FOUR-DIMENSIONAL VARIATIONAL DATA ASSIMILATION OF GROUND-BASED GPS WATER VAPOR DATA IN THE JUNE 12TH IHOP CASE

Slant-path water vapor data from 32 GPS receivers to each satellite in view were obtained every 10 min in the June 12th IHOP case and assimilated into the MM5 4DVAR system.

It is found that GPS slant-path water vapor plays a crucial role in increasing moisture in the squall line area around the KS-OK border, which leads to the improvement of the subsequent prediction of moisture and rainfall. The positive impact of data assimilation lasted for more than 6 hours. The impact of the GPS water vapor data is sensitive to the initial time of the assimilation window and to the grid resolution of the assimilation model. We need high enough resolution (15-km grid in this simulation) to resolve the scale that the slant-path water vapor can characterize.

In the comparison with the 4DVAR of precipitable water obtained from the same 32 GPS sites at 30-min intervals, the 4DVAR of slant-path data is superior in terms of describing the horizontal distribution of moisture.

The optimal initial conditions obtained from the 4DVAR experiments imply that accurate water vapor information just before the squall line forms (21 UTC 12 June, in this case) is critical for the improvement of the severe weather forecast.