recommandations en matière de normes pour les données sur le pergélisol

Using standards for your permafrost data when sharing or publishing it is one of the ways to make your data more discoverable and reusable by others. Participants at the 2020 Permafrost Data Workshop identified access to standardized data and the discoverability of existing data as two of their main data-related challenges.

This page describes NSERC PermafrostNet’s recommendations for sharing and publishing permafrost data. Here we focus on ground temperature data and geotechnical measurements. These recommendations will promote data interoperability within the network and the broader community.

PermafrostNet has started to develop a python package to support file interoperability.

Contributions are welcome and we invite other researchers and initiatives to build on and contribute to the tools and practices we have. For more information, email us.

Describing permafrost datasets with CF Standard Names

The CF Standard names are a list of terms used to unambiguously identify the kinds of measurements in a dataset. Nick Brown and Michel Paquette (NSERC PermafrostNet) have coordinated the inclusion of 12 additional permafrost-related terms (see table below) in the latest release of the Standard Names vocabulary (version 78). These terms will be particularly useful for field scientists wanting to make their published datasets more interoperable and for data publishers who host permafrost-related data. An additional 14 terms relevant to permafrost science have since been requested. The addition of permafrost-related terms contributes to improved data interoperability with the atmospheric science and modelling communities, where the CF terms are widely used.

For self-describing formats like netcdf, you can include these attributes directly in your data file, following the CF standards. If you are publishing your data in a text file, like a CSV, you should include metadata (like the standard_name) about each column in your dataset in a separate metadata file. This information should also include the units and anything else that someone else might need to re-use your data.

Permafrost related terms added to the CF Standard Name vocabulary by NSERC PermafrostNet

Ground temperature data

Here we provide two recommendations for file formats and associated standards for permafrost ground temperature data. If you’re publishing your data in an online repository, don’t hesitate to include your data in more than one format.

Text Files

Text-based, or ASCII files are commonly used to store ground temperature measurements. These are familiar and simple to edit. However, you must include appropriate metadata so that your data can be reused by others.

• Choose text formats (*.csv, *.txt) over excel-based or proprietary formats (*.xls, *.xlsx)
• Follow an existing template for ground temperature measurements rather than coming up with your own
  • The Northwest Territories Geological Survey (NTGS) has developed standard templates for permafrost data and metadata:
    • For example, look at recently-published NTGS open files (e.g. Open Report 2019-007) or this example file.
    • You can check your own file against the template using the PermafrostNet file validator (you can register a new account to log in)
    • Note that the current template does not include time zone information, so you should include this somewhere in your metadata.

NetCDF

NetCDF files are commonly used in a variety of scientific disciplines to share data. These files are self-describing and are able to contain any required metadata. An example of netCDF files being used for permafrost data can be found in Nordicana-D issue 39. Or, you can download the file directly. Check back to this page for more examples of netCDF files for permafrost data.

• Follow the CF conventions; these describe a set of standards for structuring and describing netCDF files.
• Use the standard_name attribute to clearly specify what your data represents. The CF standard name table provides a list of accepted standard names. We recommend using the soil_temperature standard name for ground temperature.
• Include a units attribute to describe the units for each variable.
• Consider adding discovery-level metadata using the ACDD attributes.

Geotechnical data

Return to the main data page.