Studies On The Visualizations And Simulations Of Spatial Pattern In Natural Regenerative Forest Of Western Mountain Area Of SiChuan Basin

There are many factors impacted the plant's spatial distribution patterns, the most plants steady patterns are aggregated distribution pattern in the natural. Studies on the forest distribution pattern is the base to the inter/intra-specific of population and the 3D visualization of forests, and the forest 3D visualization is the one of the most foundation of the digital forest and the modern forest management. This paper studies on using Ripley's univariate L(d) and bivariate L₁₂(d) functions for the spatial distribution patterns, inter/intra-specific spatial associations and different Life-form associations in three plots of natural regenerative forest located in western mountain border of Sichuan basin are analyzed, then make used of the quadtree(QT) model and the geographical information system (ArcGIS) simulated the main trees' the distribution patterns in the plots and convert the attribution information, at last we using a new index called as approximation degree was defined in this paper to testing the simulated distribution patterns of tree species, the result is better, then within the software OnyxTree6 building the 3D model of main trees, finaly realized the trees and forestland 3D visualization. All conclusions can be summarized as follows:1. We found the result of judgement the main trees' spatial distribution patterns is the same as the result of using the methods of the variance/mean T testing and negative bivariate items K testing and the methods of Ripley'K function and L function, but the result we can see clearly main trees spatial patterns different changes with the distance in the plot regional scale we selected by using the L function.2. Ripley's univariate L(d) and bivariate L₁₂(d) functions are used for analyzed the Litsea moupinensis Lec.var.szechuanica (Allen) Yang et P.H.Huang and the Aralia elata (Miq.)Seem, making up six pairs according to the different age-classes in the same trees and the different age-classes in the different trees in the plot 3. the result show: they were significantly positive associations at the plot regional scale, except the Aralia elata (Miq.)Seem when the scale beyond 8m. this is tell us the major trees require the environment resources similarity.3. Communicated the principles of population ecology, community ecology, forestry, computer graphics, calculation of ecology and so on, we using the quadtree(QT) model and the tools of ArcGIS, making the simulated the spatial pattern of main trees in three plots came true, we found using this methods simulated the main trees (number great than 30) spatial distribution patterns in plots the results are accessible.4. The index called as approximation degree to texting the results' precision of the simulated the main trees. Approximation degree is used as describing the validity of simulated pattern of the main trees in the plots, which use the buffer of individual tree to model the horizontal tree crown projection zone and evaluate whether the simulated points of individual tree fall into the zone or not in the form of(?),where (?) ,λis the real individual trees inthe plot; d_y is the distance of the simulated points to the real trees; d is the average radius of real trees' crown. The proximity of spatial pattern between simulated trees and real trees is related closely to the individual number of tree species and aggregate intensity. This index was important principle and reality value.5. The main trees' attribution information converted to the simulated trees in the plots. Base on the geostatistcal principle, using the Kriging method in ArcGIS, we convert the real attribution values (height of trees, diameter in breath of trees, crown of trees, the height behind the first branch of trees, and so on) of the main trees to the simulated trees, so after do this the simulated trees have the attribution information. Then we calculated the major values (for example: means value of height, sum value of height, the maximum value of height, the minimum value of height, the standard variance of height, and so on ), compared the each values of the simulated trees and the realization trees, judgement the precision of simulated trees' attribution values converted.6. The main trees' attribution values and spatial distribution patterns obtained by automatic and semi-automatic with ArcGIS in the three plots. This paper got the simulation trees spatial pattern's attribution values and the simulated procession semi-automation by Model Builder Geoprocessing in ArcGIS Desktop ArcToolBox, the results is better, and it nearly reach our expectations of we hoping.7. Communicated the Type theory, Space self-similarity and Mathematical models of Raster of GIS, simulated the main trees' spatial distribution patterns in the region of the based on the main trees pattern which realized or simulated in plots, and the attribution values (height of trees, diameter in breath of trees, crown of trees, the height behind the first branch of trees, and so on) contained in the simulated patterns of the main trees in the region of the plot we studies in.8. Forest 3D visualization. Base on the famous 3D forest model software-OnyxTree6, we had modeled the main tree's 3D models, and created the forest 3D model style library in ArcGIS, then put the points layer of the main trees over the forestland grid of the DEM, made the 3D forest models instead the trees points in ArcScene or ArcGlobe, so we can see the real 3D forests on the 3D landscape.In conclusion, at the early succession stage, the main trees were aggregated distribution patterns, the inter/intra-specific and Life-form of major trees were significantly positive associations. Base on above conclusion, we used the principles of population ecology, community ecology, forestry, computer graphics, calculation of ecology and quadtree(QT) model comunicated the tools of ArcGIS, realized the forest and forestland 3D visualization, which is obvious significance for 3D visualization of forest management and digital forestry in the future.