運用馬可夫鏈模型與細胞自動機理論模擬植生復育之研究
No Thumbnail Available
Date
2006-11-??
Authors
張政亮
張瑞津
Journal Title
Journal ISSN
Volume Title
Publisher
地理學系
Department of Geography, NTNU
Department of Geography, NTNU
Abstract
植被是全球陸域最具規模的地表景觀，而植被覆蓋率的空間型態變化常受人類社經活動與自
然環境等因素的衝擊而有所影響，因此探其變遷與演育歷程，對土地規劃、災害防治與生態保育，
都有一定程度之助益。隨著科學知能的進展，許多新的技術與方法也逐漸運用於植被變遷與空間
分析的研究領域，因此本文選擇921 地震後地貌破壞十分嚴重的南投九九峰地區進行研究，文中
先運用1999 年3 月（921 地震前）、1999 年10 月（921 地震後）、2002 年11 月、2005 年11 月
等四期的SPOT 衛星影像，分析常態化差異植生指數(Normalized Difference Vegetation Index，NDVI)
之變異，作為植物覆蓋率、崩塌裸地及植生復育等變遷之判釋；接著以地理資訊系統（GIS）軟
體為分析工具，來探究影響研究區地表植被復育的機制與原因；最後結合馬可夫模式與細胞自動
機之理論，進行模擬及分析，來瞭解九九峰地區植被變遷的空間分布與意涵。歸結研究成果顯示：
1.藉由遙測與GIS 之分析得知，1999 年921 震災前NDVI 影像呈現顯明之淡白色澤，顯示本區之
綠色植被繁茂；921 後大量的地形崩塌，裸露地取代植被，九九峰地區內NDVI 反逆成一片闇黑，
80％地表均呈裸露狀態，說明地景破壞之劇烈，惟經過一定時間之生養，研究區內的NDVI 由幽
黑轉灰白，代表植物漸趨生長覆蓋，以2002 年為例，草地植被覆蓋達60％的比例，到了2005 年
草地植被進一步生長成茂密樹林，六成以上之地表林相已恢復往昔常態；2.植被生長與地表景觀
復原程度深受高度、坡度與坡向等地形因子的影響與控制；3.相較而言，結合基於細胞自動機理
論的馬可夫預測模式比隨機決定的馬可夫預測模式能更合理與有效地模擬本區植生復育與地景變
遷的過程。
The recovery of vegetation is the most significant landscape in the terrestrial globe. In recent years it has been changed rapidly due to human socio-economic growth and environmental impact. Therefore, many research methods have been applied so as to comprehend and forecast these dynamic changes. This paper intends to apply some of theses methods and processes to quantify the change of vegetation recovery in Mount Jiou Jiou area. First of all, the graphic data of study area is collected by using the SPOT satellite images of four representative periods (Mar.1999, Oct.1999, Nov.2002 and Nov.2005). Subsequently the vegetation types and their changes are examined through the Normalized Difference Vegetation Index (NDVI) analysis. Second, the Geographic Information System (GIS) technology is adopted to recognize the distributions of different vegetation types, characteristics of topography and their correlation. Finally, by combining the methods of Markov chains and Cellular automata this study models and simulates the spatio-temporal change of vegetation recovery. The results can be summarized as follow: (1) After 921 earthquake, 80% of vegetation on Mount Jiou Jiou area was destroyed seriously in 1999, therefore the index of NDVI was very low. After that, the vegetation is recovered gradually for recent years under stable weather conditions. (2) The healing situation of vegetation is intensely affected by geomorphologic factors, such as height, slope and aspect. (3).After comparing two models testing, the present study shows that CA-Markov model is more suitable for simulating the change trend of vegetation recovery.
The recovery of vegetation is the most significant landscape in the terrestrial globe. In recent years it has been changed rapidly due to human socio-economic growth and environmental impact. Therefore, many research methods have been applied so as to comprehend and forecast these dynamic changes. This paper intends to apply some of theses methods and processes to quantify the change of vegetation recovery in Mount Jiou Jiou area. First of all, the graphic data of study area is collected by using the SPOT satellite images of four representative periods (Mar.1999, Oct.1999, Nov.2002 and Nov.2005). Subsequently the vegetation types and their changes are examined through the Normalized Difference Vegetation Index (NDVI) analysis. Second, the Geographic Information System (GIS) technology is adopted to recognize the distributions of different vegetation types, characteristics of topography and their correlation. Finally, by combining the methods of Markov chains and Cellular automata this study models and simulates the spatio-temporal change of vegetation recovery. The results can be summarized as follow: (1) After 921 earthquake, 80% of vegetation on Mount Jiou Jiou area was destroyed seriously in 1999, therefore the index of NDVI was very low. After that, the vegetation is recovered gradually for recent years under stable weather conditions. (2) The healing situation of vegetation is intensely affected by geomorphologic factors, such as height, slope and aspect. (3).After comparing two models testing, the present study shows that CA-Markov model is more suitable for simulating the change trend of vegetation recovery.