Abstract:Forms of human habitation within rural settlements are a major component of rural landscapes. The spatial patterns and evolutionary characteristics of rural settlements are influenced by multiple factors, including natural, social, and economic elements. It is, therefore, important not only to properly identify and assess the spatial distribution patterns and evolutionary modes of rural settlements, but also to uncover causative laws and influential factors behind their evolution. This can facilitate scientific rural planning and enhance optimal distribution of urban and rural land resources. Our study is based on a series of digitalized vector data, including the locations of rural settlements, land slopes, rivers, and road networks obtained from remote sensing (RS) images of the city of Dujiangyan in Sichuan Province for 2005 and 2010. We applied analytical techniques derived from Geographic Information Systems (GIS) and RS to quantitatively analyze the processes, patterns, and trends of spatial evolution in rural settlements. Specifically, we explored landscape pattern indexes to examine in depth the factors influencing the distribution characteristics of rural settlements in the study area. Our results showed that: 1) In general, while there was significant clustering in the spatial distribution of rural settlements in Dujiangyan during both 2005 and 2010, this settlement pattern was even more significant in 2010. Along with an increase in the spatial scale, the degree of the clustering tendency also increased, reaching a maximum value at a scale of 13 km. 2) The presence of land slopes, roads, and rivers significantly influenced the distribution of rural settlements in Dujiangyan as evidenced by the following findings. More than 80% of settlements were located in areas with slopes of less than 10 degrees; over 50% of settlements were located within 500 m from a road network; and almost 60% were distributed within 1000 m of a river. 3) In 2005 as well as 2010, there was no obvious change in the concentrated locations of clusters, which were consistently found in the southeastern region of Dujiangyan City. However, during both time periods, fewer settlements were found in the townships of Hongkou, Longchi, and Qingcheng Mountain. The reason for this imbalance in the spatial distribution of rural settlements was that besides the influence by natural factors such as the landscape, the distribution patterns of rural settlements in Dujiangyan were also closely associated with several planning elements. These included a policy of protecting national regions of scenic interest and a World Cultural Heritage site located in the city, the rural land comprehensive consolidation projects and post-earthquake reconstruction planning. 4) Based on the above findings on factors that influence the distribution of rural settlements, our research suggests that comprehensive consideration of topographical and geological conditions is required for optimal adjustment of rural settlement patterns. Scientific planning procedures can promote the relocation of rural residents from villages to townships and urban areas, and help to lessen disruption in environmentally protected regions. They can also reinforce the effective use of land resources and increase land supporting capabilities. We expect our research to provide a methodological reference for similar research conducted in the future. It can also provide a theoretical and practical reference for rural land-related decision-making and technical support, for example, dynamic monitoring of changes in rural settlements, assessments of rural land consolidation, and the "New Socialist Countryside" rural planning policy.