In spring 2002, the ALADIN cycles have experienced a tremendous jump in their numbering: cycle 15 was phased directly to 25T1, 10 ALADIN cycles created in only 2 months !

Of course, this gap is simply due to the new rule, which states that the ALADIN cycles will in future have the same number than their ARPEGE counterpart. From a technical point of view, the phasing went on quite smoothly, which was a nice change compared to the hectic phasings of last years (too large delays of ARPEGE / IFS cycles, or ODB ...). Yet, several new features have appeared in the code.

The new geometry package MAKDO has been introduced in the model (Martin Janousek, Bodo Ahrens). This creates some modified numerical values for the geometry parameters, when compared to the old routines. Thus, it would be suitable for every partner to test the use of coupling files produced by AL25T1 in Toulouse, in order to check that the geometry frame-description can be read properly by the local (older) cycles. Thanks to the important work by Martin Janousek, the new geometry can be written in the usual file-frame format, so that old cycles can access and understand the data. The Toulouse system support should come back later to this point, via e-mail coordination.

A scientific change due to MAKDO is that the domains now are defined by their centre point and X and Y half-lengths, instead of corner point positions. For local adaptations, the namelists of ee927 and Fullpos will have to be changed. An automatic tool to highlight these changes has been prepared by Martin Janousek.

Some documentation is available and can be retrieved from Martin Janousek himself, Jean-Daniel Gril or Claude Fischer. It concerns changes in the code and the use of the namelist translator.

The revolutionary non-hydrostatic predictor/corrector scheme has been informatically introduced in the code. It was not yet tested. This development is the direct result of Jozef Vivoda's PhD work and parallel developments by Deborah Salmond in IFS.

Small changes were brought to the variational code (merge of CAIN and ECAIN routines, Claude Fischer).

A modified version of the test routines "aatestprog " for spectral transform was designed, to avoid the reading of a GRIB file, and replace this by the reading of a standard FA file (Siham Sbii).

Other changes were mostly adaptations from cycle CY25 of ARPEGE / IFS (Andrey Bogatchev, Marek Jerczynski, Zahra Sahlaoui, Gergö Boloni, the Toulouse team and Gabor Radnoti in Budapest for the preliminary phasing of the spectral transform package).