Application of IASI water vapour channels to the Unified Model data assimilation system

This study attempts to utilize optimal IASI channels to the Unified Model (UM) data assimilation system, focusing on the improvement of data assimilation performance in weather forecasting with the UM model. The IASI channels were selected from 314 channels suggested by Collard (2007) using the degree of improvement of retrieved atmospheric parameters from 1 D-Var over background atmospheric parameters. In doing so, IASI 191 channels were selected, including additional water vapour channels and short-wave IR channels (IASI band 3) as compared with operational IASI 183 channels that are operationally used in the UM data assimilation system. In order to assess the impact of newly selected IASI channels on the data assimilation, the data assimilation trial test was performed for summer and winter period of 2013. It was found that additional water vapour channels in newly selected IASI channel have an impact on reducing humid biases of relative humidity in the upper troposphere that occur in the output produced from the data assimilation system with operational IASI channels. However, in the O-B analysis, the mean biases of water vapour channels of AMSU-B, CrIS, and AIRS sensors appear to be large negative in contrast with slight positive biases with operational IASI channels. In order to improve large negative biases of water vapour channels of various satellite sensors, it is necessary to recalculate the coefficients of predictors that are operationally used in the process of bias correction.