Detection of spatiotemporal dynamics of land cover and its drivers using remote sensing and landscape metrics (Pingtan Island, China)

Abstract

Land cover change (LCC) and its drivers are important inputs for tracking and tackling environmental problems. Pingtan has been experiencing rapid LCC recently, but it is not well documented. The general objective of this study is to detect land cover dynamics using three Landsat images from the years 2000, 2009 and 2013 based on metrics of land cover transition and pattern, together with driving factors. The overall trends in LCC depicted the expansion of forest land (6.75%), shrubland (4.81%), urban land (12.05%) and rural settlement (15.54%), as well as the reduction of grassland (− 17.86%), water body (− 0.71%), wetland (− 3.23%), farmland (− 17.16%) and bare land (− 0.19%). The land use intensity was also increased. Transformation of each land cover type was characterized by unbalanced expansion and contraction simultaneously and a spatial shift in their locations. During 2000–2013, a class-to-class transformation was dominated by the expansion of urban land and rural settlement (70.26% of gross gain), as well as the reduction of farmland and grassland (59.69% of gross loss). Landscape-level metrics indicated the decline of landscape heterogeneity and shape irregularities. However, there are some inconsistencies at the class level. LCC was driven primarily by policy factors and socio-economic development. Detailed quantification of spatiotemporal LCC will have significant contributions for sustainable land resource conservation and to improve land use planning policies.