As the largest carbon pool in terrestrial ecosystems, the dynamic changes in soil organic carbon (SOC) storage directly affect the global carbon cycle and climate system. Mechanistic process models play a crucial role in understanding the response and feedback of soil carbon storage to climate change. In recent years, scientists have recognized the critical role of microbial activity in the decomposition and stabilization of soil organic carbon (Allison et al., 2010; Cotrufo et al., 2013). Therefore, it is imperative to incorporate microbial factors into soil carbon cycle models.
Existing soil carbon cycle models that account for microbial activity have distinct characteristics, each with its own strengths and limitations. However, identifying a model structure that achieves both comprehensive process representation and manageable complexity remains an unresolved challenge. On the one hand, overly simplified models may fail to capture the complex interactions between microbial activity and soil carbon dynamics accurately (Wieder et al., 2015; Abramoff et al., 2022). On the other hand, overly complex models, while capable of providing a more detailed description of these processes, introduce significant challenges in parameter identification (Wang et al., 2015; Huang et al., 2018).
This study improves the microbial-explicit model Millennial V2 (Abramoff et al., 2022), which is based on measurable carbon pools, to develop five different model structures. A global dataset of measured soil carbon components is used to optimize model parameters using the non-dominated sorting genetic algorithm (NSGA). The differences in simulation results among the five model structures are then compared, leading to recommendations on the appropriate model structure for simulating soil organic carbon under different application scenarios. Then model parameters are extrapolated globally using geographic similarity, enabling global-scale simulation of soil organic carbon cycling. The simulation results are then compared with existing datasets and model outputs. This work not only enhances our understanding of the internal mechanisms of soil ecosystems but also provides a powerful tool for accurately predicting changes in soil carbon storage under global change.
 

Organic carbon,Soil carbon cycling,Model simulation,Model structure,Microbial mechanism