Dissertation > Excellent graduate degree dissertation topics show

The Research on Forestry Structural Factor by High-Resolution Remote Sensing Image

Author: ZuoGuangPing
Tutor: LinHui
School: Central South University of Forestry Science and Technology
Course: Forest Management
Keywords: data of forest structure model ridge trace selection procedure main component analytical method method of least square
CLC: S771.8
Type: Master's thesis
Year: 2011
Downloads: 124
Quote: 0
Read: Download Dissertation

Abstract

In the thesis, the studying area was set in Huangfengqiao forestry of Zhuzhou city, Hunan province. During the research, consider the GIS factors how influence the structures of forest and select the factors by the means of Model Ridge Trace Selection Procedure and Main Component Analytical Method to decide which will be effect in the Model of Forest Structures. The main results of research as follow:1) The total number of plots is 116。The data of collected in the plots concludes: geographical coordinate, soil type, soil depth, status of forest, height and DBH of tree, coverage, canopy size of tree and so on..4) The coefficients, determination coefficients and amendatory coefficient of Average of Height Model are 0.885,0.821, and 0.784, respectively. The coefficients, determination coefficients and amendatory coefficient of Average of DBH Model are 0.865,0.806,0.775,respectively. The results show that the two model both achieve the aim.5) The results of the testing accuracy of the model show that the all the models had good estimate effects. The accuracy of the height model is 82.1% and 83.4% according to the six trees gauge and enumeration with angle gauge. The accuracy of the DBH model is 79.2% and 81.5% according to the six trees gauge and enumeration with angle gauge.2) Extract the GIS and RS factors. The total number of Initialize variables is 16. They are elevation, aspect, slope, coverage,Bo、B1、B2、B3、B0/B3、B1/B2、B2/B0 EVI、NDVI1、NDVI1、NDVI2、RVI1、RVI2. Some of these factors influence the Models of Forest Structure so little that they can be eliminated from the Model. The elimination of factors is processed by Model Ridge Trace Selection Procedure and Main Component Analytical Method. The factors that needed to establish the average height of tree Models are B0、B1、B2、B0/B3、B1/B2、EVI、RVI2、slope and coverage. The factors The factors that needed to establish the average DBH of tree Models are B0、B2、B0/B3、B2/B0、EVI、RVI2, aspect and coverage.3) Selected data of 58 plots to establish the Model by method of least square.The Model of Average Height of Plot is:Y=11.25+16.2452ybd+0.0443aspect+0.0021B0-0.1524B1+0.2462Bo/B3-1.2031 B1/B2+0.00156 EVI-0.0426 RVI2+0.0014slopeThe Model of Average DBH of Plot is:Y=14+8.2454ybd-0.0025ele+0.0571aspect+0.0034Bo+0.2462B2-0.89368Bo/B3-0 .0.1786 B2/B04) The coefficients, determination coefficients and amendatory coefficient of Average of Height Model are 0.885,0.821,0.784, respectively. The coefficients, determination coefficients and amendatory coefficient of Average of DBH Model are 0.865,0.806, and 0.775, respectively. The results show that the two models both achieve the aim.5) The results of the testing accuracy of the model show that the all the models had good estimate effects. The accuracy of the height model is 82.1% and 83.4% according to the six trees gauge and enumeration with angle gauge. The accuracy of the DBH model is 79.2% and 81.5% according to the six trees gauge and enumeration with angle gauge.Paper was supported by "high-resolution earth observation system of major projects (E0305/111202)" and forestry research & special public service sectors "stand structure and growth simulation studies (201104028)".

Related Dissertations

CLC: > Agricultural Sciences > Forestry > Forest Engineering and forestry machinery > Forest measurement, forestry survey > Forest remote sensing
© 2012 www.DissertationTopic.Net Mobile