Using GPS-IPW in a 4-D data assimilation system

The NOAA Forecast Systems Laboratory (FSL) has been continuously calculating integrated atmospheric precipitable water (IPW) from GPS signal delays since 1994. Using rapid orbit information, these data have the accuracy required for use in a numerical weather prediction model through data assimilation. Parallel cycles with and without GPS-IPW data have been running at FSL since November 1997 using the 60-km Rapid Update Cycle (RUC). Verification of the analysis and the 3, 6, and 12-h forecasts against rawinsondes has been ongoing throughout the experiment. Results from these statistics show a consistent improvement in short-range forecasts of relative humidity when the GPS data are included. Precipitation verification has also been calculated for this experiment, and results show that GPS data also improve these forecasts. Recently, the average number of available GPS observations jumped from 18 to 56, and results for November-December 1999 show that the previous slight positive signal is now amplified to a stronger positive impact on the short-range moisture forecasts.