Spatially explicit mapping of phenological transition zones: A fuzzy-logic approach

Spatially explicit mapping of the Phenological Transition Zone (PTZ) is the prerequisite to PTZ dynamics monitoring and characterization, and is significant for more effective management of natural resources, biodiversity, and conservation areas in the context of climate change. However, to our knowledge there have been no studies to delineate spatially explicit PTZs and there are some potential issues associated with the mapping of other biogeographical transition zones. This paper developed a method to explicitly delineate PTZs continuously from 2003 to 2017 in the study area of the Forest-Grassland Mosaic Region (FGMR) in northeastern China, highlighting both their transitional and phenological characteristics. Annual Normalized Difference Vegetation Index (NDVI) time series were fed into the dynamic-time-warping (DTW) based fuzzy c-medoids (FCMdd) clustering to produce 12 fuzzy phenoregions. Potential PTZ pixels were identified for each pair of fuzzy phenoregions via the fuzzy intersection operator. The Delaunay tessellation field estimator was applied to convert the discrete potential PTZ pixels to continuous weighted density surface. The 7 consistently and 8 intermittently existing PTZs were exhaustively derived via thresholding. Three different types of gradient, inclusive and mosaic PTZs were identified. The PTZ dynamics were rapid and dramatic due to its phenological and transitional nature. This PTZ mapping method can be easily extended to other areas with no or minor adjustments, and it can also be adapted to map other biogeographical transition zones.