A new 4D variational assimilation method of spaceborne data for moisture budget studies

For the last years, a great number of new space-borne platforms and radiometric instruments have been launched in order to document the dynamic, thermodynamic and microphysic properties of the atmosphere. For instance SSMI on DMSP, TMI on TRMM, MODIS on AQUA...brings informations on precipitation and water vapor; GOES, MSG...allow to access to horizontal winds; AIRS on AQUA, IASI on MetOp-1 to vertical profiles of temperature and humidity. Thus a need of new analyses to process and merge these data is appeared. In this context, this paper presents a new 4D variational data assimilation method to retrieve 4D fields from space-borne data such as wind, specific humidity and related quantities (pressure, temperature, vorticity potential, apparent heat source, apparent moisture sink, evaporation rate, condensation rate). This method is based on an analytical representation of these fields which are adjusted to the data through a variational process using additionnal constraints linked to physical equations (mass conservation, boundary condition, water mass conservation, vorticity production ...). This paper will focus on this mathematical formulation and its relevant avantages. Sensitivity tests on used dataset synthetised from a non-hydrostatic meso-scale numerical simulation will be discuss. An illustration will be also given on a particular real dataset (SSMI, TMI, GOES, QuikSCAT products and dropsoundings, radiosooundings) devoted to the documentation of the moisture budget of the BRET hurricane observed over the Gulf of Mexico in august 1999.