An updated atmospheric transmittance model including sulfur dioxide absorption developed for geostationary satellite hyperspectral infrared sounders

In numerical weather prediction systems, the assimilation of hyperspectral infrared sounding data relies on rapid radiative transfer models. A key component of these models is the atmospheric transmittance coefficient, which directly influences the calculation of absorption characteristics across various spectral bands and impacts the accuracy of data assimilation and meteorological forecasts. In this study, an improved method is developed to predict the total atmospheric gaseous transmittance of the Advanced Radiative Transfer Modeling System (ARMS) by incorporating additional sulfur dioxide ( S O 2 ). The statistical regression-based transmittance predictors are enhanced with a correction term, which reduces training time while improving simulation accuracy. Compared to the Line-By-Line Radiative Transfer Model (LBLRTM), the updated transmittance model shows significant improvements in the 850–925 c m − 1 spectral range. Specifically, the new model achieves a mean brightness temperature deviation of approximately −0.25 K , compared to the sharp transition from −0.75 K to 0.25 K observed in the current original version of ARMS. The Radiative Transfer for TIROS Operational Vertical Sounder (RTTOV) model also exhibits notable fluctuations in this range, with its mean deviation gradually approaching 0 K from around −0.75 K . In the 950–1065 c m − 1 range, the new model outperforms alternatives, achieving a mean standard deviation of 0.25 K , significantly lower than the original ARMS error of approximately 1 K . RTTOV displays considerable variability in this range, with its mean deviation oscillating between −0.5 K and 0.5 K . Overall, the updated atmospheric transmittance coefficients reduce the mean biases with respect to LBLRTM, offering an enhanced accuracy compared to both the original ARMS version and the RTTOV model.