26th EWGLAM and 11th SRNWP Meetings

4-7 October 2004, Oslo (Norway), Local organiser: met.no

Participants

Marco Arpagaus, Richard Forbes, Lars Meuller, Niko Sokka, Nicole Asencio, Dominique Giard, Tiziana Paccagnella, Lora Taseva, Doina Teodora Banciu, Stefan Gollvik, Neva Pristov, Martina Tudor, Jean-Marie Bettems, James Hamilton, Jean Quiby, Per Unden, Dag Bjørge, Andras Horanyi, Evelyne Richard, Filip Vana, Michael Bush, Stjepan Ivatek-Sahdan, Kai Sattler, Josette Vanderborght, Massimo Bonavita, Marit Helene Jensen, Jan-Peter Schulz, Hans Volkert, Gerard Cats, Marek Jercynski, Yann Seity, Young Wang, Terry Davies, Dijana Klaric, Jürgen Steppeler, Mark Zagar, Maria Derkova, Martin Leutbecher, Niko Sokka & Marek Zijerczy.

1. Agenda with the national and the scientific presentations
2. Report of the SRNWP Business Meeting
3. Report of the EWGLAM Final Discussion
4. List of the operational NWP models (Oct 2004)
5. In Memoriam: Guenther DOMS (powerpoint)

1. Agenda

2. Report of the SRNWP Business Meeting

Jean Quiby, NWP Programme Manager

Stand of the Consortia

At the last COSMO Annual Meeting (22-24 September 2004), the Coordinator has been informed by the Chairman of the COSMO Steering Committee (the Steering Committee is the highest authority of the COSMO Consortium) that the National Meteorological Service of Romania has been accepted as 'Associate Member' of COSMO. The Representative of Romania at the COSMO Meeting has told the Coordinator that the relation with the ALADIN Consortium remains unchanged. The Coordinator has made it clear that the group of scientists working with the Local Model must be clearly separated from the group working with the ALADIN model. Moreover, software exchange between the two models have to follow the Sofware Exchange Rules of the SRNWP Programme.

State of the web site of the SRNWP Programme (srnwp.cscs.ch)

It has been possible these last months to improve the state of the web site. The largest part of the pages are now up-to-date and new pages have been created as, for example,  the pages 'Events 2004' and 'Events 2005' which list most of the European workshops, conferences and symposia related to Numerical Weather Prediction. The site, which is hosted by the Swiss National Computing Centre has reached a size of 580 Mbytes primarily due to the copies of the presentations made at the SRNWP Annual Meetings and Workshops.

The site is also equipped with a counter: there are about 240 visits a month, included the visits for maintenance.

Dissemination of hourly SYNOPs in Europe

The drive to higher model resolutions characterizing the development of the numerical weather prediction requires data at a high spatial and temporal resolution for the determination of the initial conditions as well as for the verification of the model results. In the last 12 months, a few new NWS have joint the action. In the next 12 months, the Coordinator shall invite the NWS of Central and South East Europe to participate.

We have noticed that some NWS produce more SYNOP than the ones that are disseminated. The reason for the non-dissemination of these other SYNOP is due to the fact that they provide from stations without an official  WMO identifier.

In the frame of  FP6, SRNWP Proposal in the instrument 'Human Resources and Mobility'

After revision by the EC of the instrument 'Human Resources and Mobility', calls have been published again the 11th of September 2004. The Assembly has decided to submit a proposal in the SRNWP frame for the call 'FP6-2004-Mobility-1' with closure date on the 2nd of December 2004. Already 8-10 NWS - mainly from the ALADIN Consortium - have expressed an interest to participate. Planned is a 'Research and Training Network' for 'NWP at very high resolution'. All the aspects of this technique, inclusive EPS and code optimization, will be considered.

State of the SRNWP PEPS Project

13 NWS are participating in the project and two supplementary commitments have been made during the meeting (Meteo-France and UGM).  The first attempt - based on a very simple scheme - shows already results: the probabilities for respective >5mm, >20 mm and >50 mm of precipitation in 24-hours. Next to these EPS products, the mean of the different forecasts is also computed. Method and results will be presented at the next SRNWP Short Range EPS Workshop (7-8 April 2004 in Bologna).

Recommandations of the last SRNWP Workshop on Meso-scale Verification

(Second SRNWP Workshop on Mesoscale Verification, KNMI, 14 - 15 June 2004).

This Workshop made several important recommandations, all related to the observations.

This proves again that the trend to the NWP high resolution puts high requirements on observations.

