AMMA-MOD Presentation

The AMMA model database aims at:

  • providing to the whole AMMA research community products widely used to interpret model results as well as observation data (such as ECMWF analysis, climatologic fields…);
  • providing to the different modeling groups facilities of storing, comparing, distributing and discussing their results;
  • stimulating the exchange of information and collaboration between researchers from different disciplines or using different methods (typically modelers versus observers);
  • building a durable archive of pertinent model products for the AMMA scientific community, to be mirrored in Africa as early as 2008.

Many modelers are involved in the AMMA project and planned to simulate different features of the Western African monsoon (atmospheric and ocean processes, chemical and aerosol transport, surface-atmosphere interactions…) and their variability using a large variety of models. Many particular situations documented during the SOPs will also constitute case studies for meso-scale simulations. Some intercomparison projects have already been designed (ALMIP, AMMA-MIP …).

Model re-analyses (ECMWF and NCEP) are already available, and operational model routine analyses will be archived (atmosphere and ocean). The open question concerns all research model (atmospheric GCM and meso-scale models, atmospheric chemistry, land surface, hydrology and energy exchanges with atmosphere, ocean specific model simulations / re-analyses, etc…). The different modelers groups will be contacted to get them sensitized to the necessity of furnishing relevant products to the database.

The design and development of the AMMA model database has been achieved in a first time, to provide first operational model outputs, then specific research simulations in a second time.

The AMMA model database is managed and developed at IPSL. A first structure satisfies the three first purposes (designed hereafter as “the 2006 model database”). The last target (designed as “the 2008 model database”) will be achieved when the researchers will transfer key products to the database in due time and standard format.

The 2006 model database

In a first time, the 2006 actual database is composed of three different parts corresponding to different uses and facilities:

Common products:

Widely used products are or will be centralized in the database in order to help the AMMA community work.

They include the products asked by the AMMA researchers among:

  • a subset of the ECMWF analysis/forecast outputs. The products are archived on a regular longitude/latitude grid over the AMMA region. An inquiry mail has been sent to a wide diffusion list in order to determine the parameters, resolution and times to be stored ;
  • a subset of the ECMWF ERA40 reanalysis;
  • a subset of the NCEP analysis
  • a subset of the ocean MERCATOR analysis ;
  • a subset of the AMMA-ALADIN analysis ;
  • a subset of the Météo-France ARPEGE analysis ;

The parameters for the different models above are not defined precisly at this time, except for ECMWF products, widely used and asked by the different communities.

We remind that the AMMA database constitutes an interface between the AMMA researchers and the operational centers (ECMWF, Météo-France…), which are not supposed to provide data service to the AMMA community. The AMMA database offers a free data service for our community. The AMMA people requirements will be taken into account and satisfied as far as possible.

Public or private workspaces:

Some modeling groups expressed the need of a workspace allowing the storage and the exchange of data and information. For these groups, we propose to set up a storage space with upload and download facilities associated with a web site and a forum. Depending on the group requirements, the access to the workspace will be either public or reserved.
At present, three groups have been identified by the model database team, which are the AMMA Land surface Model Intercomparison Project (ALMIP) group, the AMMA-MIP modeling group and the CLIMATE modeling group.

Links to other AMMA initiatives:

Some groups have already organized their own workspace, archive and exchange ways. If agreed by the groups, it will be possible to display a presentation page of the project and a link to the webpage of each group from the AMMA model database website.

Up to now, three modeling groups already organized were identified: the MESO-NH group, the AMMA-MIP project and the WAMME project.

The 2008 model database

The durable AMMA model database will be set up at IPSL and mirrored in Africa. It will include two different types of model outputs:

  • the “Common products” part, directly inherited from the work AMMA model database.
  • the research model outputs. It will contain the simulations or other products which will be evaluated as pertinent by the different modeling groups, such as:
  1. forcing fields
  2. reference simulations
  3. case studies
  4. long coupled climate simulation

Each group has to communicate results to the durable model database. In order to plan and ease the set up of the model database, each group has:

  • to propose a timeline indicating when they may be able to provide products.
  • to send an example of the model outputs in order to check that the format is manageable and the required metadata indicated.

Last Updated ( Tuesday, 19 December 2006 )