Improved Track and Intensity Predictions Using Cyclone Bogusing and Regional Assimilation - Monitoring and Prediction of Tropical Cyclones in the Indian Ocean and Climate Change - page 246

It is also noteworthy that the assimilation is deteriorating the results by
increasing the mean track errors. It is likely that the modified initial analysis
and the lateral boundary conditions create some discontinuity in the absence
of updated boundary conditions. Therefore bogusing and assimilation are carried
out to include the improved location and updated boundary conditions
(BOGVAR). The results suggest that the cyclonic vortex not present in CNTL
and VAR experiments is present in the IC after TC-bogusing is carried out
(BOG and BOGVAR experiments). However, in the BOG experiment the
cyclonic vortex introduced through bogusing technique is not sustained always
in the forecast period. The BOGVAR experiment suggests that this problem
can be overcome by using BOG along with VAR, may be because the fields
are more dynamically consistent after assimilation in VAR experiment. Figure
3(a-c) shows that BOGVAR, i.e. bogusing followed by assimilation, gives
significant reductions in track errors. The percentage reduction in track errors
in BOGVAR experiment over VAR experiment is given in Table 3.

Table 3 : Percentage reduction in track errors of BOGVAR experiments
over VAR experiments

Cyclone name % reduction in track errors of BOGVAR experiments

over VAR experiments

12 hr 24 hr 36 hr 48 hr

Laila 71.7 76.81 46.74 44.5

Giri 41.67 87.3 61.67 -

Phet 35.42 -7.81 51.55 9.96

Except for the case of cyclone Phet at 24 hr, all other values show reduced
track error in BOGVAR compared to VAR. It can also be seen that in case of
cyclone Laila, and Giri, the reduction in track error is as high as 70 to 90% in
the 24 hrs forecast. The official track forecast errors (km) provided by IMD for
'Giri' cyclone during FDP-2010 with lead time at 12 hourly intervals are 45
km (12 hr), 73 km (24 hr) and 68 km (36 hr) (Mohapatra et al., 2010) while the
respective track errors for Giri from BOGVAR experiment are found to be
43.64 km (12 hr), 17.96 km (24 hr) and 48.51 km (36 hr).

3.3 Impact of DWR Data Assimilation

Doppler weather radar (DWR) plays an important role for detecting and
forecasting severe weather, since it can probe atmosphere with high spatial
and temporal resolution. Assimilation of radial velocities (V) is primarily
important for triggering and controlling convection, while reflectivity (Z)
assimilation improves precipitation forecast (Sugimoto, 2007). Here an attempt
is made to assess the impact caused by assimilation of reflectivity and radial
velocity observations of DWR, for cyclones Jal (4-7 Nov 2010) and Thane
(27-30 Dec 2011). The DWR data from Chennai and Machilipatnam are used.