Study On Dynamic Monitoring Technology Of Forest Resource

"3S"technology-based methods for forest resources dynamic monitoring has the features of macroscopic, comprehensive, cycle short and low-cost and become the main direction of development of update forest resources dynamic monitoring technology. Probe for appropriate methods of forest resources dynamic monitoring technology, research the application of"3S"technology in forest resources dynamic monitoring technology, especially the most important two indicators for forest resources surveying, area and volume, discuss the rapidly and accurately prediction methods, it has important significance for speeding up the pace of forestry informationization construction, improve forestry management decision-making levels and promote the sustainable development of forestry.This study take Gaoligong Mountain Nature Reserve as the research area, using Indian remote sensing satellite and Landsat TM remote sensing data, research the area change of forest resources land use/land cover and the model of forest volume remote sensing estimation by data pretreatment, imagery interpretation and classification, overlying space data, build and optimize volume model, the main content and conclusion of this research is as follows:(1) Establish the digital elevation model DEM of the research area, and extract elevation, slope, aspect which used for build the model of forest volume based on DEM.(2) Fusion images about panchromatic image and multi-spectral image of IRS-P6 by IHS, the result indicate that the fusion image have higher spatial resolution, also includes rich spectral information.(3) Set up land use/land cover classification system relay mainly on forest type according to the《land use classification》,the characteristics of the research area and the image. Classificy by visual interpretation, the overall accuracy is 88.27 percent, acquired land use/land cover classification map in 1987 and in 2006, from the classification map can be seen the broadleaf's area is the largest, followed by farmland, the broadleaf mainly distributed in the Gaoligong Mountain Nature Reserve and the farmland distributed outside.(4) Analyse the changing statistics of land use/land cover area of the research area from 1987 to 2006, the results indicate: the broadleaf's and shrubbery's area have larger change, the broadleaf's area increased 1078.49hectares, the shrubbery's area decreased 1559.74 hectares; the uncovered rock and shrubbery have a larger change of the area percentage, the uncovered rock increased 72.34 percent, and the shrubbery decreased 23.56 percent.(5) Overlaying the two classification map, a convertion matrix of land use/land cover from 1987to 2006 is obtain, analysis the area and the type change of land use/land cover, the results indicate: the shrubbery's area have the largest divert from 1987 to 2006, divert area is 3105 hectares and mostly divert to broadleaf and farmland; the broadleaf's area have the largest supply from 1987 to 2006, supply area is 2567.27 hectares and mostly supply by shrubbery and farmland; the resident have the largest divert probability and supply probability, the divert probability is 74.68%, supply probability is 78.89%.(6) Example by broadleaf, build remote sensing estimation model of forest volume by one image of 2006. The scatter graph and the correlativity analyse indicate the volume per ha have linearity relation with the independent variable factor and estimation volume by multivariate linear regression model.(7) Optimize the initial model by multicolinearities diagnostics and the influence point detection, and make linear regression test, residual normal school test, equal-variance test, the test results prove the volume per ha have a stronger linearity relation with the independent variable factor, and the hypothesis of regress model is right. The coefficient of determine, the modificatory coefficient of determine and the correlation coefficients are 0.63, 0.527, 0.794, model have a good fitting effect.(8) check up the multivariate linear regression models by sample, the indexs of the check up show applicability of the model, the estimation accuracy P = 76.89% when confidence is 95%, satisfy the accuracy requirements of the Category II of forest inventory, the model is applicable for Gaoligong Mountain Nature Reserve.(9) This research pick up 996 broadleaf sub-compartments, all of the volume is 3750741.778 m~3. The volume is very large, so the entire forest resources in Nature Reserve is rich, which play important role in ecological balance, scientific research and social and economic benefits, the Nature Reserve should pay attention to protect forest resources and the method that estimate the forest volume using 3S technology is feasible.