Assessing spatio-temporal changes in forest cover and fragmentation under urban expansion in Nanjing, eastern China, from long-term Landsat observations (1987–2017)

Abstract

Assessing changes in forest cover and fragmentation under urban expansion and the correlation analysis between these changes can provide decision support for urban forest managers. In this study, forest cover and morphological spatial patterns (cores, islets, perforations, edges, loops, bridges, and branches) were firstly mapped by the vegetation change tracker (VCT) algorithm and morphological spatial pattern analysis (MSPA) respectively in Nanjing from 1987 to 2017. Next, the visible red and near-infrared-based built-up index (VrNIR-BI) was derived to extract urban impervious surface (UIS) areas, and the neighbourhood-based urban expansion model was proposed to describe the expansion types (edge-expansion, infilling and outlying). Finally, the relationships among the forest cover, fragmentation changes and urban expansion were determined using the Pearson correlation coefficient (r). The results showed that the distribution of forest cover in Nanjing was relatively scattered and it decreased by 94 km2, while the UIS area increased by 893 km2 from 1987 to 2017, resulting in some forests being gradually surrounded or encroached upon by UIS areas. The main forest morphological spatial pattern was identified as cores, with many small-area cores. Moreover, islets exhibited a higher proportion of 0.18, and these suggested that the fragmentation of Nanjing's forests was severe. Forest net changes had the higher correlation with cores, edges and branches. In addition, there were strong correlations between edge-expansion and cores, edges, branches with the r values greater than 0.75, which meant that most of the urban areas extended to core regions from the forest edges or branches. The derived information will help forest managers monitor forest dynamics in response to urban expansion and achieve sustainable development.