本研究以立霧溪事業區為例,藉由多變值分析完成林地區劃,就林木經營區進行分級之結果,探究林地分級之適宜性,以供林地使用規劃或管理之參考。 因集水區可視為一均質之地景單元,本研究使用數值地形模型(DTM)資料,自動化產生集水區作為基礎單元,推導12種地文變數,以多變值分析之因素分析與群落分析,進行統計分析,將立霧溪事業區劃為自然保護區、國土保安區及林木經營區;另配合林務局之土壤分級資料,建置林地分級資料庫,為討論林木經營區之林地分級及林地利用之適宜性,採美國農業部林務署研發之生態系經營決策支援系統(EMDS)為工具,結合林地分級之階層式架構,建置多尺度林地分級決策支援系統,由ArcMap外掛之EMDS來進行邏輯判斷,分為不適宜及適宜二類。所完成之林地區劃及林木經營區之分級結果,期供為林地使用承受力評估及未來林地經營計畫之參考。
This study assesses the forest land classification with Liwoshe Working Circle by using multivariate statistics, and probe into the suitability of the forestland classification as the basis of forest management. By using watershed as a basic homogeneous landscape unit that generated automatically with Digital Terrain Model(DTM), this study assessed multivariate statistical method of factor analysis and cluster analysis with numerous estimated terrain variables include 12 main site indices, and divided Liwoshe Working Circle into nature protection area、land protection area and timber production area by consulting the analysis results. In other way, this study also established forestland classification database with the obtain results of soil-classification, and developed a Multi-scale Forestland Classification Decision Support System with hierarchical structure applied by Ecosystem Management Decision Support (EMDS) System, which was developed by the Forest Service of USDA. At last, the Multi-scale Forestland Classification Decision Support System divided forestlands into two sorts, unsuitable and suitable, as the final result by conducting a logic trial on ArcMap. This study assesses the forestland management and planning suitability that wishes to supply forestland carrying capacity and forest management consultation in future.