Improving predictions of soil heterotrophic respiration: A parameter-optimized microbial model for alpine meadows in the three-river source region, Qinghai-Tibet plateau

As a critical pathway in soil organic carbon release, soil heterotrophic respiration (R h ) has simulation accuracy that impacts regional and global carbon cycle assessments. As a climate change hotspot, the Three Rivers Source Region (TRSR) exhibits considerable uncertainties in R h modeling, constraining in-depth research on regional carbon cycles. In this study, the MIcrobial-MIneral Carbon Stabilization (MIMICS) microbial model with an added moisture effect function is used for modeling R h of alpine meadow in the TRSR. To initialize the model and optimize its parameters, we utilized four years of field-measured data from three sites (Maqin, Tongde, and Chenduo), including R h , soil temperature, soil moisture, soil organic carbon, and microbial biomass carbon. Our results determine a set of generalized carbon pool parameters suitable for alpine meadow ecosystems across the study region. After selecting and optimizing high-sensitivity parameters, the model's simulation performance for soil R h improved significantly, with R 2 remaining above 0.7 and the relative root mean square error (RRMSE) controlled within 11%–14%. Furthermore, relative to the original MIMICS model, integrating the moisture effect function increased the model's R 2 by 0.19–0.30. This study represents the attempt to refine the MIMICS model for R h regulation, which not only provides a more accurate tool for simulating the carbon cycle of alpine meadow in the TRSR but also offers new insights into the coupling mechanisms between microbial models and environmental drivers. • The modified microbial model was first used to model soil R h in alpine meadows. • A set of universal parameters applicable to alpine meadows was identified. • Parameter optimization improved model performance with R 2 exceeding 0.7. • 1 km grid smoothing reduces R h spatial heterogeneity by 12.77%.