Substantial uncertainties in global soil organic carbon simulated by multiple terrestrial carbon cycle models

Soil organic carbon (SOC) is critical to the terrestrial ecosystem carbon cycle and global climate change. The carbon cycle model coupled with microbial processes can improve the projection of SOC. However, it remains unclear whether microbial models are superior to multiple terrestrial carbon cycle models and how large the simulation uncertainties of SOC are. Therefore, we simulate the spatial patterns of global SOC by the MIMICS (explicit nonlinear microbial carbon cycle model) and the DCC (implicit linear carbon cycle model), compare the SOC with that of the CMIP6 MME (multi-model ensemble) and the observation to obtain the uncertainties of SOC, and analyze the sensitivity of the parameters in the two models. The results show that the SOC simulated by the MIMICS is 1615.4 +/- 54.3 PgC, which is higher than that of the DCC (668.5 +/- 102.1 PgC), the CMIP6 MME (1443.7 +/- 795.8 PgC), and the observation (1519.1 PgC). The SOC of the DCC and MIMICS are regulated by NPP, and that of DCC is more sensitive to climate, while that of MIMICS is mainly influenced by the Michaelis-Menten equation and microbial carbon use efficiency. We reveal the spatial patterns of uncertainty in global SOC simulated by the explicit and implicit microbial processes of multiple terrestrial carbon cycle models. Unfortunately, neither the DCC, the MIMICS nor the CMIP6 MME simulate SOC satisfactorily, especially in the high latitudes of the Northern Hemisphere, and the simulation of SOC in this region needs to be improved.