To address the issue of insufficient inversion accuracy of CO₂ leak source parameters in CCUS (carbon capture, utilization, and storage) projects, a hybrid optimization method combining SSA (sparrow search algorithm) and PSO (particle swarm optimization) was proposed. First, a forward leakage model was constructed based on the Gaussian plume dispersion model to analyze the influence of leakage rate, wind speed, and atmospheric stability on surface CO₂ concentration distribution. Then, combining the global search capability of SSA with the local development advantages of PSO, an inversion model for leak source location and leakage rate was established. Simulation results show that the SSA-PSO algorithm achieves a relative error of only 0.01% for leakage rate, and relative errors of 0.10% for leak source coordinates. Compared with PSO, SSA, GA, and GA-PSO algorithms, the leakage rate inversion accuracy is significantly improved, with a maximum relative error reduction of 27.99%, and leak source location accuracy increased by up to 22.3%. Case studies further demonstrate that the relative errors for leak rate and location inversion using this algorithm are 10.50%, 4.00%, and 3.00%, respectively, all better than the comparison algorithms, with higher computational efficiency. The study indicates that the SSA-PSO hybrid algorithm can achieve fast and accurate localization of CO₂ leak sources in CCUS projects, providing effective technical support for emergency response and risk prevention of leaks.