Similar to the establishment of an XML based published interface between third party modules and DELFT_FEWS through the General Adapter, the U.K. Environment Agency has established XML schemas as a standard for exchange of data between the various sub-systems of the Agency. This is particularly relevant to the Flow Forecasting Systems under development as it will include the exchange of data between this system and external data sources (RTS, DDS etc.).
The approach taken was compared to that used in the published interface described in this document. Although the two definitions are strictly separate, where the published interface described here is used to communicate between the DELFT-FEWS system and third party modules, while the EA Schemes are used for exchange of data between the different systems within the EA, it is seen as a potential advantage that the different schemes are comparable. It is accepted that the schemes may differ.
In the comparison of the schemes attention has only been paid to the Schema established for hydrometric data and the time series definition used here. The differences between the two schemas have been reviewed and where appropriate amendments have been made to the published interface time series format such that relevant information can be easily mapped from one scheme to the other. In the appendices of this document a spreadsheet is supplied indicating this mapping.
The main differences between the DELFT-FEWS Published Interface and the EA XML schemes comprise:

    1. The EA Schema contains more extended header information. Where appropriate these fields have been included in the published interface time series formats. Although the available fields are extensive, most are optional and need not be supplied if not available.
    2. The geographic datum included in the EA scheme considers only the British National Grid. In the published interface different datums may be considered and must be explicitly named.
    3. Both the published interface and the EA hydrometric data schemes apply a philosophy of a single data stream for each data type. This is in contrast with the tidal scheme where multiple data types are contained in each data stream. This approach is not advocated as this may endanger the openness of the system.
  • No labels