Example of 12-hour high-resolution forecast
created at 15-12-1999.
During the night of December 13 to 14 1999, a small, developing low moved over the north of the Netherlands to the northeast. A small scale trough caused strong winds in a small area at the rear of this low. The maximum 10-minute average wind speed was 47 knots in IJmuiden. The regular, 55 km, HiRLAM did a good job of forecasting the small scale low. However, the accessory trough was not forecasted by the 55 km HiRLAM. Two experimental versions of the XHiRLAM (11 km grid) did forecast the small scale trough and the associated wind.
Below we show the 12-hour forecast of another experimental XHiRLAM. In this forecast the small scale low deepened too much and the trough did not materialize. The difference between this run and the other experimental high resolution runs are a few artificially generated vertical profiles that were assimilated extra in the two other runs. This shows the effect that a profiler network can have on the high resolution forecast.
This example nevertheless gives a good indication of the features that can be resolved by XHiRLAM. This animation starts at 13-12-1999 18UTC and ends at 14-12-1999 06UTC.
Blue lines: isobars (interval 1 hPa).
Black lines: temperature (interval 1 degr C).
Green areas: 1-hour precipitation sum.
Red: windflags.