ALADIN and ALATNET News

1. A rainy summer

Weather was not very nice last summer, especially over Central and Eastern Europe which had to face very severe floods. ALADIN forecasts proved quite good in these circumstances. More details can be found in the dedicated paper from the Austrian team and in the LACE report.

A dedicated SRNWP workshop was organized in De Bilt (Nl) on October 10th, after the annual joint EWGLAM/SRNWP meeting. Presentations and reports should be available soon on the SRNWP web site : http://srnwp.cscs.ch/ .

2. LACE training course "on physics"

A training course on physical parameterisations was organized in Prague, from the 16th of September to the 4th of October. This was a pure LACE action, aiming at training a team likely to start research in physics within the new LACE structure. Students were few : Alexander Kann, Harald Seidl (Au), Martina Tudor (Cr), Martin Janousek, Richard Mladek, Petra Smolikova, Alena Trojakova, Filip Vana (Cz), Laszlo Kullmann (Hu), Martin Bellus (Sk), Kay Suselj (Si) (and not all expected to focus on physics afterwards, as one may guess).

The program was quite ambitious and training very intensive : general lessons about the NWP physics and special ones about the parameterisations used in ARPEGE/ALADIN, comparison between code and documentation, study and report on papers, work in small groups. For more details, contact Filip.Vana++at++chmi.cz or the teacher : Jean-Francois.Geleyn++at++meteo.fr .

3. "ALADIN maintenance and phasing training" workshop

Compared to the previous one, this training course was short, only one week (25-29 November) long, and more widely open (5 non-LACE students among 18). Considering the a-priori daunting topics concerned and the standing difficulty to find volunteers for phasing, it was a success : only 10 students were expected at first. The program was quite dense, with lectures, exercises and discussions. Details on the organization and presentations may be found on the dedicated web site :

This workshop also gave an opportunity to renew or write documentation on ALADIN. The final discussions are summarized hereafter, thanks to András Horányi.

4. EWGLAM and SRNWP events

a. Annual 2002 EWGLAM and SRNWP meetings + SRNWP workshop on floods

De Bilt (Nl), 7-10 October 2002

The corresponding LAM Newsletter and some pictures are already available at : http://www.knmi.nl/hirlam/EWGLAM2002/ , thanks to the efficiency of Gerard Cats.

b. HIRLAM workshop on Mesoscale Modelling

Dublin (Ir), 14-16 October 2002

Dijana Klaric was the ALADIN representative at this workshop. For more details, contact her or have a look at the HIRLAM web site. The proceedings are also available in Toulouse.

c. SRNWP workshop on Numerical Technics

Toulouse (Fr), 12-13 December 2002

The corresponding report should be available soon on the SRNWP web site (updated recently) : http://srnwp.cscs.ch/

d. SRNWP workshop on Statistical and Dynamical Adaptation

Vienna (Au), 5-7 May 2003

"The scope of these meetings has been enlarged : not only the statistical adaptation will be the topic of the meeting, but also the dynamical adaptation. This includes presentations on, as examples : 1d models, road condition models, trajectories, orographic precipitation models, down-scaling methods, etc ... " (extracted from the official announcement). More details on the dedicated site :

e. Annual 2003 EWGLAM and SRNWP meetings

Lisbon (Pt), 6-9 October 2003

The scientific topic will be : Use of new data for short-range NWP.

f. SRNWP workshop on High Resolution Modelling

Bad Orb (De), 27-29 October 2003

with special emphasis on physical parameterizations.

g. SRNWP workshop on Verification Methods

De Bilt (NL), autumn 2003 ?

focussing on precipitations.

h. A new SRNWP Lead Centre

HMS, as a member of the ALADIN group, volunteered to be the new Lead Centre for LAM EPS, but there are 3 candidacies ...

i. SRNWP report for 2002

The annual SRNWP report, written by Jean Quiby, may be found hereafter. It should be available soon on the SRNWP web site.

