Spatial homogeneity from temporal stability: Exploiting the combined hyper-frequent revisit of Terra and Aqua to guide Earth System Science

Abstract

The terrestrial component of the Earth system has witnessed considerable changes in the past decades due to anthropogenic action. Throughout this period, the NASA Terra mission has been constantly monitoring the surface with the Moderate Resolution Imaging Spectroradiometer (MODIS) instrument. When combined with the MODIS instrument on-board of the Aqua platform, we obtain a hyper-frequent revisit capability providing sub-daily observations globally at a moderate resolution of ~250 m, but with a strong multi-angular variation of the observation footprint. Here we propose to exploit this particular configuration provided by the combined Terra + Aqua constellation to infer spatial homogeneity from the temporal stability of the observation time series. We thus designed a Temporal Coherence Index (TCI) based on the concept of information entropy. We test it on both synthetic and real landscapes showing its capacity to isolate which time series correspond to homogeneous targets based only on the MODIS measurements themselves. The index is calculated annually for the entirety of emerged surfaces for the period 2003–2020 by leveraging on the scaling capacity of the cloud computing Google Earth Engine (GEE). We foresee multiple practical applications for this index as a novel information layer that should foster the integration of surface data with satellite records, a practice that is increasingly necessary in Earth System Science.