Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/66131
標題: 地理資訊系統在森林經營管理上之應用 ── 以自然保護區為例
The Application of Geographic Information System to Forest
作者: 陳朝圳
Chen, Chaur-Tzuhn
關鍵字: Geographic Information System;地理資訊系統;Forest Management;Nature Reserve;森林經營管理;自然保護區;遙感探測;數值地形模型
出版社: 森林學系
摘要: 
本論文係以自然保護區為例,探討地理資訊系統如何應用於生態資源調查
、分析與自然保護區之經營管理。其結果摘要如下: (一)、地區性之地
理資訊系統應優先設立。本研究完成自然保護區之資訊系統架構。 (二)
、數值地形之建立以距離平方倒數加權平均法內差成網格寬度 40m×40m
時,其效果最佳,本研究發展坡度、坡向、水份梯度、日輻射潛能及全天
光空域等物理環境因子之空間性資料推估。 (三)、應用數值地形模型進
行天然林之衛星遙測影像地形常態化處理,結果 SPOT 遙測影像以二階段
統計糾正法效果最佳,LANDSAT 遙測影像則以非藍氏反射模式糾正法最佳
。 (四)、以 ARC/INFO 地理資訊系統配合自行設計之電腦程式,建立台
灣穗花杉自然保留區及霧頭山自然保護區之生態資料庫。 (五)、利用地
理資訊系統建立立木之空間及屬性資料,文中討論台灣穗花杉族群之生育
地分析、立木空間分布及生長分析,並推導出台灣穗花杉族群於該保留區
之生態最適界範圍。 (六)、山羌之棲息地分析可知地面調查資料所建立
之評估準則,為應用地理資訊系統推估空間性資訊不可或缺之過程,而地
理資訊系統則為資料整合之有效工具。 (七)、以 59 個地面樣區之植群
及環境因子資料,配合該地區之地理資訊系統,進行分類其結果可知,植
群分類之準確度為 71.4%,Weibull直徑分布 B介量為 60.7%,C 介量為
67.5%,而植群歧異度為

This paper, taking nature reserve as an example, explores how
to apply GIS to the analysis and inventory of ecological
systems as well as to the general management of nature
reserves. The summary of the obtained results is as follows:
As the locality-oriented analysis of information system in
nature reserve shows, local GIS should be first built because
one nature reserve should lay a great emphasis on different
stages in management planning operation in order to
correspond with the request of efficiency and cost, this
study has established a structure of information system for
nature reserves. According to the accuracy test and efficiency
analysis of DTM and the focal point of the research is
on interpolation treatment, the best result can be
obtained if grid elevation interpolation is proceed by
the method of weighted mean of distance square and 40x40
m grid width. Using DTM precede normalization of
topographic effect of satellite remote sensing imagery in
natural forest through three normalization techniques. The
result indicates that the application of the Two-stage
statistic method for SPOT imagery and that of Non-Lambertian
reflectance model for LANDSAT imagery are the most
successful technique respectively for reducing topographic
effects. The installation of auto-camera can collect animal
data from sample plots. By making use of forest
classification data and digital terrain model information
provided by remote sensing imagery in GIS, matched with
Fuzzy set theory, the habitat analysis of Munticaus reeve
micrurus can be obtained. By matching GIS with relational
models of ecological variables and environmental factors, can
be used to deduce the spacing of ecological information.
And the accuracy of vegetation classification is 71.4%,
B and C parameter of Weibull diameter distribution are
60.7% and 67.5% respectively, and vegetation diversity is
62.4%.
URI: http://hdl.handle.net/11455/66131
Appears in Collections:森林學系

Show full item record
 

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.