Home page  
Home   Your Room   Login   Contact   Feedback   Site Map   Search:  
Discover this product  
About Us
Overview
Getting here
Committees
Products
Forecasts
Order Data
Order Software
Services
Computing
Archive
PrepIFS
Research
Modelling
Reanalysis
Seasonal
Publications
Newsletters
Manuals
Library
News&Events
Calendar
Employment
Open Tenders
   
Home > Newsevents > Training > Rcourse_notes > DATA_ASSIMILATION > INVERSION_METHODS >  
   

Inversion methods for satellite sounding data
April 1991


By J. R. Eyre




 
  Training Course Notes Front Page >>
Table of contents >>
Next Section >>
Previous Section >>



5 . Satellite sounding data in numerical weather prediction

In recent years it has become increasingly difficult to show that temperature/humidity profiles retrieved from satellite sounding data have a positive impact within operational NWP data assimilation systems, particularly in areas where other observation types are available. This problem has arisen partly because NWP systems have improved to a point at which great care is required in the treatment of any observation type; appropriate quality control is necessary and the error characteristics of each data type must be taken into account. However, there are additional problems with satellite sounding data.

As discussed in section 1.4, the intrinsic vertical resolution of the satellite sounding system is low, both in relation to other temperature sounding observations (i.e. radiosondes) and to the vertical resolutions of modern NWP models. Because of this the background and constraint information used in the inversion affects the retrieved profile considerably. In practice low-order vertical structures in the retrieval are obtained mainly from the radiance information, but high-order structures come largely from the background information. Consequently, considerable care must be taken to avoid components of the retrieved profile which are not derived from the radiance data, but are artefacts of the inversion method, contaminating an otherwise good NWP analysis. Another symptom of the same problem is that retrieved profiles have systematic error structures of a very subtle and specific nature (see Eyre 1987). It is difficult for many analysis systems to suppress the harmful effects of these error characteristics without simultaneously losing the real information contained in the radiance data.

Recent developments in NWP data assimilation seek to solve these problems by making more direct use within the NWP system of radiance observations themselves, rather than retrieved temperature profiles. The French Direction de la Météorologie (Durand 1986) and the UK Meteorological Office (Eyre and Lorenc 1989) both run operational TOVS processing and assimilation systems based on these ideas. At ECMWF, systems for both one- and three-dimensional variational analysis of TOVS radiances are being developed, based on the theory presented in section 2.3 (b) (see Eyre 1990, Pailleux 1990).


Training Course Notes Front Page >>
Table of contents >>
Next Section >>
Previous Section >>







 

Top of page 07.06.2002
 
   Page Details         © ECMWF
shim shim shim