Interactive
Forecast Preparation System (IFPS)
The creation
of NDFD starts with the formulation of digital forecasts at each
WFO. This, in turn, required the development and implementation of
efficient methods whereby the results of the thought processes of
WFO forecasters could be put into digital (or numerical) form. The
Interactive Forecast Preparation System (IFPS) has been developed
for this purpose. It has been under development for a number of years
(Ruth et al. 1998, Peroutka et al. 1998), but the processing power
of the Advanced Weather Interactive Processing System (AWIPS) was
necessary to make it feasible nationwide.
Interactive forecast
preparation represents a substantial change for forecasters. Instead
of manually typing a myriad of forecast products tailored for specific
user communities (e.g., public, aviation, marine), forecasters rely
on various data interpretation and editing techniques to prepare forecasts
of weather elements in a common digital database from which forecast
products will be automatically composed and formatted.
IFPS incorporates three methods for entering digital forecasts into the database
which are: matrix entry, graphical entry, and model interpretation.
Matrix entry--The
Interactive Computer Worded Forecast (ICWF) (Ruth and Peroutka 1993)
employed this method, making interactive earlier technology for producing
text from digital forecasts (Glahn 1970, 1979), and has been integrated
into IFPS. While this method was successful (Rezek 2002; Dickman 2002)
and still provides a method for "fine tuning" data entered
by other methods to specific times and locations, it has been largely
replaced by the graphical method.
Graphical entry
can insert forecasts into a gridded database at a spatial resolution
essentially bounded only by computer power (Wier et al. 1998). The
graphics depict an underlying grid of values. The forecaster works
with the graphics until he or she is satisfied with them. Then, the
underlying grid becomes the "forecast."
Model Interpretation
involves working directly with numerical model output (Ruth 1998) or
its interpretation (e.g., MOS) to forecast sensible weather elements.
Model interpretation, through IFPS, requires a higher level of understanding
and training and is not yet being used at most WFOs. As weather prediction
models improve, direct interaction with the models through this interface
will probably become more prevalent.
Each of these interfaces
provides a viable way of entering data into the digital database. The
forecasters at WFOs choose the interfaces they use. The forecasts can
have a time resolution of 1 hour at short ranges and 3 or 6 hours at
longer ranges. Spatial resolution is expected to be at least as fine
as 5 km with full implementation of IFPS in September 2003. These digital
forecasts automatically result in the routine textual and voiced forecasts
and provide a foundation for the development of a new generation of
grid-based NWS products, including the digital forecast database itself.
For more details
on IFPS, visit http://www.werh.noaa.gov/AWIPS/ifps/why2.htm
|