Buried natural gas pipelines are susceptible to leakage caused by natural or anthropogenic factors. Significant safety hazards are posed by these leakage events. In order to address the insufficient consideration of gas-soil heat transfer characteristics and the lack of theoretical guidance for distributed fiber optic detection system deployment, numerical simulation was used to investigate the effects of transmission temperature and soil temperature on leakage diffusion behavior and fiber optic alarm response. The results show that the mass flow rate at the leakage hole increases compared to the adiabatic model when heat transfer characteristics are considered. The consistency between simulation results and experimental data is improved. The mass flow rate at the leakage hole increases as the transmission temperature decreases. The influence of soil temperature on the leakage mass flow rate is relatively small, with a difference of less than 1%. The diffusion range of natural gas decreases with the increase of transmission temperature. The range expands with the rise of soil temperature. Both the decrease in transmission temperature and the increase in soil temperature lead to a delay in the fiber optic alarm. Specifically, the alarm response time is extended by 58.1 s when the transmission temperature decreases by 30 K. The response time is prolonged by 10 s when the soil temperature increases by 30 K. The maximum error of the prediction model for the buried fiber optic alarm time is 10%, and the average error is 4.18%. The accuracy of the prediction results is verified. It is concluded that thermal characteristics significantly affect the leakage process. This study provides a theoretical basis for the division of leakage emergency areas and the optimization of distributed fiber optic detection system deployment.