The following recommandations have been made:

• GPS: Dissemination of the ZTD's coordinated by EUCOS
• Collecting non-GTS data of precipitations:
  KNMI has shown interest to become active in this important matter.
        Reporting of the snow depth in the SYNOP
    The Co-ordinator will liaise with the Chairman of the PB-Obs about that matter
• Dissemination of the hourly SYNOP: To gain other countries
• Velocity Azimuth Display (VAD) Radar Winds: Ensure an improvement of the dissemination in the frame of OPERA 2
• Radial winds: Dissemination of the radar radial winds should also be considered by the OPERA 'Programme

These recommandations will define a large part of the future work of the Coordinator.

Model inter-comparison

In the NWP Programme Decision, which is the document elaborated by the EUMETNET Council and defining the tasks of this Programme, one of the stipulated tasks is To take initiative for model inter-comparisons.

How this task could be practically realised has already been thoroughly discussed at the last SRNWP Workshop on Meso-scale Verification (14-15 June 2004, KNMI, De Bilt). Two working groups had to make propositions and they arrived at very different conclusions. But both groups agreed that it will not be an easy task. Whatever the strategy will be, it will always be possible to find weaknesses, even to claim that the chosen strategy was unfavourable for a given model.

Nevertheless we should embark in the course of the next 12 months on this endeavour.

A necessary condition to arrive at a result is that the strategy must be simple. It would be hopeless to try to compare between them all the models in operation in Europe.

In essence, there are only 4 models in the SRNWP Network:ALADIN, HIRLAM, Local Model and the Meso-scale UM. Thus the comparison should be limited to only one version for each of these four models and the version to retain should be the one with the largest integration domain in order to have a common overlapping area.

The LACE Programme leader insisted to also have in the comparison a model of one of the LACE partners. The Coordinator accepted this request. It remains to be seen whether this will be technically possible.

3. EWGLAM Final Discussion

minutes: Jean Quiby, SRNWP Coordinator, Wednesday 6th of October 2004

Proceedings (or documentation) of the EWGLAM/SRNWP Annual Meetings

At last year's meeting (Lisbon, 2003), it was decided to collect the scientific contributions (articles, not the presentations) and to make an electronic Newsletter only (no production of a Newsletter book as in the previous years). Unfortunately, IM Portugal has been unable to make this electronic Newsletter. Instead, all the presentations (Consortium reports, National posters and scientific presentations) have been collected by the SRNWP Coordinator and can be viewed in the SRNWP web site).

For this year's EWGLAM/SRNWP Meeting, met.no will produce an electronic Newsletter with the reports of the Consortia, the national reports and the scientific contributions ('papers', not the presentations). The national reports are limited to 4 pages. No limitation has been defined for the other types of reports and for the scientific contributions. The contributions must be sent to met.no till November the 30th. And the SRNWP Coordinator will collect all the electronic presentations and display them in the SRNWP web site in the same way as this has been done for the 2003 Annual Meeting.

Format of the EWGLAM/SRNWP Annual Meetings

It has been decided that future meetings will last 3 full days only or 3 and a half if there is an excursion. The local organizer decides solely whether it will organize an excursion.

Special topic and special session for the 2005 Meeting

After a lengthy discussion, it has been decided that the Special Topic for next year's meeting will be 'Very high resolution'. Presentations are expected on all aspects of the very high resolution: data assimilation, numerics, physics, etc. There will be no Special Session, but the Local Organizer may still propose in the invitation a topic of particular interest. A possible such topic could be the interpretation of model's output from the modeler point of view and the assessment of the adequacy of presenting the results of the very high resolution in the standard model output.

An important discussion took place following a wish that our Annual Meetings should also consider the needs of the NWP users and that contacts with them should be envisaged. The majority of the opinions expressed were that our EWGLAM/SRNWP Meetings should concentrate on the models only (inclusive data assimilation) and that the link with the users should be dealt with by the Lead Centre for Statistical and Dynamical Interpretation. The Coordinator will discuss with this LC a possible extension its scope towards the needs of the NWP users.

Date and place of the 2005 EWGLAM/SRNWP Meeting

This meeting will take place in Slovenia from Monday the 3rd to Wednesday the 5th of October (or to Thursday 12h; see Point 2 above) and will be organized by the Environmental Agency.

Place of the 2006 and 2007 Meetings

The Swiss delegate invited the 2006 Meeting in Switzerland (still to be approved by the Director of MeteoSwiss).

The Croatian delegate invited the 2007 Meeting in Croatia (still to be approved by the Director of the Meteorological and Hydrological Service).

4. Numerical models in Europe