To analyze the effects of the real spectral response of microwave remote sensors on microwave radiance simulation in the assimilation modules of numerical weather prediction systems, the real and ideal spectral responses of the fourth channel, which reflects the atmospheric and terrestrial radiance simultaneously, of FengYun-3D Microwave Temperature Sounders-2(FY-3D MWTS-2) were applied. Based on line-by-line atmospheric absorption models and predictors of Radiative Transfer for TOVS(RTTOV), different optical depth fitting parameters by real and ideal spectral responses of FY-3D MWTS-2 were generated by applying the least square method. Radiance background fields by real and ideal spectral responses can be simulated from four dimension variation assimilation module of the China Meteorological Administration-Global Forecast System (CMA-GFS) by using the Final (FNL) global operational analysis dataset from the National Center of Environmental Prediction(NCEP) as the meteorological background fields. Observation Minus Background (OMB) under different sea and land types could be calculated using the observations of FY-3D/MWTS-2 and simulated radiance backgrounds. The results show that microwave radiance backgrounds simulated by real spectral response were higher than ideal spectral response under low radiance observation situations. Simulated radiance backgrounds were closer to observations when the radiance observation increased. Radiance backgrounds by real spectral response were closer to radiance observations than by ideal spectral response under global region except for a small part of the Antarctic region. Compared with using ideal spectral response, when the real spectral response was applied, the Root Mean Square Error (RMSE) of OMB on land region decreased by 1.8%; the RMSE of OMB in sea region decreased by 2.6%; the RMSE of OMB in land sea border region decreased by 4.72%, the total RMSE of OMB decreased by 2.17%. Standard Deviation (STDV) and Mean Absolute Error (MAE) also decreased by 0.75% and 4.6%.