•A Parallel distributed HPC framework (RS-DPCF) was developed for the RZ-SHAW model.•RS-DPCF requires less time and effort in calibrating parameters for multiple sites.•RS-DPCF provides faster calibration for the RZ-SHAW model with scalability.•Multi-threading delivers better efficiency compared to multi-processing for RZ-SHAW. The winter/spring season in cold climate regions has been recognized as a critical period for cropland nutrient loss and greenhouse gas emissions and has been predicted to be vulnerable to climate change. Agricultural system models serve as essential tools in anticipating potential outcomes. The RZ-SHAW model, combining the Simultaneous Heat and Water Model (SHAW) with the Root Zone Water Quality Model 2 (RZWQM2), comprehensively describes snow and soil freezing dynamics. Showing promising results in simulating winter frost dynamics it, however, remains computationally intensive. High-performance computing has been recognized as a promising approach to reducing simulation runtimes. A modularized parallel distributed high-performance computing (HPC) framework (RS-DPCF) with both multi-threading (MT) and multi-processing (MP) capability was developed for RZ–SHAW calibration and simulation. Two case studies demonstrated that the developed RS-DPCF program could significantly reduce RZ–SHAW calibration time, allow improved computation resource scalability, and perform field-scale simulations of overwintering conditions for multiple croplands across Canada. For the RZ–SHAW model, MT provided a parallel-computing efficiency superior to that of MP.