Role of Sea Surface Temperature in Simulation of Arabian Sea Cyclone - Monitoring and Prediction of Tropical Cyclones in the Indian Ocean and Climate Change - page 342

Fig. 6: The track of the Arabian Sea cyclone is shown at 12 hourly intervals
obtained from JTWC and model simulations with different SSTs (track
of same SST experiments is not here since it does not produce any cyclone).
The track is shown from 12UTC of 5th May 2002 for model simulated
results and 00UTC of 6th May for JTWC to 00UTC of 11th May of 2002
(JTWC location of the system at 12UTC of 5th May is unavailable).
compared to observed cyclone location. The movement of the cyclone in
SSHA2SST experiments is westwards initially and then northwest wards till
the landfall. In the Control experiment, the landfall of the cyclone occurs far
northwards, while in SSHA2SST experiment, the landfall occurs little
southwards as compared to observation. However no landfall of the cyclone is
found in Inserted SST grad experiment. It is noticed from the results that the
translational motion of the cyclone is faster in Control than other two
experiments. Figure 6 indicates that the track of the cyclone obtained from the
SSHA2SST experiment is closer to observation. In order to estimate the
quantitative error in the location, displacement error of the track has been
computed for all experiments with respect to JTWC track and shown in
Table 2. The initial error is 320 km in all the experiments which indicates that
the model integration has been conducted with large error in vortex location.
On day-1, the displacement error is 1025, 610 and 301 km for control, Inserted
SST grad and SSHA2SST experiments, respectively. This result illustrates that
the initial error is largely magnified in Control experiment than other two
experiments. This may be due to presence of a positive SST gradient towards