Figure 4-1. Structure of the MAP Data Centre.
All data gathered in or used by MAP will be stored and made available by the MAP Data Centre (MDC). The MDC is responsible to the MAP Programme Office, and is physically located at ETH Zurich. Access is through the Internet at address
Build-up of the MDC has started in 1995. It is funded by financial contributions of the European Weather Services through an official EUMETNET project, that grants funding until the end of 1999. In the meantime, the MDC has become operational, both on a technical and administrative level. Registration to users is open since 1 August 1996. At the time of this writing, a wide range of data is already available. Further data will be included as it becomes available from the numerous data providers.
Access to data stored at the MDC is regulated by the MAP Data Access Policy (see Schär and Binder 1996). The Intergovernmental Panel of MAP and all major data providers have approved the MAP Data Access Policy on 22 May 1996. Access to the MDC is open to all scientists upon registration, regardless of whether they are project participants or not. The registration procedure requires to sign a brief statement that specifies the conditions under which the data may be accessed and used. This simple procedure applies for all the field phase and case study data. For a few other data categories (such as long-term climatological series), additional conditions are implied by the providing agencies.
The tasks of the MDC are primarily to
Existing data sets are required for climatological studies and analyses. Operational data relates to GTS-type data whose collection should start at a very early stage of the programme. Special data refers to data either measured or produced in special pre-field experiment activities in so-called warm-up activities. Supporting data includes physiographical data, digital maps, address lists, bibliographies, etc. As its name suggests, field phase data are data collected in the field experiment. Secondary data, finally, refers to results of simulations, model output, etc.
The location of and access to data which is relevant for MAP, but is already stored elsewhere (e.g. in the InfoClima system) is facilitated by including the respective data catalogues in the MDC.
An electronic bulletin board provides the latest information on the status of observing systems, experimental activities, and other news of interest to the MAP community.
The MDC is responsible for the compilation and continuous updating of a data catalogue, listing the information available from MDC, and providing instructions on how to retrieve data in a Data User's Manual.
It is the MDC's responsibility to ensure that poorly or only partly documented data sets are properly described. This must be done in co-operation with the data producer or, in simple cases, by the MDC alone.
Data arriving at the MDC must be formally quality controlled by
the MDC. Depending on the available manpower, the MDC might also
run quality control programmes. However, quality control is basically
in the responsibility of the data producer. Additional independent
projects related to data quality control are encouraged.
The MDC will have a modular structure: A Central MAP Data Bank (CMDB) to which all requests for data are made, is connected with a number of Special MAP Data Banks (SMDB). Each data bank has its own Data Bank Manager(s) who is (are) responsible for the tasks listed in Section 4.1 (of course, this applies only with respect to the data archived in the data bank in question).
Data are to be submitted to the data banks in form of sequential files. Archiving of the data is either file structured or for certain data categories in relational data bases. The transfer of the data submitted as sequential files into the relational data base is the responsibility of the data bank manager. While relational data bases are very user friendly, not all data types are suitable for archiving in this way. In addition, the transfer is quite demanding on manpower. Consequently, the extent to which relational data bases will be established, will depend on feasibility and available resources.
Access to the CMDB and later on to any SMDB is only possible through a Guard and Logging Facility. Here all data traffic is monitored, users may be identified, passwords may be requested, etc. Two main security problems exist:
The different data banks making up the MDC will be connected via Internet. Efficient and user friendly transfer protocols are used. However, it should be born in mind that on many Internet connections traffic is quite heavy. For the transfer of data files larger than, say, 10Mbytes, line capacity will most likely be a problem. Consequently, the exchange of large amounts of data (e.g. radar and satellite images), is realized by alternate means such as compact disks.
In order to allow data in BUFR and GRIB format to be included in the data banks, binary as well as text formats are supported. For a given data type (e.g. aircraft data, profiler data, surface data, etc.) as few different formats as possible are used, formats allowing for quality flags are preferred. Apart from transferring file-structured data to a relational data base, it is not the responsibility of the data bank managers to re-code data to a different format!
The Central MAP Data Bank Manager carries the responsibility for the operation of the MDC as a whole. The manager is not connected with a specific MAP research programme but is funded from central services through EUMETNET. The Special MAP Data Bank Managers will, as a rule, be experts involved in and funded via the particular project that gathers the data in question (e.g. aircraft and radar data, etc.).
Figure 4-1. Structure of the MAP Data Centre.
