查詢結果分析
來源資料
頁籤選單縮合
題 名 | 地文因子對旗山溪集水區莫拉克風災崩塌地植生復育影響之研究=Effects of Topographic Factors on the Vegetation Recovery for the Landslides Caused by Typhoon Morakot in the Chishan Streams Watershed |
---|---|
作 者 | 林于筌; 江孟玲; 林昭遠; | 書刊名 | 水土保持學報 |
卷 期 | 44:3 2012.09[民101.09] |
頁 次 | 頁295-312 |
分類號 | 351.8 |
關鍵詞 | 崩塌地; 多變量分析; 植生復育; 地文因子; 旗山溪; Landslides; Multivariate analysis; Vegetation recovery rate; Topographic factors; Chi-shan stream watershed; |
語 文 | 中文(Chinese) |
中文摘要 | 莫拉克颱風於2009 年8 月6 日至10 日為台灣中南部帶來嚴重災害,創紀錄之超強雨勢造 成南台灣山區多處村落遭到崩塌而摧毀掩埋、人員傷亡與房屋毀損,並造成土石鬆動以及植生 破壞,地形地貌已全然不同。然風災至今已近三年,有必要對崩塌地進行植生復育評估,瞭解 崩塌區位植生復育變遷情況。本研究利用旗山溪集水區莫拉克風災前後之衛星影像,萃取崩塌 區位,並以風災後一年及二年之衛星影像,分別計算崩塌區位兩期之植生復育率。結果顯示風 災後一年及二年;旗山溪崩塌地植生復育率分別為6.16%與28.61%,植生復育率提升了 22.45%,表示崩塌區位植生有持續復原之趨勢。另以19 項集水區地文因子;利用主成分分析, 探討集水區地文因子與植生復育率之關係,可歸類為五個主成份軸,五個主成份軸約能解釋 90.82%之成份。再以Kmeans 群集將崩塌地植生復育率分為優、好、普通及差等四類,利用判 別分析具顯著影響之集水區地文因子,可得四組之Fisher’s 線性區別函數,其分類準確第一年 達60%、第二年可達90%;兩年間植生復育差值之分類精準可達77.5%,得知集水區地文因子 對崩塌區位之植生復育率有顯著之影響,其解釋能力隨演替時間越久有越好之趨勢。 |
英文摘要 | Typhoon Morakot dumped torrential rains that caused serious flooding in the central and southern Taiwan. This serious storm had washed out the villages in the mountain and caused large amounts of landslide. Three years passed, the vegetation recovery rate (VRR) of the landslides needs to be evaluated to know the changes of vegetation recovery status. SPOT satellite images of the Chishan streams watershed were applied to extract the spatial distribution of landslides and calculate the VRR of the landslides. Results show that the average of VRR is reached to 6.16% for the 1st year and 28.61% for the 2nd year. The increase of 22.45% means that the recovery status is better. Principle component analysis depicts that five principle components can be derived from 19 topographic factors and there is about 98.82% interpretation. Furthermore, four categories (excellent, fine, ordinary and worse) of VRR were classified using K-Means clustering method, which coupled with discriminant analysis can get four sets of Fisher’s Linear Discriminator to understand the effects of topographic factors on the VRR of the watershed landslides. The accuracy can reach up to 60% for the 1st year and 90% for the 2nd year, while the difference of the two phases is 77.5%. It shows that watershed topographic factors can efficiently interpret the VRR of the landslides. |
本系統中英文摘要資訊取自各篇刊載內容。