You are here

Warning message

Attention! This event has already passed.

Contribution to severe weather and multimodel ensemble forecasting in Belgium

Monday, 6 November, 2006 - 17:00
Campus: Brussels Humanities, Sciences & Engineering campus
auditorium P. Janssens
David Dehenauw

The vision of the Royal Meteorological Institute of Belgium (RMI) is to be a reliable
service for Belgium in the fields of meteorology, climatology and geophysics, for public
and government, based on continuity, innovation and research. To realise this vision, the
RMI-mission is to deliver a permanent service to provide information for and increase the
security of the Belgian population, including the socio-economic and scientific
community, by qualified staff and in national and international coordination.

The main task of a national weather service is to warn the public and the government of
severe weather ahead. This permits to take the necessary precautions. Generally, it is
anticipated, that as public awareness of the effect of weather on their everyday lives and
businesses increases, then the need for improved severe weather prediction becomes

This work aims to make a contribution to a better daily forecasting process in certain
severe and non-severe weather circumstances. Severe weather events are here defined as
damaging or at least disrupting weather patterns and do not occur on a daily basis, so
there is also a need to study non-severe weather, if higher quality forecasts are desired
every day.

The severe weather forecast contribution focuses on the prediction of large hail and the
risk of severe convective windgusts in Belgium, as tornadoes and downbursts. These two
phenomena are very hard to estimate but have a usually devastating impact on society. In
the literature, some forecasting techniques are mentioned, but these are rather old and/or
US-based. However, the climatology of dangerous weather of the US and Belgium differs
significantly. In this work, a different approach was taken and ten cases of Belgian
tornadoes or downbursts were studied and simulated with an operational weather
computer model. A tornado/downburst index was developed to warn forecasters of
potential wind damage during thunderstorms. Also, the synoptic and mesoscale features
of tornadoes or downbursts were looked into.

The hail forecast algorithms are meant to estimate the probable and maximum hailstone
size during thunderstorms.. These are based on a weather model simulating updrafts and
convective precipitation.

The non-severe weather forecast contribution tries to improve the performance and
increase the reliability of the current short-range numerical weather models, used in the
weather room of the Royal Meteorological Institute of Belgium (RMI). Techniques were
developed, based on multi-model ensemble modelling and Kalman filtering, to reduce
forecast errors.

The developed forecast methods will be available for RMI-meteorologists, on a daily
basis, as a guidance tool for making weather predictions.