The rapid and stable monitoring of CO₂ emissions from point sources in localized regions remains a key challenge in energy conservation and emission reduction efforts. To address this challenge, the Gaussian plume model is adopted for the rapid prediction of carbon emission dispersion from multiple point sources, and an inversion model for carbon emission intensities is constructed based on the Simplex search algorithm. By incorporating elevation data, the Gaussian plume model is modified to adapt to undulating mountainous terrain, and the impacts of the Gaussian diffusion model on the CO