Premium
The comparative impact of the assimilation of SSM/I and TMI brightness temperatures in the ECMWF 4D‐Var system
Author(s) -
Marécal Virginie,
Gérard Élisabeth,
Mahfouf JeanFrançois,
Bauer Peter
Publication year - 2001
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1002/qj.49712757322
Subject(s) - geopotential height , environmental science , microwave radiometer , data assimilation , special sensor microwave/imager , water vapor , meteorology , radiometer , brightness temperature , precipitation , humidity , liquid water path , remote sensing , microwave , climatology , atmospheric sciences , computer science , geography , physics , geology , telecommunications
This paper studies the impact of assimilating TRMM (Tropical Rainfall Measuring Mission) Microwave Imager (TMI) brightness temperatures in the European Centre for Medium‐Range Weather Forecasts fourdimensional variational assimilation system (4D‐Var). The methodology is similar to the one developed operationally for Special Sensor Microwave/Imager (SSM/I) radiometers. First, a one‐dimensional variational (1D‐Var) approach is used to retrieve specific‐humidity profiles, sea‐surface wind speed and cloud liquid‐water path. Then, the first two quantities are assimilated in the 4D‐Var analysis system. Results show an improvement of 1D‐Var product accuracy when using TMI instead of SSM/I brightness temperatures. This is related to the characteristics of the TMI radiometer which allow a better estimate of surface parameters and of water vapour in tropical conditions. Consequently, 4D‐Var analyses of the total column water vapour are improved using TMI data. The experiment with TMI observations gives a smaller increase in global humidity than the experiment with SSM/I data, leading to a reduction of the model precipitation spin‐down at the beginning of the forecasts. The impact on the low‐level wind analysis is small with either SSM/I or TMI data. Forecast performance with TMI data is generally improved compared with a control experiment without any SSM/I or TMI data. The TMI experiment scores are better than those obtained in the SSM/I experiment for the 1000 hPa geopotential in the northern hemisphere and for tropical winds at 200 hPa.