ASSIMILATION OF SEVIRI RADIANCES OVER LAND IN THE M′ET′EO-FRANCE MESOSCALE MODELS

摘要

This work aims to improve the assimilation of low-levelrnSEVIRI (Spinning Enhanced Visible and Infra Red Imager)rnIR (Infra-red) observations over land to better constrainrnatmospheric analyses in mesoscale models operatedrnat M′et′eo-France. To date, only high-peaking WaterrnVapour channels are operationally assimilated over landrnand IR channels are entirely rejected over land surfaces.rnThe assimilation of IR observations over land is possiblernonly if several limitations are accounted for: a reliable descriptionrnof the surface emissivity, an accurate estimationrnof the surface temperature and an effective bias correctionrnscheme.rnSome feasibility studies have been undertaken in orderrnto assimilate more SEVIRI observations in the ALADINrnand AROME French systems. The land surface emissivityrnwas described using climatologies from the EUMETSATrnLand-SAF (Satellite Application Facilities). Thernland surface emissivity and SEVIRI Tb were used as inputrnparameter in the radiative transfer model to retrievernthe surface temperature (Ts) over Europe. The retrievedrnTs was compared with independent Ts estimates (Land-rnSAF products, T2m…) and was then used within thernassimilation process to constrain the analysis of surfacerntemperature.rnSeveral assimilation experiments were run during bothrnsummer and winter periods on the basis of the operationalrnversions of ALADIN and AROME. We show thatrnthe forecast impact is generally neutral to positive, especiallyrnduring summer. In particular, SEVIRI data pointrnto positive impact on humidity fields over Southern Europe.rnSEVIRI data are also shown to improve the qualityrnof analyses, particularly those of Total Column WaterrnVapour (TCWV), and this is substantiated through comparisonsrnwith independent GPS measurements. With regardrnto the spanish synoptic stations, forecast of temperaturernand geopotential are improved.