High Resolution Numerical Weather Prediction Project
Website of the ALADIN Consortium
First stage abstract
Article published on 19 January 2005
last modification on 15 September 2008

by Patricia

Stormnet abstract
as submitted in the first stage proposal submission (form A1)

Local short-range numerical weather prediction (NWP) in Europe is based on high-resolution limited-area models and relies on many, usually small, national teams who joined their efforts within four consortia.

Since, on one side there is a continuous need for training early-stage researchers on NWP-related issues, and, on the other side expertise is widely spread among teams, a common and cross education effort appears necessary.

The few past initiatives at the level of groups were fruitful but are not enough. Organisation of training should now be thought at a higher level, so as to encompass more schools of thought and more scientific or technical disciplines.

This should also tighten links and favour exchanges between teams and groups. An increased training effort is also all the more important since new challenges are emerging.

All teams have now to face simultaneously a quick march towards further higher resolution for limited-area deterministic forecast models (involving significant changes in the conception of models and new scientific problems), a parallel improvement of data assimilation systems (as concerns the sophistication of algorithms and the density of observations to be used), an increased effort on the problem of forecasts reliability (both on old issues, such as verification, and on new ones, such as the short-range predictability of local extreme events) and their interaction with other applications.

Besides attention must be paid to the design of softwares, so that increased complexity keeps compatible with efficiency and portability (i.e. with an operational application), and with the management of input (meteorological observations) and output data, the volume of which will grow very quickly.