Skip to content

Latest commit

 

History

History
46 lines (35 loc) · 3.73 KB

Coriolis_DAC.md

File metadata and controls

46 lines (35 loc) · 3.73 KB

Coriolis DAC - Standard Operational practices

How to provide glider data to Coriolis ? (users point of view)

To provide glider observations to Coriolis, providers have to push the data to Coriolis ftp server : 'ftp eftp.ifremer.fr' Contact [email protected] for user name and password. Glider data submission directory structure : submit/XXX/XXX_YYYYMMDD/ where:

  • XXX: glider name
  • YYYYMMDD : deployment date of the glider

Inside each deployment repository, we need raw files (m and dat files for slocum, netcdf or bpo files for seaglider and gz files for seaexplorer) and json files containing all the necessary information (global attributes, sensors information, calibration coefficient) if the provider wants us to decode data. Otherwise, we need final files if the provider/DAC generates itself the files that we must disseminate.

Data recovery

If we must decode data:

  • We check the json files format and informations
  • We use a matlab decoder developed by Ifremer (https://www.seanoe.org/data/00343/45402/) to produce the files to disseminate (explained in the next chapter). Note that a compiled version is available. Note that it’s not necessary to have a matlab license to run the decoder.

If we get already decoded data:

  • We check the format of the netcdf files received. If needed, we produce the profiles files from the EGO/OCEANGLIDERS netcdf file. Note that we can get/produce/disseminate data in real time on a daily basis.

Data dissemination

We produce and disseminate two types of netcdf files.

data Diffusion

All decoded and recovered data are available on an ftp space ( ftp://ftp.ifremer.fr/ifremer/glider/v2/ ). We generate index files at the root of this space to facilitate the search for data

An ERDDAP is link the this space (http://www.ifremer.fr/erddap/tabledap/OceanGlidersGDACTrajectories.html). This ERDDAP is used for the map of the https://www.oceangliders.org/ website.

The profiles files are ingested into Coriolis database then automatically disseminate to different projects like Copernicus or Emodnet physics It’s also possible to assign DOI to each deployment dataset (https://www.seanoe.org/data/00453/56454/ )

Summary of tools / documentations / link used

loead to improvement

By experience, we have some difficulties. For provider: provide complete and correct jsons. The way to fill/retrieve this metadata should be facilitated. As long as we have to check and correct the jsons, the ingestion cannot be automated. Get coefficient calibration: Very important for example to compute, validate DOXY. We have difficulty in recovering these coefficients from providers. Manage a common format between DACs as all DAC use their own tools. The future OceanGliders format will help a lot.