: Public Package
| Created: |
24/06/2015 11:08:42 AM |
| Modified: |
25/06/2015 10:28:58 AM |
|
Project: |
|
| Author: |
dlowe |
| Version: |
1.0 |
| Phase: |
1.0 |
| Status: |
Proposed |
| Complexity: |
Easy |
| Difficulty: |
|
| Priority: |
|
| Multiplicity: |
|
Advanced: |
|
| UUID: |
{435897A4-6C6E-4b6d-859A-3F80261D7903} |
| Appears In: |
WIGOS Categories |
Have a 'DataProcessingMethod' class- very basic at first with data description, name etc. and possibility to link out to algorithms etc.<br /></p><p><br /></p><p>OSCAR attributes.<br /></p><p><br /></p><p>processing organisation<br /></p><p>proc.. method<br /></p><p>processor version<br /></p><p><br /></p><p>Much of category 7 goes in DataProcessingMethod. Level of Data in DataProcessingMethod. Reference.<br /></p><p><br /></p><p>'referenceTime' becomes 'referenceTimeSource'. - the type of instrument used to measure time. Codelist.<br /></p><p><br /></p><p>Latency of reporting - typical latency e.g. 3 hours. (TM_Duration) - goes in data processing.<br /></p><p><br /></p><p>Need to consider whether Data Processing and Reporting should be separate classes (both datatypes?)<br /></p><p><br /></p><p><br /></p><p><br /></p><p>resolution on data processing [NOTE: This is very similar to resolutionScale in METCE but not identical]<br /></p><p><br /></p><p>processingparty - CI responsible party.<br /></p><p><br /></p><p>reporting Latency should probably go in the ObservationProcess.<br /></p><p>We may want to re-use the TSML TimeseriesMetadata class for the temporal reporting properties, aggregation period etc. Note that this is a DataType so can't be byReference.<br /></p><p><br /></p><p>LevelOfData<br /></p><p>DataFormat and Version of data format would go in the MD_Metadata record. Out of scope of the observation metadata.<br /></p><p><br /></p><p><br /></p><p>ReferenceDatum is *critical* to interpreting the result. Is it part of the observed property? (name, type, value, unit). WHERE TO PUT THIS.<br /></p>