From observation to state variable: embedding microwave-constrained biomass into Big Earth Data system dynamics

Rapid advances in Earth observation and geospatial infrastructures have enabled the emergence of the Big Earth Data (BED) paradigm for analyzing complex Earth system interactions. A key challenge is the physically consistent representation of vegetation dynamics, as biomass uncertainties propagate through carbon-cycle, hydrological, and climate–biosphere interaction. This study introduces a BED modeling framework that transforms microwave-derived biomass from an observational product into a dynamical state variable within the Climate–Nature–Society System (CNSS), implemented in the Geoecological Information-Modeling System (GIMS). The framework incorporates microwave remote sensing observables into CNSS dynamical equations within a BED architecture that processes multidimensional satellite data cubes. Microwave attenuation is physically inverted to retrieve vegetation biomass, which is directly embedded into coupled ecosystem dynamics. This formulation enhances sensitivity to vegetation structure and reduces saturation effects associated with optical indices. Analytical and numerical analyses demonstrate that microwave-constrained biomass enhances the stability of coupled system dynamics. Numerical simulations indicate reductions of up to 45% in vegetation carbon variance and 38% in gross primary productivity uncertainty. The proposed CNSS–GIMS framework supports the development of Digital Earth and Digital Twin Earth systems and improves the reliability of large-scale environmental monitoring aligned with sustainability objectives and the United Nations Sustainable Development Goals.