MCI GPP: ensembling a global model- and climate-independent gross primary productivity for 2001–2023

Gross primary productivity (GPP), the starting point of carbon entry into the land, is crucial for understanding the global carbon cycle. Previous research have debated incorporating the CO₂ fertilization effect (CFE) and canopy structural traits into GPP modeling. This study systematically evaluates their influence, demonstrating that CFE improves GPP estimation accuracy and significantly alters long-term trends. Interestingly, a two-leaf model (TLM) achieved comparable accuracy to the production efficiency model (PEM). Leveraging these insights, we generated 12 distinct GPP datasets and integrated them into a novel model- and climate-independent (MCI) GPP product using random forest regression and spatio-temporal tensor models. The MCI GPP estimates average global GPP from 2001 to 2023 at 141.9 ± 4.0 Pg C yr⁻¹, with a significant global increase of 5.7 Pg C yr⁻¹ per decade. Validation against AmeriFlux data shows MCI GPP outperforms other global products (MOD17, GOSIF, X-Base Fluxcom), achieving an R² of 0.72 and RMSE of 1.86 g C m⁻² d⁻¹. Available on Zenodo, this robust 0.05° monthly dataset provides a valuable resource for carbon-climate feedback studies.