|
MAIN CONCLUSIONS OF THE SO-CALLED
“AAA MEETING”, PRAGUE, 13/2/2004
(document prepared by
Jean-François Geleyn and sent to ALADIN directors and correspondants on February 25)
The proposal for
clarification of the terminology was accepted, with the replacement
of the name ‘TOOLBOX’ by the name ‘INTERFACES’
for one of the sub-projects.
It was decided to
disconnect for the time being the launching process of ALADIN-2 from
the HIRLAM problematic.
The need for the
ALARO 'intermediate step' between two transitions (on the way from
ALADIN to AROME) was confirmed by the nine Partners attending the
meeting (representing a majority of the ALADIN workforce). The first
transition will be in the source code (ALADIN => ALARO) and it
shall be harmonized within Partners as much as possible, also
because of the associated change in file structures and of the
externalisation of ISBA. In its standard version it should also
imply some additional costs, for the parts where one will know that
the Meso-NH physics is worth adapting to any scale (at least
microphysics and turbulence, quite likely). The second transition
will be in the cost (CPU and memory) of going to the scale of
resolved convection and of 3D turbulence and will be on individual
choices of each Partner, depending on its mastering of the new tools
and on its computing capacity. The separation between these two
transitions shall make the whole exercise as smooth as possible in a
still coordinated ensemble. Indeed some evolution of the NWP
applications during the ALARO intermediate step will be encouraged,
especially if some progress happens in the treatment of the 'grey
zone' problem.
The question of
"critical mass of manpower" was raised, with the comment
that so many subprojects might cause an exaggerated spread of the
current human resources, something that may even slow down the
ALADIN-2 progress. ALADIN Partners were again made aware that the
chosen "intermediate step" between two transitions will
imply some additional effort, but they judged it still appropriate
to their situation with respect to AROME and ALADIN.
The respective
role of operational codes, prototypes and Meso-NH software in the
R&D of most ALADIN-2 actions was clarified, even if some
uncertainties can only be treated later.
There exists now a
rather detailed ALADIN-2 workplan for 2004 that should:
be the basis for
first urgent actions;
help to a
reformulation of the strategic document, this leading, with the
help of Slovenia, to a political translation of the now chosen
orientations (effort targeted for the next Assembly in Split);
allow the quick
mobilisation of especially dedicated manpower transversal resources
fostered by Meteo-France and LACE, in order to reach the missing
critical mass of coordination and supervision work, in particular
on ALARO.
The 'tool-box' and
'convergence' basic principles still exist to structure the ALADIN-2
actions, but they should be taken more pragmatically. In short,
their concretisation should be sought only when really necessary
(physics-dynamics interface and coupling file creation vs. use, for
example).
Concerning the
question of a potential internal competition between solutions
either coming from differing origins or meant for different
purposes, only three areas in the physics of ALARO were identified
as really requiring the security of separate efforts (deep
convection, macro vs. micro physical aspects in the 'grey-zone' and
shallow convection), this leading to a small but unavoidable
overhead. Elsewhere one should rather aim at specific declinations
(e.g. for efficiency and stability at long time steps) of the same
'AROME inspired' software. Dynamics and data assimilation are not
supposed to require the same level of attention on this particular
point.
Priorities within
the preliminary synthetic planning document were examined, work
force was redistributed whenever necessary (except that the problem
of lateral coupling is obviously undermanned, at least for the time
being) and the harmonisation with the LACE plan required the
suppression of two items of the latter (plus the consequences of the
above-mentioned transversal adaptations). The individual cases of
people having to change topic as soon as feasible were also
scrutinised whenever identified. Dominique Giard will produce an
up-to-date version of the 2004 plan for topics put in ‘priority
one’, the other (non-discussed) items being left in Appendix.
Concerning the
preparation of the Split Assembly (29-30/10/04):
one aims at a one
and a half day meeting;
steps should be
soon taken to ensure the best possible level of attendance, given
the importance of the event;
it is assumed
that the work plan coming out of the present Prague meeting helps
extending the ALADIN current official scientific plan until the end
of the second ALADIN-MoU (hence one can concentrate at the Assembly
on the other issues, also by strongly diminishing the time devoted
to non-urgent standard agenda items);
on top of the
definitive anchoring of ALADIN-2 (see above), the follow-on process
of the ALADIN MoU should be central to the discussions at this
Assembly; given the link between this and the ALADIN-2 planning and
supervision, it is expected that the outcome of the AAA-meeting
will allow compiling a list of important modifications necessary to
the MoU. After discussion by Directors, this could be converted in
Split into draft proposals for the official replacement of the MoU
(targeted to the Bratislava Assembly of Fall 2005).
There is a need
for more gatherings of the type of community assembled at the
occasion of this ‘AAA’ ad-hoc meeting. This should
mainly happen at the occasion of other events (ALADIN Workshop,
EWGLAM-SRNWP, Assembly). Additionally each big item of the working
plan should have a coordinator and/or contact points. All this
should progressively lead to a new scientific steering structure for
the new project.
One ought to
extend the scope of the next ALADIN workshop: discussion sessions
are required on a few well-targeted topics, and one could extend the
already identified main one to "developments and plans in
physics", in order to attract more scientists.
|