土地利用與土地覆蓋變遷是人與環境交互作用的展現。隨著都市化的快速發展，人們對土地的使用需求增加，改變了原有的環境地貌，提高災害發生的風險。因此，土地利用變遷分析與預測模式為學術界所關注的重要議題之一。早期的土地利用變遷研究大多透過田野調查來獲取資料，需耗費較多人力與時間，限制了研究的深度與廣度。近年來，隨著遙測科技與地理資訊系統的發展，讓研究者能在較短時間內取得高解析度的遙測影像將資料進行分析、模擬地表環境變遷。本研究以2006年為起始點，利用福衛二號影像與馬可夫鏈模式建構基隆河臺北市段的土地覆蓋變遷分析模式，並透過熱點分析探討土地覆蓋的空間分布特性。本研究分析之地區為基隆河臺北市段，係因基隆河自第二次截彎取直計畫結束後，河道環境已少有重大變革，加上截彎取直後的新生地有利於人為經濟活動的介入，並改變既有的土地利用。期望透過本研究瞭解人類活動對基隆河二次截彎取直後的土地覆蓋變遷影響，以及在時空上的分布特性，作為日後土地利用變遷相關研究的參考。

Land use and land cover changes (LUCCC) demonstrate the interactions between human beings and their environment. For several decades, along with rapid urbanization, the increasing demands on limited lands not only change land surfaces, but also increase the risk of disasters' impacts. In addition, land use and land cover changes are a hot topic in the field of geography. In the early years, field work was the widely accepted method to acquire data; however, it was laborious and time-consuming, so research depth and extension are limited. With the advances in Remote Sensing (RS) and Geographic Information System (GIS), researchers nowadays are able to acquire high-resolution satellite imagery to build land change models to simulate environmental changes. This study applied these technologies, RS and GIS, to examine land use and land cover changes. We chose the Keelung River in Taipei as the study area since its channel has fewer changes after two straitening of the channel curves project and more human activities influenced on this area. Starting in 2006, we used FORMOSAT-2 imagery and Markov Chains to analyze land cover changes in the Keelung River. Additionally, hot-spot analysis was used to characterize land change patterns. We expect this study will allow people to understand the impacts how human activities influence on the Keelung River after the second straightening project, and how land covers change over space and time.