial distributionsof stands, a novel and station based forest survey method that does not need a closed traverse was developed.2) Based on self-similarity theory of stand structures and combining modified IFSfractal algorithm andangularscale model, the spatial configuration of trees was simulated by developing a spatial structure unit in which an object tree was centered with four neighboring trees competed with each other and then shifting and rotated this unit.3) Based on compettion of trees within stands, a comprehensive competition index model was proposed. The model took into account both light and neighboring competition and combined Open Comparison Model, Simple Competition Index Model, and Competition Pressure Index Model. Within the range of competition effect, the light competition and neighboring competition were weighted. The obtained competition model accounts for the various competition components in a new way.In a word, in this study tree competition and forest growth were modeled using competition model, stand growth partitioning model and adaptivealgorithm that simulatedstand structures. This approach was reasonable and provided a great potential for visualization of tree and stand growth and forestry information technology. |
PDF Full Text Request