June is Bustin' Out All Over Rogers and Hammerstein
The Summer is here and this June Newsletter is bustin' with interesting
articles. A glance at the `Table of Contents'
on the back cover will show the great range of current activities. I would
like to thank all of you who have taken the trouble to prepare and submit
articles and contributions. The vibrancy and vitality of the Newsletter
depends completely on your input. There are some (pleasant) surprises for
everyone in this issue.
The recent, very successful All-Staff Meeting in Oslo is reviewed on page 3. A special session of that meeting focussed on turbulence; this followed from an evaluation meeting at KNMI in March, reported on page 7. The reference first-order vertical diffusion scheme and a higher-order turbulence scheme are described in the next two articles. A note on the automatic scaling of the horizontal diffusion coefficient follows.
Starting on page 19, the System Manager reviews recent collaboration to improve the performance of the reference system at ECMWF; the optimum arrangement is to split execution across two computer platforms. Recent tests with the re-coded surface parameterisation scheme are described by Ernesto Rodríguez (p21) ant the improved performance of ARPEGE/ALADIN resulting from the ISBA scheme are reported by Eric Bazile and Dominique Giard (p26).
Viel Ødegaard describes a problem in the DNMI humidity analysis (p31). This problem is also relevant for the Hirlam analysis (NL26, p6). Viel also reports on progress with the implementation of the Rasch-Kristjánsson scheme. Following this, Bent Hansen Sass discusses the application of two cloud and condensation schemes in the context of a BOMEX case, illustrating some advantages of the straco scheme.
The errors associated with lateral boundaries are a nagging headache for limited area modellers. Work towards `moving the boundaries out of harm's way' by use of a stretched co-ordinate system is in progress (Enda Kelly and Jim Hamilton, p40). Another exciting research project, underway at Tartu Observatory, is the development of a version of Hirlam using the complete, non-hydrostatic equations in pressure co- ordinates; this is described by Rein Rõõm.
On page 46, Carl Fortelius discusses the use of budget diagnostics for detection of systematic model deficiencies and proposes a list of quantities to be computed and archived. Plans to use a digital filter as a gravity wave constraint in variational assimilation are outlined by Loïk Berre on page 48. Then Nils Gustafsson and Magnus Lindskog deliver some good news on preliminary tests of 3DVAR in Hirlam.
The operational configurations of Hirlam in seven member institutes are summarised starting on page 57. Comparisons with last year's reports (NL27) show significant enhancements, especially in terms of model resolution. Operational developments at Météo-France during the past year are presented on page 81.
The Newsletter ends with two notes on verification. The rms errors in sea-level pressure for the first trimester of this year are compared for several Hirlam implementations (page 88), with a discussion of problems and current work to resolve them. Finally, a report of some comparative runs using regular and rotated grids is presented. This arises from recent studies which produced inferior results in rotated geometry. It appears that the discrepancies can be explained in terms of proximity of the boundaries to the verification area, and that they are not due to any more fundamental problems or to coding errors.
Other news:
Forthcoming Meetings: