Exploring Spatiotemporal Phenological Patterns and Trajectories Using Self-Organizing Maps

摘要: Consistent satellite image time series are increasingly accessible to geoscientists, allowing an effective monitoring of environmental phenomena. Specifically, the use vegetation index has pushed forward large-scale phenology. Most these studies derive key phenological metrics from Normalized Difference Vegetation Index (NDVI) on a per-pixel basis. This paper demonstrates approach analyze synoptic spatiotemporal patterns over large areas, rather than per pixel. The selected involves data mining using self-organizing map (SOM) and Sammon's projection. To illustrate our approach, we trained SOM 13 years ten-day NDVI composites Systeme Pour l'Observation de la Terre-VEGETATION Kruger National Park, South Africa. resulted in topologically ordered set states. projection was then used create simplified representation that reflects similarities among Subsequently, depicted trajectories for each season show how development changes between years. provides holistic characterization main regional dynamics effectively summarizes information present series, thus facilitating further interpretation.