Geoscience Reference
In-Depth Information
In the operational mode, the Global Data Assimilation (GDAS) cycle runs
four times a day (00 UTC, 06 UTC, 12 UTC and 18 UTC). The analysis and
forecast for seven days are performed using the HPCS installed in IMD Delhi.
One GDAS cycle and seven days forecast (0 -168 hour) at T382L64 (~35 km
in horizontal over the tropics) takes about 30 minutes on IBM Power 6 (P6)
machine using 20 nodes with seven tasks (seven processors) per node, while
the same for GFS T574 (~25 km in horizontal over the tropics) is approximately
1 hour 40 minutes. Details of data presently being processed for GFS at
IMD are available at http://www.imd.gov.in/section/nhac/dynamic/data_
coverage.pdf .
3. Data and Methodology
Forecasted fields of sea-level pressure, 850 hPa wind, vorticity, and divergence
are examined for a monsoon depression formed over North BoB during 16-19
June 2011. In particular, day 1, day 3, day 5 and day 7 forecasts of these fields
are generated for the GFS. The initial analysis of a low-pressure system by the
CWC is taken as the time of cyclogenesis. In most cases, this initial analysis
occurred several hours prior to classification as a tropical depression. The model
forecasts are verified against the surface analyses produced by GDAS. Forecasts
of mean sea level pressure (mslp), 850 hPa wind and divergence fields are
verified against the corresponding satellite analyses from the Kalpana-1
Meteorological Satellite images.
The accuracy of a forecast for this study includes both spatial location and
timing of the model-generated vortex. The low-pressure system must have
developed within a six degree radius of the predicted location to be deemed an
accurate forecast. This radius was calculated from the average distance that an
easterly wave moves in a day. The forecasted trend in terms of location is also
noted, to determine the forecast accuracy relative to the analysis as well as to
each other. Consecutive forecasts of cyclogenesis events at different forecast
periods are deemed successful, as opposed to cases where a model predicts
cyclogenesis in a 5-day forecast, and then loses it until the 1-day forecast.
To determine how successful the GFS T574 and T382 models are at
anticipating tropical cyclogenesis, the forecasts of two deep depressions are
examined for the 2011 monsoon season. In this manner, the propensities of a
model to either over-develop or under-develop tropical lows can be determined.
In turn, the forecast skill of the models can be seen relative the models' ability
to handle cyclogenesis events as a whole, regardless of further development of
the low into a classified system.
4. Result and Discussions
4.1 Monsoon Depression (16-23 June)
In order to assess the ability of the model to capture cyclogenesis, episodes of
a monsoon depression and heavy rainfall are illustrated. The satellite imageries
Search WWH ::
Custom Search