data in January 2002, NCEP implemented a higher resolution global forecast model of T254, L64 configuration in October 2002. This effectively increased the horizontal resolution to about 60 km from the previous T170, L42, which is 100 km in the horizontal resolution. It was then deemed necessary to investigate the impact of higher resolution (50 km) with the purpose of making effective use of the inherent mesoscale features associated with the QuikSCAT wind data. Two parallel data assimilation experiments were conducted — one with the finer resolution QuikSAT wind (50 km) data, and the other without the data — for a total period of about 60 days, starting January 8, 2004, and ending March 8, 2003. The major findings of these two parallel experiments based on the average of the 60 cases of forecasts reveal a significant positive impact on heights and winds at all levels for both the northern and southern hemispheres, especially for winds over the tropical oceans (Yu, 2003). Figure 5 shows anomaly correlations of zonal and meridional winds at 850 hPa over the tropics for forecasts with the use of QuikSCAT winds and those without the use of QuikSCAT winds. Note that the most significant improvement occurs in Day 4 and Day 5 forecasts and over the mesoscale features (of waves from 10 to 20).

Radiometer wind speed data from SSM/I measurements of the US Defense Meteorological

Satellite Programs (DMSP) since the 1980's constitute a continuing source of ocean surface wind data. The SSM/I wind speed data from DMSP has been used in the NCEP operational global data assimilation system since March 1993. The justification for implementing the SSM/I wind speed data was based on the results from a number of impact studies showing that the assimilation of wind speed data was slightly beneficial to the NCEP numerical analysis and short-range forecasts (Yu and Deaven, 1991; Yu et al., 1992). However, these operational SSM/I wind speeds were derived using an algorithm developed by Goodberlet et al. (1989), which assumed a linear dependence of wind speed on brightness temperatures. This assumption is acceptable when the level of moisture, both water vapor and liquid water, in the atmosphere is very low. As soon as the level of moisture increases, the dependence of the wind speed on brightness temperatures becomes significantly nonlinear, and errors in wind speeds retrieved by the Goodberlet algorithm become very large. For this reason, during the early days of operations at NCEP, only SSM/I wind speed data over the clear-sky area were used, and a large number of data points over active weather regions had to be rejected. This was rather unfortunate, since it is these developing weather systems that are most interesting and important in weather forecasting.

Tropical 850 mb mind (u) (waves 10-20 AC)

Tropical 850 mb mind (u) (waves 1-20 AC)

0 0

Post a comment