Spatiotemporal analysis of prehistoric settlement clustering based on terrain constraints in the Linfen area of China

Ancient people's choices of where to live were strongly influenced by geography, especially by terrain. However, most settlement clustering algorithms, which use the Euclidean distance, do not consider terrain. Here, we present an improved clustering algorithm for the cluster analysis of ancient ruins that takes terrain‐related factors, such as slope and elevation, into account and uses the resistance distance (also known as the cost‐weighted distance) instead of the Euclidean distance. Furthermore, we select the prehistoric settlements (during 7‐3 ka BP within four cultural stages) in the Linfen area in the Fen River Basin as research objects to analyze the settlement distribution and spatiotemporal evolution of the four cultural staging sites with reference to the climate. We conclude that the proposed site clustering method based on terrain constraints can simulate ancient people's travel routes and improve the clustering results. The spatiotemporal analysis of prehistoric settlements in the Linfen area reveals that most of the settlements are distributed in the alluvial region of the river. Over time, the dependence on the rivers gradually decreased. The clustering and spatiotemporal analysis results show that the proposed method is feasible and applicable to ancient settlements.