5. ALADIN cooperation

a. 7th Assembly of Partners

The 7th assembly of ALADIN partners was held in Bucharest, Romania, on the 28th of October, 2002. It welcomed participants from 14 ALADIN countries and an ECMWF observer.

The minutes should be soon available, but some points may already be mentioned :

- agreement concerning the use of the ALADIN model in climate mode for scientific purposes (climate simulations), by the International Arctic Research Center (IARC) of the University of Alaska in Fairbanks, Alaska (USA);

- agreement on the cooperation with African countries, with Morocco as main intermediate (see hereafter);

- problems with ODB (lack of documentation and support) underlined once more;

- idem for the subcritical size of some ALADIN teams, the need for more publications and documentation, etc;

- presentation by Jure Jerman of the "Prototype for common objective verification at synoptic scales". No agreement was found on the problems of financial and technical support to the Slovenian team, so it was suggested to continue the prototype development and present both the results and a detailed documentation including the foreseen economical efforts next time;

- presentation of some opportunities within the 6th Framework Program of the European Union;

- creation of a "rotating foreign bureau" hosted by the last organizer of the Assembly of Partners, starting with Romania. This committee will monitor the MoU modifications required by the cooperation of the ALADIN Consortium with ACMAD (see hereafter);

- no volunteer to host the next ALADIN workshop.

Next meetings : Cracow, end of October 2003; Croatia in 2004.

b. Cooperation with African countries

Following the conclusions of the working group's meeting in Ston-Dubrovnik (Croatia - 04.2002) and of the WMO meeting in Pretoria (RSA - 10.2002), and within the cooperation between ACMAD and Météo-France on one hand and between ACMAD and DMN Morocco on the second hand, to develop NWP in Africa, a new configuration of ALADIN, called ALADIN-NORAF, is running operationally in Casablanca since the 24th of February. The NWP products issued from this new model should be sent to the covered NMSs via RETIM 2000 or Internet. The final agreement is for the following list of African countries : Algeria, Benin, Burkina-Faso, Chad, Egypt, Eritrea, Gambia, Ghana, Guinea, Guinea-Bissau, Ivory Coast, Liberia, Libya, Mali, Mauritania, Niger, Nigeria, Senegal, Sierra-Leone, Sudan, Togo (Morocco and Tunisia being full ALADIN members).

The ALADIN-NORAF domain covers nearly all the African countries to the north of the equator (from 43.86°N to 0.70°S and from 36.20°W to 56.48°E). This model is run twice a day, at 00 and 12 h UTC, with respectively 72h and 60h forecasts. Data assimilation is realized every 6 hours, using optimal interpolation in a first step. By the end of this year Morocco plans to use raw radiances from TOVS AMSU-A with a 3d-var implementation. The coupling files are provided by Météo-France from the ARPEGE stretched global model. In parallel a local model ALADIN-ALBACHIR, covering Morocco, is run without data assimilation and coupled with the regional model ALADIN-NORAF.

A new users' group, the Committee of Users of the Model ALADIN-NORAF (CUMAN), has been created by ACMAD to define the best products which should be produced and distributed to the NMSs, and to discuss on the different topics linked to the exploitation of this new model. The first meeting of the CUMAN will be held in Casablanca by the end of April.

For more details contact Radi Ajjaji in Casablanca or Philippe Caille in Toulouse.

c. Scientific reports

Thanks to the thematic coordinators, a detailed scientific report covering the first 10 months of 2002 (prepared for the Assembly of Partners) is available on the ALADIN web site :

Many other documents are now available on these pages. A scientific plan for ALADIN in 2003 has not been prepared yet, because of the uncertainties resulting from the problematic relationships between ALADIN and AROME (see below).

d. New LACE organization

The organization of the LACE group (including Austria, Czech Republic, Croatia, Hungary, Slovakia and Slovenia) has been modified. From the beginning of 2003, the common operational LACE suite stops. Each partner is responsible for its own operational applications, only the coupling files computed from ARPEGE are the same for all domains. Hence several teams were very busy with the design of new operational domains along the last months. Cooperation remains in research (3 working groups), training (next topic : ODB) and maintenance actions.

