A neural-network-based approach was suggested to estimate CO₂ emissions using satellite-based net primary productivity (NPP) and XCO₂ retrievals. XCO₂ anomalies were calculated for each year using OCO-2 retrievals. A Generalized Regression Neural Network (GRNN) model was then built; NPP, XCO₂ anomalies, and ODIAC CO₂ emissions from 2015 to 2018 were used as a training dataset; and, finally, CO₂ emissions were predicted for 2019 based on the NPP and XCO₂ anomalies calculated for the same year.