Abstract: The split-window algorithm has been widely applied for surface temperature inversion of various satellite payloads. The iterative simulation of large datasets during the fitting of split-window algorithm coefficients is often time-consuming and inefficient. Therefore, it is important to develop a highly efficient simplification method for split-window simulation computations. MODTRAN was to simulate and analyze the impact of variations in background parameters on total radiance. Subsequently, we performed a simulation analysis of the relationship between key parameters and total radiance under two adjacent thermal infrared channels of FY-3D MERSI-Ⅱ, exploring the variation patterns of total radiance under different coupling scenarios of these key parameters. Simulation results reveal that under the MERSI-Ⅱ thermal infrared channels, changes in land surface temperature have a greater impact on total radiance than surface emissivity. The effect of the atmospheric water vapor content concentration on the total radiance increases with an increase in both the land surface emissivity and land surface temperature, whereas the influence of the land surface emissivity and land surface temperature on the total radiance decreases as the atmospheric water vapor content concentration increases. When determining the coefficients for the split-window algorithm, narrowing the range of the atmospheric water vapor content concentration can reduce the number of simulations required, thereby enhancing the efficiency. For land surface temperatures ranging from 300 to 320 K, the atmospheric water vapor content concentration should be within 0.5 to 5.5 g/cm²; for temperatures ranging from 270 to 300 K, this range narrows to 0.5 to 4.0 g/cm². The saved simulation runs account for 18.23% of the total number of runs, which reduces the simulation time by 26 min. A comparison of the split-window coefficient fitting and absolute difference calculation results before and after simplification shows that the simplified scheme has minimal impact on the fitting outcomes.