The new team of coordinators is the following :

- project leader : Dijana Klaric

- dynamics and coupling : Petra Smolikova

- data assimilation : Gergö Bölöni

- physics : Thomas Haiden

- maintenance : Oldrich Spaniel

- data management : Stjepan Ivatek-Sahdan

e. Next ALADIN workshop ?

Though the problem was raised again and again, no partner volunteered to organize the 2003 ALADIN workshop till the end of February ! Since the Newsletter is late (as often), we are glad to announce that the Czech team offers to welcome the next workshop in Prague, during the last week of November. More details soon.

Austria candidated for 2004.

f. ALADIN and AROME

AROME is a Météo-France project aiming to design a NWP LAM of very high resolution (2-3 km) within 2010, "as far as possible in cooperation with ALADIN partners".

However, though the ALADIN non-hydrostatic dynamics (and the ALADIN source code) was chosen as the starting point for AROME, a future cooperation may really be questioned. A report, issued from CSSI discussions, was sent to Directors and a dedicated scientific+political workshop will be organized in Prague (11-12 April), just before the next WMO session.

6. ALATNET events

a. Timetable of stays

The ALATNET stays of Christopher Smith in Prague, Ilian Gospodinov in Bruxelles, Cornel Soci and Gianpaolo Balsamo in Toulouse are now finished. Cornel and Gianpaolo should defend their PhD thesis before the end of Spring 2003.

b. Workshop on coupling

A one-week (17-21 February 2003) workshop on coupling issues has been organized by the Slovenian team in Ljubljana. The program is available at the following address :

c. Reports

The next annual report must be sent before the end of April. A meeting of the steering committee is scheduled on the 24th of March in Ljubljana.

More informations are now available on the personal web pages of the ALATNET students.

d. Last workshop

A workshop gathering ALATNET students, their mentors and an additional member of each ALATNET team will be organized by the Hungarian team from the 15th to the 17th of October 2003. The SRNWP coordinators and representatives of the other consortia are also invited.

This meeting was scheduled after a demand from the EU representative, Mrs Jane Shield, at the mid-term review. Its aim is to allow students to present their results and discuss together once again. Part of the workshop will also be devoted to the preparation of the final review.

e. And afterwards ?

A proposal for a new RTN in the framework of the 6th FP of the EU (second call, November 2003) is to be considered. When compared to ALATNET, the network should focus more on training and involve more European countries. Along informal discussions during the EWGLAM meeting, COSMO expressed some interest in a joint venture.

7. Some practical informations

a. About 2003 stays

In Toulouse :

Many visitors are expected in Toulouse in 2003 !

Embassy fundings for Central and Eastern Europe asked for 2002 are now available, while in the meantime maintenance fundings slightly increased and the department of international affairs of Météo-France (D2I/INT) obtained extra fundings for ALADIN (both as a counterpart to the disappearance of Embassy fundings). Embassy fundings for Morocco and Tunisia will be negotiated soon, and are likely to keep the same level as last year, at least. And ALATNET stays are to be added.

However the situation is quite messy. The management of fundings depends on their origin, let us recall the rules :

All stays should start before the end of 2003, but EGIDE/Paris is not yet ready, and there are now few offices available at GMAP for visitors. If we don't obtain some more (the demand is already a few years old) we won't be able to welcome everyone. Small (very small) indices that discussions might start soon appeared along the last weeks.

The timetables are available on the ALADIN web site.

In Budapest :

A grant for a stay in Budapest is offered by HMS, as in 2002. Contact András Horányi (horanyi.a++at++met.hu) for further details.

