遙測技術應用於溪頭地區柳杉人工林之森林健康調查

Forest Health Inventory by Remote Sensing Technology for the Cryptomeria Plantation Forest in Chitou Area

衛星遙測之光譜資料，可大面積推估與森林健康有關之常態化植生指標（Normalized Difference Vegetation Index, NDVI）及葉面積指數（Leaf Area Index, LAI），為森林健康調查之重要資料。本文以前人研究所建立之林木健康指標調查表，進行溪頭地區柳杉人工林之地面健康調查，利用因素分析，建構柳杉人工林4種等級之森林健康度，調查資料經因素分析後，7個林木健康指標萃取出4個因素，分別命名為樹冠活力指標、根部損傷指標、活冠層指標及林木競爭指標。利用遙測資料所計算之NDVI與地面調查之LAI資料，建立兩變數之關係模式，藉由關係模式，繪製溪頭地區之柳杉人工林LAI分布圖，並以森林健康度調查樣區之點位進行套疊，以驗證NDVI、LAI與森林健康度之關係，結果顯示LAI超過3.00以上時，柳杉人工林分之健康度有趨於劣化之趨勢，而LAI分布介於2.51-3.00範圍之間時，樣區林分屬於健康等級中之等級2，唯因本次調查缺乏LAI值低於2.00之樣區，對於低LAI林分，其森林健康狀態，有進一步研究之必要。 The spectral characteristics of satellite images can be used in large areas for estimating normalized difference vegetation index (NDVI) and leaf area index (LAI). The NDVI and LAI are important indices for grading forest healthy. This study therefore uses factor analysis to establish tree health indictors in order to assess and to classify the tree health condition. It has extracted four common factors from seven health-related variables, and these factors have been named crown vigor indicator, root damage indicator, live crown indicator and tree competition indicator. This research was based on NDVI and LAI, it used to construct a simulating relation model for mapping LAI distribution map, and carries on the telescope by the sample health plots, and confirms NDVI, LAI and relations of the forest health. The result showed when LAI surpasses above 3, then health of the Cryptomeria plantation forest has tends the decline risk, but the LAI distribution is situated between 2.51-3 scopes the levels, the sample plots presents the healthy rank. In this research, only LAI lowers the region was non-forest healthy sample plots distribution. So NDVI and LAI between the forest health's relations needs to be further studied.

The spectral characteristics of satellite images can be used in large areas for estimating normalized difference vegetation index (NDVI) and leaf area index (LAI). The NDVI and LAI are important indices for grading forest healthy. This study therefore uses factor analysis to establish tree health indictors in order to assess and to classify the tree health condition. It has extracted four common factors from seven health-related variables, and these factors have been named crown vigor indicator, root damage indicator, live crown indicator and tree competition indicator. This research was based on NDVI and LAI, it used to construct a simulating relation model for mapping LAI distribution map, and carries on the telescope by the sample health plots, and confirms NDVI, LAI and relations of the forest health. The result showed when LAI surpasses above 3, then health of the Cryptomeria plantation forest has tends the decline risk, but the LAI distribution is situated between 2.51-3 scopes the levels, the sample plots presents the healthy rank. In this research, only LAI lowers the region was non-forest healthy sample plots distribution. So NDVI and LAI between the forest health's relations needs to be further studied.