Geoscience Reference
In-Depth Information
CTRL experiment for all the levels except 200 hPa. The MODIS run also shows
improvements in the humidity profiles with lower RMSE of dew point temperature
at almost all the levels except 950 and 850 hPa due to assimilation. The SSM/I run
shows lower error values of dew point temperature at 950, 920, 850, 780, 700, 400
and 300 hPa levels against the CTRL run on 03 September 2006 12 UTC. On the
second day of forecast, the ATOVS run has produced significant improvements
in the dew-point temperature profile by reducing the RMSE of the dew point
temperature at all the levels except 950 and 300 hPa. The MODIS and the SSM/I
run have shown higher RMSE of dew-point temperature at almost all the levels.
However, on the second day of the forecast, the assimilation experiments have
failed to show marked improvements in the humidity profiles on 5 September 2006
12 UTC as compared to the CTRL experiment.
26.4.3
Depression That Formed During 18 to 22 June 2007
The depression that formed over the Bay of Bengal during the third week of
June 2007 was initially seen as a low pressure area over the east central Bay and
neighbourhood on 20 June 2007. The low pressure area intensified to a depression
and lay centred at
;86:0
ı
E on 21 June 2007 03 UTC. The depression further
intensified into a deep depression and lay centred over
15:5
ı
N
84:0
ı
Eat12UTC
on the same day. At 03 UTC of 22 June 2007, the system was over coastal Andhra
Pradesh. The system subsequently moved westwards.
16
ı
N,
26.4.3.1
Numerical Experiments
The model settings and the domain used are the same as that for the two previous
cases. The NCEP-GFS forecast data available at a horizontal resolution of
1
ı
1
ı
and a time resolution of 6 h are used to develop the initial and lateral boundary
conditions. Four numerical experiments are performed to study the impact of
temperature, humidity and total precipitable water in the simulated structure of
monsoon depression. All the simulations are started with the same initial conditions
on 18 June 2007 18 UTC, but the observations are assimilated at different times
depending on the availability of the satellite data over the domain. The CTRL
experiment, has utilized the NCEP-GFS data for creating initial and lateral boundary
conditions. The model integrations are performed from 18 June 2007 18 UTC to 22
June 2007 06 UTC without any assimilation of satellite observations. The second
experiment, called the ATOVS run is started from 18 June 2007 18 UTC. For the
ATOVS run, the ATOVS temperature and humidity profiles have been ingested
initially into the model using 3D-VAR after 6 h of forecast, i.e., on 19 June 2007
00 UTC and again up to 20 June 2007, 00 UTC in a 12 hourly interval. The WRF
model is then subsequently integrated for the next 54 h in a free forecast mode
Search WWH ::
Custom Search