Applying Double Sampling to Estimate Forest Volume Stock in North-east Taiwan

• Main Authors: Wang, Hsu-Chih, 王序之
• Other Authors: Wang, Chao-Huan
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/29123936253533040792

碩士 === 國立宜蘭大學 === 森林暨自然資源學系碩士班 === 105 === The main purpose of forest inventory is to estimate the total area and the volume of forest, and those are the basic elements needed in forest management. The accuracy and precision of the survey result will directly affect forest management planning and decision making, so forest managers should use the available budget to get the highest efficiency. The remote sensing has been using frequently in many kinds of researches. Comparing to the ground survey, we can use less human resources, time wasting and expenditure to get the main factors of the forest volume by using remote sensing. This study uses the data set of the 4th forest resources and land use inventory in Taiwan, and selects Luodong forest district and Hualien forest district as the study area, and results 493 records of them have both ground-plot data and aerial photo-plot data. Using stand height, crown width, crown closing, elevation and regeneration method as independent variables, this study applies regression analysis technique to fit model for predicting stand volume. Due to the amount of data is limited, using the dummy variable by combining artificial forest and nature forest, the data set after carefully selected has 19 records of broadleaf, 76 records of coniferous and 39 records of mixed stands. There are 134 records with both the ground-plot volume and the aerial photo-plot variables. This study shows there is no difference between artificial forest and nature forest in fitting volume equations, so combine these two forests for further analysis. In linear regression, stand height, crown width and elevation are the significant factors for predicting ground-plot volume.