LACE actions :

Fundings are available within LACE for research stays : 8 months (4 for dynamics, 3 for data assimilation, 1 for physics), in Budapest (1), Prague (6) and Vienna (1).

b. Documentation and publications

A significant effort was dedicated to documentation along the last months, thanks to the workshop on maintenance. The corresponding information is available on the (internal) GMAP web site and on the ALADIN pages, referring to the HMS web ones. This will be pursued along 2003, with some reports already scheduled : general ALADIN documentation (update by Jan Masek), 923, DFI,... PhDs and publications will also be considered (and in practice are restarting). Several maintenance stays in Toulouse have been scheduled for such a work. The next step should be an update of the ALADIN and ALATNET web sites, to take this into account. Any volunteer ?

c. New cycles and export versions

· Export version for cycle 25T1 (AL25T1_export_01)

A new export version, based on cycle 25T1, has been prepared recently. Apart from the reference version, whose main advances have been described by Claude Fischer in the previous Newsletter, it includes further debugging, especially in data assimilation configurations, and the latest changes for operations in Toulouse (in 2002). A short description of changes and the source code of convertors (for geometry and namelists) are available in the export package.

This cycle has been implemented operationally in Toulouse (17 December), and the change of geometry within coupling files was quite smooth. It has also been installed in Slovakia, for research on NH dynamics mainly, and in Morocco, for operations.

Let us briefly recall the contents of this version :

- a lot of pruning (removing unused options etc), cleaning, debugging (as usual), and more safety locks, especially in physics;

- new ALADIN geometry (and induced changes);

- implementation of a "unified" predictor-corrector time-stepping scheme;

- introduction of the source code for blending;

- update of data assimilation configurations, new management of satellite data, small changes in ODB (but with no ascending compatibility), improved use of land-sea mask in some "surface" observation operators;

- use of local climatological ozone profiles in radiation (on demand), changes in the anti-fibrillation scheme;

- update of options LVERCOR (relaxation of thin layer hypothesis) and LVERTFE (finite elements for vertical discretisation) in dynamics;

- refinements for the vertical-plane model, with still some problems;

- and pure ARPEGE modifications.

· Cycle 25T2

A new cycle was created last autumn, including:

- new developments in NH dynamics : new prognostic variable d4, predictor-corrector scheme, new sponge, advection of vertical velocity, corrections in the vertical-plane model ... (this was the main purpose of this intermediate cycle);

- tendency-coupling for surface pressure (only for 2TL schemes and standard DFI);

- 4d-screening in ALADIN;

- introduction of an "obs pre-processing" for ALADIN : geographical selection of observations, in C+I zone;

- improved handling of "mean wind" in variational applications;

- use of raw radiances in ALADIN;

- improvements in the selection of surface observations (case of isles and lakes, use of SYNOPs with missing height information);

- merge of the new snow scheme (described in the previous Newsletter), safer handling of liquid water and ice in physics;

- optimizations, cleaning, and pure ARPEGE modifications.

There is no corresponding export version.

· And

Modifications for configuration 923, based on the export versions AL12_export_04 and AL15_export_03, are available. They allow, first to run e923 on AL15, second to solve some standing problems and use new developments. A documentation has been sent to the e-mail list aladin++at++meteo.fr.

A new export version for AL15 should be prepared soon, including the bugfix for 923 and new developments in physics, at least.

A new export version for AL25T1 will also be prepared, for changes in physics.

For more details, contact Claude Fischer : claude.fischer++at++meteo.fr , or the GCO team (with a newcomer !) : gco++at++meteo.fr .

d. ODB

Complains about ODB (lack of support and documentation, very quick evolution,...) were many and strong along the last "official" meetings. We got an answer from ECMWF : they listen but cannot provide a documentation before Spring 2003. In practice the most recent news are : ( i) ECMWF asked for the documentation written by Dominique Puech (GMAP), to translate it; (ii) Databases must be rebuilt at each new cycle.

e. New e-mail addresses in Romania

The new ones read : forename.name++at++meteo.inmh.ro (e.g. doina.banciu++at++meteo.inmh.ro).

And, thanks to Mateja Irsic, informations on ALADIN teams are now up-to-date on the ALADIN web site.