The global attribute DATA_SOURCE consists of two underscore separated fields. The first field describes the type of instrument or numerical model that created the data. The type may consist of several dot-separated words. The main constituent reported in the file shouldextend the DATA_SOURCE, e.g. LIDAR.O3, or FTIR.NO2. The second field is the acronym of the institute/organization (cf. affiliation acronyms link in Table 7.1above) that owns the instrument/model. This institute/organization may differ from the affiliations of the PI, the DO and the DS, but it is usually the PI's or DO's. Concatenated to the institute/organization acronym is a unique numeric identifier of the instrument.
The instrument identification system is mandatoryfor all instruments, including single or disposable instruments. Each organization must assure that different instruments of the same type have unique identifiers, even if they are operated in different locations. For example, if NASA owns two LIDARinstruments (LIDAR.O3), the entire entry for the DATA_SOURCE attribute for NASA's second instrument would become: LIDAR.O3_NASA.GSFC002. This instrument identification system allows each organization to create a unique identifierfor each instrument, without conflicting with other organizations instrument identifiers. Different organizations may operate several instruments of the same type at the same location without conflicting instrument identification. The instruments may be used at different locations at different times, while the instrument history remains traceable. When an instrument is taken out of service, the identifier must not be reusedfor another instrument. A particular case exists for instruments that are used as "consumables", for example weather balloons that are often lost after the sounding flight. In such cases a unique identifier may be useless. The identifier 000is therefore reserved for the non-unique case. A laboratory may re-use this particular identifier any number of times.
For satellite data from the original data release, DATA_SOURCE shall be reported with an instrument name and a platform name. e.g. SCIAMACHY.ENVISAT, or TOMS.EARTHPROBE. The main constituent reported in the file shouldextend the DATA_SOURCE, e.g. SCIAMACHY.ENVISAT.O3or TOMS.EARTHPROBE.O3. The second part of DATA_SOURCErefers to the satellite operating agency and an instrument identification number, e.g. SCIAMACHY.ENVISAT.NO2_ESA001. For data which werederived from the original data release, were reprocessed or filtered by algorithms developed by third parties, the second part of DATA_SOURCE shall refer to the third party's acronym and an instrument identification number, e.g F3C.FM1_UGRAZ.WEGC001. Note, the DATA_SOURCEappears in the file name of the reported data.
Actually used DATA_SOURCEs: Brand names shall not be included in the DATA_SOURCE. A distinction of instruments from different manufactures can be made in the second field by numbering. Alternatively a more detailed description can be given in the VAR_DESCRIPTIONfield. For a list of allowed variables please refer to Table 7.2.
Entry: Two underscore "_" concatenated fields
Format: Instrument or model type (from Table 7.2) "_" institute/organization acronym (from Table 7.1) concatenated with a unique three-digit identifier (for example 001, 007, 111, or 000for disposable equipment).
Example 1: DATA_SOURCE = SONDE.O3_NILU000
Example 2: DATA_SOURCE = FTIR.CH4_NILU001
Example 3: DATA_SOURCE = LIDAR.O3_NASA.GSFC002
Example 4: DATA_SOURCE = SCIAMACHY.ENVISAT.NO2_ESA001