EUMETNET
C-SRNWP Programme
Annual Report 2002
Workshops
The workshop activity has followed the rule that each Lead Centre has to
organise a workshop every two years. With one exception, all the workshops of
the even years have take place, the exception being the UKMO which, because of
the move of its headquarters from Bracknell to Exeter, has postponed its 2002
workshop on Variational Methods.
Workshop on LAM-EPS
At the initiative of the INM (Spain) and in cooperation with it, the
first workshop in Europe on EPS (Ensemble Prediction Systems) for LAM (Limited
Area Models) has been organised in Madrid the 3-4 October. The motivation was
the following: the EPS technique is successful for the medium-range and there
is no scientific reasons hindering the use of this technique for the
short-range. That the technique will be more delicate to apply at short range
due to a smaller spread between the members is to be expected.
Workshop on Numerical Techniques
Following the decision taken at the SRNWP Annual Meeting 2001, the
Numerical Technique Workshop moved from the odd to the even years and took
place at Meteo-France Headquarters the 12-13 December. This workshop was again
very interesting and experiences amplification: numerical problems more
specific to the climate GCM were also discussed, as the conservation of mass in
the lagrangian advection. This problem is not of primary importance for the
short-range, but is a preoccupation for the climate simulations. Concerning our
high resolution LAM's, a problem is emerging that will become very important
with the steadily increase of the horizontal resolution: to have numeric
schemes that can cope with very steep slopes.
Workshop on the August floods in Central Europe
With the full support of KNMI, the Coordinator organized a one day
workshop (10 October) on the August floods that caused large-scale damages in
Austria, in the Czech Republic and in Germany. That workshop took place
immediately after the SRNWP Annual Meeting. The purpose of this workshop was to
see how our models have coped with the huge amount of precipitations measured
on the 12th and 13th of August. The workshop has been very successful in the sense
that not only the affected countries or their neighbours presented their
precipitation forecasts, but also more distant countries like UKMO, Spain or Ireland.
The presentations were followed by a general discussion where the impression
given by the presentations has been confirmed: it was only for 24h in advance
that our LAM's gave the right amount of precipitations at the right place. It
is not known why most models put the precipitations systematically too far to
the East. Many participants shared the opinion that the source of this inaccuracy
must lie in the boundary conditions yielded by the global models. It has also
been said that the usefulness of the ECMWF EPS has been very limited.
EU projects
The Coordinator has been deeply involved in two projects:
ALATNET
The Coordinator has been asked by the Commission to act as expert
advisor for the mid-term review of the ALATNET Project (22-23 April, Brussels).
This Project which permits to 12 young scientists to be trained as PhD
candidates or as Post-Doc's in numerical and dynamical meteorology using the
structure and the support of the ALADIN/LACE Consortia is a very good promotion
for the NWP science in Europe. The Coordinator has been impressed by the large
scientific scope covered in this project and by the quality of the works
presented.
Framework Programme 6
In May, the Coordinator has proposed to the SRNWP Members to submit an
EoI (Expression of Interest). The answer has been very positive. With the help
of colleagues of 4 NWS, he has drawn up and submitted a proposal for a
"Network of Excellence", this network being the NWS of our 5
Consortia. The title of the proposal was: "Improvement of Weather
Forecasting and Risk Assessment in Europe by Very High Resolution Simulations,
Ensemble Forecasting and Modelling System Developments".
Annual Meeting
It took place in the Headquarters of the Dutch NWS in De Bilt (9-10 October).
23 European NWS were represented (which is remarkable) plus the ECMWF. The main
points of discussion have been:
EUMETSAT ATOVS Retransmission Service
Several NWS have been pleased to hear that with the new station Athens
the South and South-East of Europe will also be covered by this Retransmission
Service. The Coordinator has expressed to the EUMETNET Deputy Director his
thanks for the fulfilment of his request (cf. Annual Report 2001).
Hourly SYNOP's
With the continuous data assimilation schemes (as 4D-Var or nudging) and
for the validation of the model diurnal cycle, it becomes each year more
important to have SYNOP's on an hourly basis. The Coordinator showed, as
example, the SYNOP chart of the 25 Jan. 2002 at 11h00 UTC. The discrepancy is
maximum: either is a country covered with SYNIOP (as for example Portugal,
France, Ireland, United Kingdom, Germany, Belgium, The Netherlands, Austria) or
the country is void of any observation (as, for example, Switzerland, Italy,
Spain, Denmark, Sweden, Norway, Finland, Island, Greece). The Coordinator will
try in 2003 to find out the reason of this dichotomy.
LC for LAM-EPS
After the success of the Madrid Workshop (see above under Workshops),
the Assembly has decided to create a new Lead Centre for LAM-EPS. SRNWP Members
have had the possibility to apply till the end of the year to become this new
Lead Centre. Four NWS have applied.
Hourly frames from ECMWF
Switzerland reiterated his wish to receive from ECMWF the frames for
each forecast hour (+00, +01, +02, etc) instead of the 3-hourly frames (+00,
+03, +06, etc) as today. But no other country supported this requirement.
SRNWP Multi-Model Ensemble
All the EPS works done today in Europe and presented in Madrid last
October are single-model EPS. DWD volunteered to assess the feasibility of the
multi-model LAM-EPS. We have in Europe four models: HIRLAM, ALADIN, Local Model
and UKMO UM. These models will form the basis of the multimodel short-range
EPS. This work will be established in 2003. In order to avoid administrative difficulties,
these multi-model ensemble forecasts will be computed in a first step in
delayed mode, i.e. with a delay of one or two days. Surface pressure, wind at
10m and precipitations at +48 will be produced one a day.
FP6
At the time of the Annual Meeting, the situation was not very clear what
FP6 would be. Nevertheless it was thought that SRNWP should follow its
intention expressed in its EoI: to propose a Network of Excellence and not an
Integrated Project.
EUMETSAT SAF's
Some of the developments made in the Satellite Application Facilities
(SAF) centres are of great importance for the NWP. It was always accepted - as
far as NWP is concerned - that the benefit of these developments goes mainly to
the global models. It must today be realised that the benefits could be of the
same importance for the high-resolution LAM's that it is for the global models.
SAF on Land Surface Analysis: 8-10 July 2002. Lisbon
As the soil processes are of the utmost importance as forcing for the
boundary layer and more generally for the hydrological cycle (cloudiness
included), it has been judged important to attend this workshop also because
the LAM community was weakly represented next to the ECMWF.
The coordinator has brought successfully three amendments to the final recommendations
which should be positive for the short-range NWP community:
- Snow cover: to concentrate to the simple method (the so-called
spectral one) in order to have products as soon as possible and to have the
more complicated method (the directional one) as a second step.
- Resolution of the albedo maps: as albedo is a function of vegetation,
same high resolution for the vegetation as for albedo (and not a coarser one as
envisaged).
- In the priority list: to give to the snow cover a higher priority than
to the land surface temperature.
For the report:
J. Quiby
C-SRNWP Programme Manager