Technology Overview

We utilize advanced artificial intelligence technology to pioneer a Synthetic Multi-approach multi-Layered carbon Emission flux inversion modelling System (SMILES). First, we construct a nested domain carbon flux inversion model, which includes a coarser-resolution global model and a finer-resolution regional model; the former provides initial and boundary conditions for the latter. Second, we apply four-dimensional variational methods (4D-Var) and Ensemble Kalman Filter (EnKF) inversion schemes, combined with AI-based data fusion methods, reducing the uncertainty of flux inversion from the current 50%-100% to below 20%. This model is capable of near real-time carbon flux inversion calculations, used for detecting hotspots, monitoring emission changes, and providing high-resolution, near real-time solutions for emissions of greenhouse gases such as carbon dioxide and methane.