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 676

without any further assimilation of the satellite data. The third experiment, named
the MODIS run, ingests the temperature and humidity observations from MODIS
in a 12 hourly interval from 18 June 2007 18 UTC to 19 June 2007 18 UTC and
the model is subsequently run in a free forecast mode up to 22 June 2007 06 UTC.
The high resolution MODIS data is subjected to thinning before ingesting the data
to the observational pre-processor. The fourth experiment (SSM/I) is similar to the
ATOVS experiment except that it incorporates SSM/I total precipitable water instead
of ATOVS temperature and humidity observations. All the model runs include 6 h
interval data cycling to maintain the dynamical consistency of the model simulation.
The CTRL run is subjected to data cycling without any assimilation of observations.
The results presented are from 12 km resolution domain only.

26.4.3.2 Results and Discussion

Initial Conditions

As in previous cases, the ATOVS experiment depicts larger spatial distribution of
positive analysis increment values of 850 hPa wind speed with negative increment
seen towards the south of the domain. For the MODIS run, most of the positive
increment values of 850 hPa wind speed are seen towards north of the domain with
negative values concentrated towards the southern tip of the Indian peninsula and
the adjacent Bay of Bengal. As seen in the previous cases, the assimilation of SSM/I
total precipitable water shows very small increment values of 850 hPa wind speed
over the domain. The differences in the initial conditions are presented in Fig. 26.19
for the 18-22 June 2007 depression.

MSLP

Figure 26.20 a-d show the spatial distribution of MSLP patterns of the depression
from the NCEP-FNL analysis. An intense monsoon depression with a minimum
central pressure of 990 hPa is shown in the figure. As per the NCEP analysis, the
system made the landfall after 22 June 2007 00 UTC which is consistent with the
observation records. The results of all the numerical experiments from Fig. 26.21
show that the intensity of the depression is over estimated in all the experiments
except the ATOVS run. The ATOVS experiment did predict the intensity of the
depression with almost the same magnitude as that of the NCEP-FNL analysis.
None of the four experiments can accurately predict the landfall of the depression.
Moreover, a slow movement of depression is seen prominently in the MODIS run.

Wind Speed

The spatial pattern of the 850 hPa wind vector simulated by the numerical
experiments is shown in Fig. 26.22 for the 18-22 June 2007 depression. The NCEP-
FNL wind pattern showed landfall of the cyclonic system on 22 June 2007 00 UTC