Soundings
4. Lydie Lavanant
|
|
At CMS, three-dimensional fields of temperature and humidity are produced in a near-real time way since the beginning of 2009 from the acquisition of the advanced IASI interferometer and microwave AMSU / MHS sounders of METOP by using a 1dVar method. The purpose of this chain is to provide the satellite retrieved atmospheric fields before the release of the analysis and especially if possible with more realistic estimations of humidity and instability of the air, parameters which are useful for Nowcasting applications.
A Météo-France internal Web server displays dynamically and in real time all necessary fields for helping French Nowcasters. Besides an ‘external clone’ (http://retmetop.meteo-spatiale.fr/ ) of the site is also available for everybody outside Météo-France with the products of the last ten days over the ‘Europe-Atlantic’ and CMS HRPT area.
In the assimilation of the numerical weather prediction models, the retrieval of the atmospheric profiles from the data of satellites is very preservative with regard to the initialization profile. For example, in order not to perturb the model with regard to the other sensors, a very weak number of IASI channels sensitive to the water vapor molecules is used and to these channels are associated important errors via the error covariance matrices of observations, what limits the impact of these channels. As a consequence the fluctuations in the humidity profiles, which are very important for Nowcasters, are poorly evaluated through the assimilation of the satellite data. This chain is used to experiment new processing methods for better taking into account their needs.
The figures correspond to the processing of two IASI spectra co-registered to a dropsonde during the Jaivex experiment. The bottom figures show the impact on the retrieved water vapor profiles of a dynamic selection of 150 channels according to the processed situation from the set of 366 channels selected before the launch of METOP (case B on the figure) and of a subset of 1700 channels sensitive to the temperature and to the water vapor only (case C on the figure). The case A is set as a "reference" with 178 channels and the two error covariance matrices used by ECMWF. IASI contribution with regard to that of the initialization profile is displayed on the upper figures.
References:
Lavanant L., Towards a better retrieval of fine water vapor atmospheric structures using IASI. IASI conference n°2 proceedings. 2010.