|Title||Operational implementation of RTTOV-10 in the IFS|
|Year of Publication||2011|
|Authors||Bormann, N, Geer, A, Wilhelmsson, T|
|Secondary Title||Technical Memorandum|
This memorandum summarises the evaluation of RTTOV-10 for use in the operational assimilation system at ECMWF. The new features of RTTOV-10 evaluated here are a revised treatment of the top of-the-atmosphere, and a new sea surface emissivity model for microwave frequencies (FASTEM-4). The new version of RTTOV is evaluated through comparisons of radiative transfer simulations with RTTOV-9, and through an analysis of departure characteristics against observations. The impact on forecasts is also investigated through a series of assimilation experiments. The use of FASTEM-4 leads to significantly different bias characteristics, primarily for the microwave imager channels used in the ECMWF system. With FASTEM-4, brightness temperature biases against observations are overall within the size of biases between different instruments for imagers, even though biases for some channels on some sensors are degraded compared to using FASTEM-2. The benefits of FASTEM-4 compared to FASTEM-3 are less clear for AMSU-A for which significant biases remain for the window channels. The changes to the treatment of the top-of-the-atmosphere in RTTOV-10 lead to minor changes to the departure characteristics of channels with some sensitivity to the top of the atmospheric model. Nevertheless, changes are noticeable in the mean temperature analyses above 1 hPa. The impact of RTTOV-10 on medium-range forecasts is neutral.