Spatio-temporal variation indicators for landscape structure dynamics monitoring using dense normalized difference vegetation index time series

Abstract Terrestrial ecosystems are constantly changing at various spatial and temporal scales due to natural and/or anthropogenic factors. It is therefore of great interest to develop effective indicators based on spatio-temporal variations to detect, map, and monitor ecosystem changes on the continental and global scales. In this study, spatio-temporal variation indicators were established to characterize landscape structure changes based on temporal analysis combined with spatial information obtained using dense Landsat time series in Changting County, Fujian, Province, China. All available Landsat images from 1987 to 2017 were collected. The fractal dimension (FD) of the spatial neighbor normalized difference vegetation index (NDVI) derived from the Landsat images was calculated. Four change patterns (stable, increasing, decreasing, and wave) were explored using the ordinary least-squares method between abrupt change points performed by a Mann–Kendall test on the FD time series to characterize the spatio-temporal dynamics. The results showed that 49.98% of the landscape structure in the study area decreased over the 31-year study period. The forest and water were dominated by decreasing patterns (67.00% and 56.27%); the urban land and cropland were dominated by wave patterns (50.07% and 50.95%); bare land regions were dominated by wave patterns and increasing patterns (50.52% and 31.38%). Wide and discontinuous landscape structures with high heterogeneity were replaced by landscape patches. The regional landscape structure heterogeneity increased in the bare land regions and other homogeneous landscape structures, while the decreases may be coupled with internal landscape structure changes. It was concluded that the spatial and temporal variations of the change patterns and change processes determined based on the FD of the NDVI in the spatial neighbor region are systematic indicators and can be used to detect landscape structure changes successfully and provide insights into dynamic trends.