Dataset configuration for multi-site datasets

[John Mund - NOAA Affiliate]

Hi,

I've got various time series datasets of atmospheric measurements.  For example; flask air samples from the MLO observatory in HI over 50 years (netcdf text).  We have similar datasets for many sites and species.  So at MLO there are ~50 species measured and similarly, CO2 is measured at ~50 different sites.  The data are all in netcdf files of species_site, so I've been using the EDDTableFromNcFiles dataset with timeseries datatype.  The files have detailed metadata tied to that site/species like the PI, local partners etc.  Any suggestions on how to organize this?  I tried doing a wildcard to import a whole folder of files, but it grabs the metadata from only 1 file (last by default) and displays it for the whole dataset, which is wrong because of the site specific content.  I also tried doing single datasets per file, which works but is cumbersome and makes it difficult for users to navigate and group.  Any suggestions on how to handle multi-site timeseries data? Thanks!
John

[Roy Mendelssohn - NOAA Federal]

I answered John directly, but to have a record in the group, I am responding here also. if i understand the issue John has global metadata in netcdf files that he wants to be included in the search, and the metadata values vary across the files to be aggregated. The solution is that there is an option to elevate global metadata into a variable in the table, for example:

<dataVariable>
<sourceName>global:platform</sourceName>
<destinationName>platform</destinationName>
<dataType>String</dataType>
<addAttributes>
<att name="long_name">Platform</att>
<att name="ioos_category">Identifier</att>
</addAttributes>
</dataVariable>

Here I have "global:platform” but any of the global metadata can likewise be elevated by substituting the name into “global:metadataParameter”.

-Roy

> --
> You received this message because you are subscribed to the Google Groups "ERDDAP" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to erddap+un...@googlegroups.com.
> To view this discussion, visit https://groups.google.com/d/msgid/erddap/13878f3f-8904-49f8-a23c-2199c27f51b8n%40googlegroups.com.

[John Mund - NOAA Affiliate]

Thank you! I think this will do what I need.  I was having trouble wrapping my head around how to transform my per site/species metadata into a 2 dimensional table.  Now I'll just need to decide how much to aggregate together using this method and how much to separate into distinct datasets.  I appreciate the quick response and helpful info.

John

[Roy Mendelssohn - NOAA Federal]

Great. Let is know if you run into further problems or have further questions. This looks like a great dataset.

-Roy

> To view this discussion, visit https://groups.google.com/d/msgid/erddap/ba5dd0ff-2fe6-4094-9f30-1bbe1b700b46n%40googlegroups.com.