The Land Carbon Sink in China over 2015–22 Constrained by OCO-2 XCO₂ Retrievals with a Regional Atmospheric Inversion System GONGGA-CN

China’s terrestrial ecosystems significantly offset anthropogenic CO₂ emissions, yet large uncertainties persist in quantifying this carbon sink using atmospheric inversion. In this study, a high-resolution regional atmospheric inversion system—Global ObservatioN-based system for monitoring Greenhouse GAses-ChiNa (GONGGA-CN)—is developed, which employs an advanced data assimilation method, NLS-4DVar, and a novel dual-pass optimization framework. By assimilating Orbiting Carbon Observatory-2 (OCO-2) column-averaged CO₂ observations, the system effectively constrains net biome productivity (NBP) across China from 2015 to 2022. The GONGGA-CN system estimates China’s mean annual NBP at 0.36 ± 0.04 (mean ± std) PgC yr⁻¹, consistent with values from OCO-2 v10 MIP (0.62 ± 0.49 PgC yr⁻¹, mean ± std) and the Global Carbon Budget (GCB, 0.31 ± 0.31 PgC yr⁻¹, mean ± std). It reveals strong carbon sinks in northeastern and southeastern China, dominated by temperate (0.192 ± 0.02 PgC yr⁻¹) and subtropical (0.125 ± 0.28 PgC yr⁻¹) regions, while the Qinghai–Tibet Plateau acts as a weaker sink (44 ± 19 TgC yr⁻¹). The system further quantifies climate-driven sink reductions: during extreme droughts (2019, 2021) and the 2015–2016 El Niño event, China’s carbon sink decreased by 8.3%, 16.7%, and 5.6%, respectively, linked to combined heat and drought stress. By establishing a satellite-constrained benchmark, GONGGA-CN enhances the accuracy of carbon sink monitoring and provides a critical tool for tracking China’s progress toward carbon neutrality.