Representing cropping systems with the MEMS 2.0 ecosystem model.

Abstract

Abstract Croplands have been the focus of substantial investigation due to their considerable potential for sequestering carbon within the soil matrix. The adoption of process-based models stands as a pivotal approach for elucidating soil organic carbon (SOC) dynamics and guiding land management strategies. Accurately representing crop growth and agricultural practices in these models is critical for realistic SOC modeling. The MEMS 2.0 model incorporates current understanding of SOM formation and stabilization, measurable pools, and deep SOC dynamic is seen as a highly promising tool to inform management intervention for SOC sequestration. However, it was developed to represent grass vegetation. In this study, we further developed MEMS 2.0 to model annual grain crops and common agricultural practices, such as irrigation, mineral nitrogen fertilization, harvesting and tillage. Using four Ameriflux sites, we demonstrated accurate simulation of crop growth and development. Model performance was strong for simulating aboveground biomass (R 2 range of 0.66 to 0.96) and green leaf area index (GLAI) (0.77 to 0.91 R 2 ) across corn, soybean, and winter wheat. Good agreement with observations was also achieved for net ecosystem CO 2 exchange (NEE) (0.81 to 0.85 R 2 ), evapotranspiration (ET) (0.76 to 0.79 R 2 ), soil temperature (ST) (0.93 R 2 at 4 cm), while discrepancy with the available soil water content (SWC) data remain (0.02 to 0.49 R 2 at four depths to 100 cm). While we will continue model testing and improvement, MEMS 2.0 has now demonstrated its ability to effectively capture the dynamics of agricultural land.