Exploring light use efficiency models capacities in characterizing environmental impacts on paddy rice productivity

Abstract

Remote sensing-driven light use efficiency (LUE) models have been widely used to calculate gross primary productivity (GPP) for various terrestrial ecosystems, but there was limited knowledge on the capacity of LUE models to evaluate the GPP in paddy rice ecosystems. In this study, at seven rice-growing sites over the Northern Hemisphere and based on six commonly used LUE models, we calibrated the parameters (i.e., maximum LUE (LUEmax) and optimum temperature) by separating the growing period into four phenological transitions and evaluated the performance of models, and investigated the impact of changes in cloud conditions and environmental factors (i.e., air temperature and vapor pressure deficit) on GPP simulations. The calibrated LUEmax corresponded closely to phenology, allowing the six LUE models to track the seasonal variations in GPP reasonably well. The sensitivity of GPP estimates to sky clearness index (CI) indicated that the TL-LUE model incorporating diffuse radiation fractions outperformed other models under cloudy conditions. Environmental stressors including along with changes in the diffuse radiation fractions synergistically affected GPP simulation, resulting in distinctly variable performances of the LUE models under different water and temperature conditions, with the TL-LUE model always performing well during the suitable rice growing season. These results demonstrate that it is crucial to consider the diffuse radiation fraction and to better represent environmental stresses under certain environmental conditions in LUE models for accurate estimation of rice GPP.

Keywords

• Light use efficiency model
• Gross primary productivity
• Paddy rice
• Environmental stressors