Revising the process-based Breathing Earth System Simulator (BESS) for rice growth and evapotranspiration simulation: BESS-Rice model description and evaluation

As a satellite remote sensing derived, biophysical process based model, Breathing Earth System Simulator (BESS) performs well in estimating global gross primary productivity (GPP) and evapotranspiration (ET). However, lacking of the process of carbon allocation, which plays a critical role in ecosystem carbon cycling by shifting the products of photosynthesis, BESS could not be used for crop growth simulation and yield estimation. Meanwhile, BESS did not sufficiently take into account the water standing evaporation of paddy rice (Oryza sativa L.). In this study, we developed the BESS-Rice model by integrating carbon allocation module into BESS, and revising the evaporation module for the early growth stage with open water. Accumulated GPP was used as a scaler for growth development, and the dynamics of dry matter distribution among aboveground organs (grain, leaf and stem) during development was simulated based on the relationships between accumulated GPP and partitioning coefficients. Annual grain yield and straw production was estimated finally. Analysis of correlation shows that over 90% variations in dry matter partitioning coefficients could be explained by accumulated GPP. It is much simpler than conventional crop models and enables us to deduce the uncertainty related to the input variables while detecting key processes. Meanwhile, a Priestley-Taylor approach was used to quantify the open water evaporation during the early stages of rice growth. We evaluated the BESS-Rice simulated GPP, ET, grain yield and straw production against eddy covariance and biometric measurements at site scale. The diurnal and seasonal patterns of carbon and water fluxes and the rice growth dynamic were also investigated. The process-based BESS-Rice could be calibration-free and applied globally, which is significant for large-scale rice growth monitoring and ecosystem carbon cycling research.