The information content of clear sky IASI radiances and their potential for numerical weather prediction

The estimation-error covariance matrix associated with the analysis of the atmospheric state is used in one particular meteorological situation to examine the potential benefit of radiance data for numerical weather prediction. The gain of information content obtained from simulated Infrared Atmospheric Sounding Interferometer (IASI) data is studied and compared with the current information content present in the TIROS Operational Vertical Sounder (TOVS) radiances. Nineteen independent items of information on a typical temperature/humidity profile are available from the IASI data, compared to six from the TOVS data. In terms of temperature, fine-scale structures associated with a vertical resolution of 1 km are estimated with a 0.7 K error standard deviation. The gain of information for specific humidity is of the same order of magnitude as for temperature. Typically, humidity structures associated with 1 km vertical resolution are estimated with a relative error of 16%. The projection of the analysis-error covariance matrix on atmospheric-error structures (relevant for numerical weather prediction) gives a measure of the impact of the use of radiances on the observability of such structures. The results indicate that IASI data would be a decisive source of information for the analysis of such structures. Finally, a preliminary sensitivity study suggests that the degradation due to radiance noise associated with possible modifications of the IASI instrument, hardly affects the quality of the analysis.