Abstract: Severe convective weather systems such as tropical cyclones are often accompanied by intense precipitation. Accurate retrieval of sea surface rainfall rate at high temporal and spatial resolutions is crucial for disaster prevention and mitigation, numerical weather prediction, and the investigation of the fine-scale structure of typhoons. In this study, a coupled cloud–rain atmospheric radiative transfer model combined with a nonlinear least-squares inversion technique is developed to retrieve sea surface rainfall rate under tropical cyclone conditions. To address the overestimation of surface emissivity by the FASTEM-5 model at moderate-to-high wind speeds (25 ~ 45 m·s^-1), a wind-speed-dependent emissivity correction term is introduced.The corrected emissivity model exhibits improved consistency with that derived from SFMR (Stepped-Frequency Microwave Radiometer) observations, with a mean absolute error less than 0.02. Furthermore, the absorption effects of oxygen, water vapor, and cloud liquid water are incorporated into the radiative transfer calculations, and the corrected emissivity model is used to simulate brightness temperatures. The results show that the simulated brightness temperatures exhibit mean biases ranging from −1.56 K to −2.14 K and standard deviations between 1.64 K and 1.87 K relative to observations. Based on SFMR flight data collected from multiple tropical cyclones during 2014 ~ 2023, the retrieved rainfall rates exhibit a root-mean-square error (RMSE) of 2.54 mm·h^-1 and a correlation coefficient of 0.915, demonstrating the robustness and accuracy of the proposed rainfall retrieval algorithm in typhoon environments.