Single column modelling and intercomparison of physical packages

The general goal of this intercomparison study is to improve the operational ARPEGE/ALADIN physics while validating it on observed datasets (GATE, TOGA/COARE, BOMEX, ARM, etc...) and making predictions of ARPEGE/ALADIN, ECMWF or HIRLAM physics in a common software environment to compare their results in an easier way.

Jozef firstly introduced the ECMWF physics into the common software environment, secondly made the interface between ECMWF physics and ARPEGE/ALADIN DDH software, producing the budget of all prognostic variables. He finally made 1D predictions on GATE and TOGA-COARE Case 1 datasets.

This leads to interesting results for both technical and scientific aspects:

• T1. Now it is possible to run ECMWF or ARPEGE/ALADIN physics inside 1D model, only changing one switch.
• T2. The interface to diagnostics is safe in the sense that they produce small residuals in the prognostic variables budgets.
• T3. Work was done on 1D model source code environment and GNU Makefiles.
• S1. The first results indicated that existing biases from 3D ARPEGE/ALADIN model can be reproduced on 1D observed datasets predictions: thus we can now address these biases in an easier and more understanding environment.
• S2. Intercomparison studies on GATE dataset showed that the convection scheme ACCVIMP needs to produce stronger water vapor turbulent fluxes.

This job will be continued in 1999 in Toulouse to change the convection scheme, deal with new datasets and work on radiation scheme.

for more details on the 1D version of ARPEGE/ALADIN, please contact : jean-marcel.piriou++at++meteo.fr or eric.bazile++at++meteo.fr

Home