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 645

depicts the analysis increments due to assimilation of the satellite observations
such as MODIS, ATOVS and SSM/I on the 850 hPa wind speed; the figure shows
clear differences in the wind speed at 850 hPa. In contrast to the increment due
to the assimilation of SSM/I TPW, which are negligibly small, the assimilation
of MODIS and ATOVS temperature and humidity profiles have shown significant
and marked analysis increment values over the domain of study. Among the three
3D-VAR experiments, the ATOVS 3D-VAR assimilation has resulted in the largest
spatial extent of positive value of increments up to ms 1 and the largest negative 2:5

ms 1 when compared to the MODIS and the SSM/I

values of increments up to 2

3D-VAR experiments.

Mean Sea Level Pressure (MSLP) fields

Figure 26.3 shows the MSLP (a-d) and the lower tropospheric wind at 850 hPa
(e-h) from NCEP-FNL (Final Analysis), and 24 h accumulated precipitation from
TRMM satellite (i and j) for the 19-22 September 2006 depression. The depression
is initially shown as a low pressure area with a minimum central MSLP of 1,000 hPa,
which later intensified into a depression with a central minimum pressure of 996 hPa
as evident from Fig. 26.3 a-d. The maximum accumulated precipitation of TRMM
satellite observations are seen over the west coast of the Bay of Bengal during 20-22
September 2006.
Figure 26.4 a-p depicts the simulated MSLP patterns on 20 September 2006
00 UTC and subsequent predicted values at 24, 48 and 60 h of forecast for the
CTRL experiment (Fig. 26.4 a-d), the ATOVS experiment (Fig. 26.4 e-h), the SSM/I
(Fig. 26.4 i-l) experiment and the MODIS (Fig. 26.4 m-p) experiment, respectively.
The simulated MSLP fields are then compared with the NCEP FNL analysis. The
CTRL run simulates an intense depression with a minimum central pressure of
990 hPa. The system as simulated in the CTRL run intensifies in the 24, 48 and 60 h
of forecast. The ATOVS experiment simulates the weakest depression of all the four
numerical experiments, with a lowest central minimum pressure of 998 hPa. The
simulated MSLP pattern of the SSM/I experiment is similar to that of the CTRL
experiment, clearly indicating that the ingestion of the SSM/I total precipitable
water has not produced significant impact in the numerical simulation. The MODIS
experiment however simulates a well organised monsoon depression with the well
known west-north-westward movement of the depression ( Godbole 1977 ). Also,
the results of the MODIS experiment simulate the depression well in terms of the
location of the depression center and not with respect to the intensity. Among the
four experiments, the MODIS run shows the maximum inland penetration, though
there are some inadequacies in accurately simulating the location of the depression
centre vis-a-vis the NCEP-FNL analysis. In terms of intensity of MSLP, the ATOVS
experiment results are quite close to the NCEP-FNL values than the other three
experiments.