海岸為海洋與陸地交界之處，風、浪與潮流等自然營力長期於此不斷的侵蝕與堆積交互演替。近年來，隨著人口快速增加，人類對於海岸地區土地利用與開發的需求急遽擴張，使得影響海岸地形的變因日益複雜且變化迅速。宜蘭特殊的沙丘性海岸因抗蝕性弱，易受到外力影響而改變地形，海岸後方的沿海平原為人口與產業集中的地區，因此自然營力與人為因素對宜蘭平原海岸地形與環境的影響，備受關注。因衛星影像具多時期與大尺度的空間特性，可提供土地覆蓋變遷分析之有效資訊，故本文使用2003年、2006年與2009年宜蘭平原沿海地區SPOT 5衛星影像，利用階層式分類程序將研究試區分為水體、建成與交通用地、沙地、農地與林地等五種土地覆蓋類別，透過土地覆蓋分類之結果，比較三個時期土地覆蓋型面積的變化；建立馬可夫轉移矩陣，了解各土地覆蓋型轉移的情況；其次，量化地景指標以了解整體土地覆蓋型區塊在空間結構上的情況，並利用Shannon多樣性指標t檢定測驗兩時期間整體地景是否有明顯的變遷。進一步，利用二項式Logit迴歸分析2003至2006年與2006年至2009年間土地覆蓋型的變化與沙丘海岸變遷的關係以及參考前人宜蘭海岸變遷之研究，選擇可能影響此區海岸變遷的自然與人為環境因子，建立二項式Logit迴歸模式，探討各項因子對於沙丘性海岸的影響，並利用海岸沙丘空間分佈預測機率圖，最後以2006年與2009年沙地主題圖作為驗證資料，探討模式的可行性。本研究透過不同的空間計量方法，了解本區土地覆蓋型的變化，期本研究成果對於此區海岸保護與管理政策制定者有一參考的依據。 / Coastal zone is at the junction of ocean and land. The area constantly experiences interchanging succession of erosion and accumulation due to natural forces such as wind, wave, and tidal currents. In recent years, associated with fast population increase, the demand of lands expanded rapidly such that the effects on topography of coastal zone became more complex and changed quickly. Coastal sand dunes are dynamic and fragile terrain often regarded as environmentally sensitive areas. Sand dunes are vulnerable to erosion by natural process and human activity. The objective of this research was to examine the effects of environmental factors and land-use changes on coastal sand dunes in I-Lan County. Satellite imageries are characterized by multi-temporal and large-scale, therefore they are ideal for providing necessary information to facilitate analysis of regional land-cover changes. In this research, three SPOT 5 images acquired in 2003, 2006 and 2009 were used to analyze land-cover changes in I-Lan coastal zone. Firstly, a hierarchical classification procedure was applied to classify the image data to five land-cover types and the land-cover changes were compared. Secondly, based on the classification results, a Markov transitional probability matrix was constructed to understand the transition among different land-cover types, and the Fragstats software was used to quantify the landscape structure of three different periods. By analyzing the spatial distributions of land-cover types in different time periods, we were able to examine to the temporal and spatial changes of land-cover in the I-Lan coastal zone. In addition, a t-test based on Shannon diversity index was used to evaluate the changes of the whole landscape in the study area. Thirdly, by selecting possible natural and man-made factors that are likely to affect coastal environment based on various prior studies, the mathematical models such as Markov chain and binomial logit regression analysis were applied to predict the future overall landscape structure and to simulate the spatial distribution of the sandy coastal zone. Thematic maps derived from satellite images obtained in 2006 and 2009 were used to verify and assess the feasibility of the models. This study integrated several spatial statistical methods to understand the patterns of land-cover changes in the study area. It is expected that the results of this study may offer a valuable reference for the policy-makers of coastal protection and management.