Abstract: Objective Carbon capture, utilisation and storage (CCUS) is a key pathway to achieving large-scale carbon emissions reductions. As the primary medium for the long-distance, large-scale transport of CO₂, high-pressure underground pipelines are subject to constraints imposed by the overlying soil layer, resulting in a more complex mechanism of CO₂ leakage and diffusion. Following a leak, the throttling effect causes a sharp drop in temperature, which may trigger secondary disasters such as low-temperature brittle fracture of the pipeline material and soil frost heave. The diffusion behaviour of CO₂ in soil is influenced by multiple factors, including moisture content and pore structure, and the diffusion process differs fundamentally from that in above-ground pipelines. Methods Utilising an industrial-scale underground CO₂ pipeline leakage test platform, 11 sets of leakage experiments were conducted, varying soil moisture content, leakage direction and the nominal diameter of the leakage orifice. The dynamic evolution of the soil temperature field during the leakage process was monitored in real time. Combined with observations of the surface CO₂ jet morphology and three-dimensional measurements of the frozen soil zone following the leakage, a quantitative analysis was performed to investigate the spatial expansion patterns and evolutionary mechanisms of the low-temperature zone under different operating conditions. ResultsSoil moisture content was positively correlated with the extent of the low-temperature zone; as moisture content increased from 7.80% to 15.15%, the maximum radial extent of the low-temperature zone increased by 64%, following an S-shaped growth pattern; at high moisture contents, the freezing of pore water altered the jet diffusion path, significantly enhancing the spatial asymmetry of the low-temperature zone; The direction of the leak directly determines the preferred direction of cold diffusion, with the peak circumferential temperature drop showing a high degree of consistency with the initial direction of the jet; the nominal diameter of the leak orifice is negatively correlated with the overall magnitude of soil temperature drop. For small-diameter leaks, CO₂ remains trapped within the soil, undergoing sufficient heat exchange to form a stable low-temperature zone; for large-diameter leaks, the jet breaks through the ground surface and the cold energy dissipates rapidly, resulting in a corresponding reduction in the magnitude of the temperature drop. Conclusion This study clarifies the mechanisms by which multiple parameters influence the temperature field resulting from leaks in buried CO₂ pipelines, providing an experimental basis for optimising the layout of leak monitoring systems, assessing low-temperature risks, and formulating emergency response and prevention plans for long-distance buried CO₂ pipelines.