At the same time, the analysis and numerical weather prediction
(NWP) communities have conducted several field programs over marine
regions adjacent to the CONUS. These studies have suggested that
forecast errors for the CONUS region are very sensitive to the initial
conditions over the adjacent oceans, in particular the upstream Pacific
Ocean. Results from the 1998 North Pacific Experiment (NORPEX-98)
indicated that satellite winds can have a significant impact on NWP,
even when the data are only made available to a data assimilation system
every 6 hours.

The Pacific Landfalling Jet Experiment (PACJET), planned to occur
January-February 2003, is a field experiment sponsored by NOAA Research,
NESDIS, the United States Weather Research Program (USWRP), and the
National Weather Service (NWS). The goal of this experiment is aimed at
developing and testing methods to improve short-term (0-24 h) forecasts
of damaging weather on the U. S. West Coast in landfalling winter storms
emerging from the data sparse Pacific Ocean. PACJET also includes data
impact assessments based on use of new data in NWP, and by forecasters.
In support of this experiment, GWINDEX-3 will employ a scanning strategy
for GOES-10 that will include a series of rapid scans (RS) every three hours,
allowing for the generation of high-quality satellite wind sets every
three hours. The increased temporal resolution of these satellite wind products
offers significant potential for having a positive impact in rapidly
developing 4D assimilation systems that provide the initial conditions
for NWP. In addition, NWS field forecasters can take advantage of the
high temporal nature of these products to improve their forecasts
through assessment efforts built into the PACJET experiment. We will
work with PACJET scientists and the NWS forecast community towards
validating the quality and usefulness of the RS satellite wind products
using ground truth measurements made available for PACJET. We will also
provide the three hourly RS wind datasets to NOAA's Forecast System Laboratory
(FSL) for inclusion in the Rapid Update Cycle (RUC) model. Data impact
evaluation will follow.

NESDIS/OSO has implemented the proposed R/S scanning strategy for
GOES-10. The schedule provides one loop of 3 images at 7.5-minute
frequency every three hours around the clock over the eastern North Pacific
region for winds determinations. This schedule will be activated for a
three-month period starting in early January, and should not adversely
affect the routine NWS GOES-10 imaging requirements, although of course,
this will need to be verified by the NWS.

In the short term, GWINDEX-3 will afford the opportunity to
demonstrate improved impact of GOES RS data on quantities such as wind
vectors. If successful, a more permanent RS schedule could be envisioned
for routine winds production from both the east and west GOES (in
anticipation of major USWRP planned field programs to examine the
optimal mix of observations, such as THORPEX). There is a movement in
the international community to utilize geostationary satellites more
effectively. We believe NESDIS will take a global lead in demonstrating
the value of RS products in anticipation of next generation satellite
systems that are planned.

Christopher Velden - University of Wisconsin-CIMSS