Analysis of remote sensing time-series data to foster ecosystem sustainability: use of temporal information entropy

Remotely sensed time-series data have provided valuable information and sound foundations for ecological sustainability studies. Ecosystem sustainability has been viewed as a dynamic process that requires an ecosystem to deal with climate change and anthropogenic disturbances. Following this school of thought, ecosystem sustainability can be portrayed in terms of order and disorder using spatio-temporal analysis of entropy-related indices of Normalized Difference Vegetation Index (NDVI) time-series. Information theory and entropy-related measures have provided insights for complex systems analysis and have high relevance in ecology; however, less attention has been focused on temporal evolution and dynamics. The overall aim of this study is to propose an index called ‘temporal information entropy’ (H _t), and it is an entropy-related index able to describe the degree of order and regularity within a time-series of observations. We then assess H _t’s ability to measure the ecosystem sustainability of Yanhe watershed based on MODIS (Moderate Resolution Imaging Spectroradiometer) NDVI time-series. Our results indicate that temporal information entropy of ecological time-series data may be used as a natural indicator with respect to sustainability, and in some degree, it helps us to get a better understanding of ecosystem dynamics from a physical-based standpoint.