Studies on the Impacts of 3D-VAR Assimilation of Satellite Observations on the Simulation of Monsoon Depressions over India - Data Assimilation for Atmospheric, Oceanic and Hydrologic Applications - page 640

26.3 Literature Study of Earlier Assimilation Studies

Numerical Weather Prediction, especially in the short-range requires accurate
initial conditions. A large number of studies have shown that assimilating various
observations such as satellite data, Doppler Weather Data have improved the initial
conditions and have resulted in better model performance ( Gal-Chen et al. 1986 ;
Lipton and Vonder Haar 1990 ; Lipton et al. 1995 ; Ruggiero et al. 1999 ; Zou and
Xiao 2000 ; Pu et al. 2002 ; Fan and Tilley 2005 ; Chou et al. 2006 ; Chen 2007 ; Zhang
et al. 2007 ; Zapotocny et al. 2007 ; Govindankutty et al. 2008 ; Kelly et al. 2008 ;
Singh et al. 2008a ; Singh et al. 2008b ; Brennan et al. 2009 ; Rakesh et al. 2009a ;
Sinha and Chandrasekar 2010 ; Singh et al. 2010 ; Govindankutty et al. 2010 ; Singh
et al. 2011a , b ; Kumar et al. 2011 ; Singh et al. 2011c ). Chen ( 2007 ) investigated and
compared the impact of assimilating SSM/I, and the QuikSCAT satellite surface
winds, on the simulations of Hurricane Isidore. The results of the above study
indicated that the increment of the QuikSCAT wind analysis was higher than that
from the SSM/I analysis. Furthermore, the results also showed that the increase
in low-level wind speeds enhanced the air-sea interaction processes and improved
the simulated intensity for the hurricane in the assimilated QuikSCAT run. Also,
the non-availability of the surface wind direction information from the SSM/I data
resulted in less improved simulation as compared to the QuikSCAT assimilated run.
The above study showed that the position of the center of hurricane over the ocean
which is usually misrepresented at the model initial time can be improved due to
assimilation of high-resolution surface wind information.
Rakesh et al. ( 2009b ) investigated the impact of assimilating QuikSCAT surface
wind vectors, SSM/I wind speed and the Total Precipitable Water (TPW) for
forecasts of wind, temperature, and humidity from 1 month long assimilation
experiments during July 2006. In the above study, the control (without assimilation
of satellite data) as well as 3D-Var sensitivity experiments (with assimilation of
satellite data) using MM5/WRF were made for 48 h starting daily at 0000 UTC
July 2006. Rakesh et al. ( 2009b ) utilized the control run results as a baseline for
assessing the impact of MM5/WRF 3D-Var satellite data sensitivity experiments.
Rakesh et al. ( 2009b ) found from their results that the forecast errors in predicted
wind, temperature and humidity at different levels are lower in WRF model as
compared to the MM5 model, except for the temperature prediction at lower
level. Also, their results indicated that the rainfall pattern and prediction skill
from day one and day two forecasts by WRF model is superior to the MM5
model. Furthermore, Rakesh et al. ( 2009b ) found that the spatial distribution of
forecast impact for wind, temperature, and humidity fields showed that on an
average, for 24 and 48-h forecasts, the assimilation of satellite data did improve
the MM5/WRF initial conditions and resulted in reduced errors of the predicted
meteorological fields. Among the assimilation experiments, MM5/WRF wind speed
prediction was found most beneficial due to ingestion of QuikSCAT surface wind
and SSM/I TPW data while for the temperature and humidity prediction the