Advance Gross Primary Product (GPP) Estimation by Integrating Remote Sensing-Based Evapotranspiration into Light Use Efficiency Modeling

Gross primary productivity (GPP) is a key component of the terrestrial carbon cycle, providing essential insight into carbon–water flux dynamics and serving as an indicator of ecosystem responses to climate change. Yet, GPP products at daily temporal scale and fine spatial resolution (1 km) remain limited, constraining their ability to detect short‑term climate extreme impacts on vegetation productivity. To address this gap, an enhanced GPP estimation framework was applied that integrates remote‑sensing‑based actual evapotranspiration (ET_a) as a water-stress scalar into a MODIS‑reanalysis‑driven light use efficiency (LUE) model, producing daily GPP estimates at 1 km resolution. Model performance was evaluated using flux tower observations from the FR‑Aur cropland site (ICOS Warm Winter 2020 dataset), representing the northern study region climate condition, and compared against MOD17, the most widely used operational GPP product. A total of 387 clear‑sky flux‑GPP records across seven wheat growing seasons, identified using smoothed NDVI‑based phenology, were used to evaluate the LUE model framework. The modelled GPP substantially outperformed MOD17, yielding a mean bias of approximately 0.77 g C m⁻² d⁻¹ relative to flux observations, compared with a larger bias of −3.02 g C m⁻² d⁻¹ for MOD17. These results demonstrate that incorporating remote‑sensing‑based ET_a into LUE frameworks significantly improves daily GPP estimation, offering a more accurate basis for monitoring climate–ecosystem interactions and assessing vegetation productivity under increasing climate variability.