Additional impacts of all‐sky assimilation of temperature‐sounding channels from hyperspectral infrared sounders

Abstract This study assesses the impacts of all‐sky infrared radiance assimilation of temperature‐sounding channels, which have received far less attention than those of water vapour (WV) channels. We extend the all‐sky assimilation framework developed for WV channels to temperature‐sounding channels sensitive to the mid‐ and upper troposphere, and evaluate the separate and combined impacts of these channels in the Japan Meteorological Agency's global numerical weather prediction system. The all‐sky assimilation of temperature‐sounding channels improves short‐ to medium‐range forecasts relative to clear‐sky assimilation, although its overall impact remains smaller than that obtained from all‐sky assimilation of WV channels. When both channel types are assimilated together, the overall forecast skill is further enhanced. The weaker impact of temperature‐sounding channels in comparison with WV channels is attributed to their limited direct sensitivity to WV, their comparable influence on temperature analysis, and the larger influence of residual biases arising from suboptimal bias correction in combination with smaller observation errors. In addition, a strong dependence on background clouds due to the locality of Jacobians poses another challenge for the effective use of temperature‐sounding channels. These results clarify the role and limitations of temperature‐sounding channels in the all‐sky assimilation context.