透過您的圖書館登入
IP:54.205.116.187
  • 學位論文

應用空載光達於阿里山地區林冠孔隙分類

Using Airborne LiDAR to Classify the Forest Canopy Gap in Alishan Area

指導教授 : 陳朝圳

摘要


森林生態系經營有賴於林分結構之資訊,而林冠孔隙的形成在林分結構中扮演了重要角色。森林因自然干擾因素致使林木死亡使林冠疏開,造成林間孔隙,且足以營造出不同於原森林生態系之生育環境,稱之為林冠孔隙,簡稱為孔隙。林冠孔隙會因更新的過程及時間的變化由其他林木更替,此種森林演替過程稱之為林冠孔隙動態,林冠孔隙之監測為森林生態系經營不可或缺的工作。傳統上地面調查林分資料,需耗費大量人力、物力、與時間,不易迅速獲得林冠孔隙資料。近年來使用遙感探測技術,如利用被動式高解析力影像,配合主動偵測式的光達資料,將可以節省許多人力資源。目前國內外有許多研究,應用光達資料進行森林經營之應用,但是近10年以光達資料,進行林冠孔隙研究卻較少。空載光達因具有穿透樹冠間隙之特性,可迅速獲得林分三維結構之資料,適合於林冠孔隙之分析與探討。本研究以阿里山試驗樣區所調查之林冠孔隙資料,配合光達技術探討林冠孔隙面積大小、周長、林冠孔隙內植生高度與環境因子,進行林冠孔隙繪製與分類。研究結果顯示,以光達所產生之1 m x 1 m網格解析力為最適網格,採取二值化分類法判釋林冠孔隙模型,準確度可達91.33%,kappa值為0.83,以二值化判釋林冠孔隙區域為最快速之方法。林冠孔隙依照地真資料區分為自然植生 林冠孔隙與人為干擾林冠孔隙,以實地調查資料進行判別分析分類,結果顯示19組光達資料分類準確度為84.21%,kappa值為0.68。本研究利用集群分析之華德法進行分群數的判定,以K-means進行123個光達林冠孔隙資料的分類,結果顯示最佳分群數為2,其中人為干擾林冠孔隙有8個,自然植生之林冠孔隙有115個。阿里山試驗樣區中,因植生覆蓋物種相似,高度變異不大,所以林冠孔隙內植生高度於分類中並不顯著,其與傳統航照判釋林冠孔隙相比較,光達資料確實可以偵測林冠孔隙內之植生高度,對於林冠孔隙動態研究有所助益。

並列摘要


Forest stand structure and canopy gap are important information for forest ecosystem management. “Canopy gap” were caused by trees dying under natural disturbance that were adequate for building a totally different habitat environment. Canopy gap would be replaced by other trees during the process of succession, and this phenomenon is called “gap dynamics” that is a process for monitoring of forest ecosystem management. Usually ground survey consumes a great deal of manual labor, material and time that is inefficient to gather information for canopy gap. With the improvement of the remote sensing technique and the usage of high resolution images and laser scanning systems that have been saved many human and material resources. Over the past 10 years, there are many studies applying the LiDAR system for forest measure, but there is fewer on canopy gap. The canopy gap can be penetrated by Airborne LiDAR system and to access the characteristic of three-dimensional information that is suitable for the study of the gap size and crown delineation. In this study, the type and location of canopy gap will be classified by the gap characteristics that include the size and perimeter, vegetation height and terrain slope in Alisan. The results show that the 1 m x 1 m LiDAR raster grid size and binarization method are the best method to identify canopy gap. The accuracy and kappa value of the identification of gap are 91.33% and 0.83 respectively. 19 canopy gaps would be classified into different type of natural vegetation and artificial interference by discriminate analysis, the accuracy and kappa value are 84.21% and 0.68 respectively. Otherwise, cluster analysis would be used to classify 123 gaps into different type and forming clusters by Ward’s method and used k-means to proceed with gap classification. The partitioning results into 2 clusters yielding clusters of 8 artificial interferences gaps and 115 natural vegetation gaps. In Alishan area, the vegetation species and height of gap are not significant, so that variable of gap vegetation height is not useful in the process of gap classification. We compared with the data of aerial photos and LiDAR, the LIDAR data could be used to detect the vegetation height of canopy gaps that is useful for the study of gap dynamics.

參考文獻


魏浚紘 (2008) 應用空載光達推估阿里山地區柳杉人工林林分材積。國立屏東科技大學森林系碩士論文,92頁。
張瑞文 (2008) 應用光達資料於溪頭柳杉人工林分調查之研究。國立台灣大學生物資源暨農學院森林環境暨資源學系碩士論文,108頁。
施瑩瑄 (2005) 應用空載光達獲取林分高之研究。國立台灣大學森林環境暨資源學系碩士論文,77頁。
彭炳勳、謝依達、陳朝圳 (2008) 空載光達雷射穿透率指數與柳杉林葉面積指數之關係探討。台灣林業科學 23: 63-73。
蕭淳伊、曾義星 (2009) 應用遙測影像與空載光達資料推估森林分布面積及樹冠體積。航測及遙測學刊 14(1) : 51-64。

被引用紀錄


徐世杰(2015)。可轉移之多面額電子現金〔碩士論文,中原大學〕。華藝線上圖書館。https://doi.org/10.6840%2fcycu201500909
Liu, Y. C. (2014). 離線電子現金系統的安全分析之研究 [master's thesis, National Taichung University of Science and Technology]. Airiti Library. https://doi.org/10.6826%2fNUTC.2014.00043
林哲欣(2014)。整合航測影像與光達資料監測南仁山地區森林孔隙動態變化〔碩士論文,國立屏東科技大學〕。華藝線上圖書館。https://doi.org/10.6346%2fNPUST.2014.00235
魏浚紘(2014)。應用光達技術於人工林之經營與監測〔博士論文,國立屏東科技大學〕。華藝線上圖書館。https://doi.org/10.6346%2fNPUST.2014.00169
林威廷(2014)。基於信用卡服務之多銀行付費系統〔碩士論文,朝陽科技大學〕。華藝線上圖書館。https://www.airitilibrary.com/Article/Detail?DocID=U0078-0905201416542777

延伸閱讀