請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/17045
標題: | 遙測技術應用於金門森林調查與監測 An Application of Remote Sensing Techniques on Forest Inventory and Monitoring of Kinmen Islands |
作者: | Shiang-Ting Cheng 鄭向廷 |
指導教授: | 邱祈榮(Chyi-Rong Chiou) |
關鍵字: | SPOT衛星影像,土地利用調查,地理資訊系統,林木蓄積調查,變遷偵測, SPOT Satellite Image,land use analysis,Geographic Information Systems,Inventory of growing stock,Change Detection, |
出版年 : | 2010 |
學位: | 碩士 |
摘要: | 衛星影像具有多光譜、週期性與數值化之特性,對於地面資源調查、土地變遷監測及管理等能提供即時且準確之資料。因此本研究採用SPOT-2衛星影像,藉由衛星影像配合土地利用調查與地理資訊系統進行處理與分析,以瞭解金門森林之狀態。
為了瞭解林木蓄積狀態,除了一般的蓄積量估算外,還藉由光譜資訊之各波段與林分參數進行迴歸分析,最終選用紅光波段進行分析(R2=0.75)。以一般推估法之金門林分總蓄積量為583,759 m3,而光譜資訊所推算之總蓄積量為516,467 m3。兩種估算方法相較之下,光譜迴歸式推算之像元材積分布,能清楚的反應出金門林木蓄積狀態,可有效的提供經營管理參考使用。 在森林狀態之監測方面,採用兩個不同時期但相同季節之衛星影像,經由影像處理後,利用常態化差異植生指數之特性進行差異性分析,選取出其差異值正值及負值最大之區域,並對這些區域進行檢核。其中狀態為「佳」之區域,多屬於目前生長良好、有植生進入或受到良好好護之區域。而狀態為「差」之區域,則有火燒、開發、葉子變色或林木死亡等現象。 Satellite images sustain great help for resource inventory, management and monitoring land change. It’s characteristics of multispectral, period and digital. It can also supply us the data immediately and correctly. In this research, SPOT-2 images are used for testing materials. Based on land-use map and geographic information system to realize the status of Kinmen forest. To understand the structure of forest, in addition to the general estimates method, but also assessment of growing stock can through satellite remote sensing technology and forest parameter. Finally, we chose the red-band to analysis(R2=0.75). Results show that the general estimation method of the total stand volume was 583,759 m3 and the spectral information calculated was 516,467 m3. Comparison of two estimation methods, the volume calculated by pixel can show the forest structure, and can provide a reference base for making forest management decisions. This study used two different periods but the same seasonal satellite images. After image processing, using the normalized difference vegetation index to identify the difference between two images, then selected from the largest difference value region, detection and explore the possible causes of occurrence. The status as 'good' means many regions are growing well and have new vegetation or well protected. The status 'bad' means some area have been burned, be developed, the leaves change color or tree death. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/17045 |
全文授權: | 未授權 |
顯示於系所單位: | 森林環境暨資源學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-99-1.pdf 目前未授權公開取用 | 4.86 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。