资源环境科技发展态势分析平台

Urbanization and agriculture can be considered as main drivers of landscape change in many parts of the world, including France, producing particular interacting environmental gradients, which in combination with natural components, give rise to the urban-rural-natural (URN) gradient. Along this gradient, many complex transitional landscapes, in which dynamic processes and unstable conditions occur, can be identified. This research focused on the development of a broadly applicable methodology to classify and analyze the entire URN gradient, beginning with widely available Land Use/Land Cover (LULC) data. Kernel Density Estimation was used to model LULC density gradients and Principal Component and ISODATA cluster analysis were combined to classify the landscape types composing the URN gradient. A particular adjacency graph analysis supported the visualization, a better understanding of the gradient organization, and the discovery of key landscape sequences in the area under investigation. A subsequent characterization of landscape patterns along these sequences, using a database management system and spatial metrics, produced a deeper comprehension of the structures and organization of the landscape along specific portions of the gradient. The discovery and characterization of all the landscape types nested along the URN gradient is crucial to facilitate more effective land use planning in which transitional landscapes assume a key role.