CONTEXT: Agricultural mechanization exacerbates global soil compaction, while subsoiling serves as an effective method to alleviate soil compaction. However, current research is still confined to small-scale experiments and lacks available process-based agroecological models incorporating subsoiling for quantitative simulation studies. This gap hinders scientifically evaluating the long-term impact of subsoiling on crop yield and soil organic carbon (SOC). OBJECTIVE: This work aimed to integrate a novel subsoiling module into the existing DNDC model framework to optimize subsoiling treatments and predict their long-term impacts on yield and SOC dynamics. METHODS: We developed an integrated model called DNDC-SubS 1.0 that combines the DNDC model with subsoiling processes. Specifically, the model characterizes depth-specific subsoiling impacts on soil properties, integrating these parameters with SOC decomposition, soil water-oxygen content, and soil microbial activity in the DNDC model to simulate yield and SOC dynamics. RESULTS & CONCLUSIONS: The model effectively simulates spring maize yield and SOC dynamics across five treatments in Qinggang County, Heilongjiang Province, from 2019 to 2025. The results showed that crop yields under different treatments followed the order: T1 < FP < T3 < T4 < T2, demonstrating general agreement with field measurements. Further analysis identified that a subsoiling depth of 34 cm combined with straw incorporation was the most effective treatment, designated as T-best. In future projections from 2026 to 2040, the T-best treatment achieved the highest yield stability along with a moderate yield increase. Our modified model demonstrates value for simulating yield-SOC responses to subsoiling under both historical and future climate conditions. SIGNIFICANCE: This study demonstrated that the DNDC-SubS 1.0 model provides a reliable tool to assess the long-term impacts of subsoiling on agricultural productivity and soil health.