A lot of our datasets have been around a while, but new datasets come online occasionally, such as Shaking Layers, and a few of the smaller volcano datasets, for example scanDOAS. As a data user you’ll see a new dataset you are interested in, but probably little of what is wrapped around that dataset. Here we aren’t talking about the data delivery mechanism, though that is very important, but the documents, plans and references that help us keep a dataset alive, accessible and generally “well managed”. Most of what we do is also used in other parts of GNS Science as the framework was developed by GNS Science’s Information and Records team.

The steps are:

• a Data Management Plan

• a Dataset Catalogue entry

• a Digital Object Identifier (DOI)

• updating the GeoNet Data Policy web page

• triggering an evaluation of FAIRness for the dataset

This is the list for GeoNet data. Other agencies will have a different list, but the intent will be the same – to ensure well managed data. Having these should, we hope, contribute to our data community trusting a specific dataset and GeoNet as a dataset collector and deliverer in general. Let’s dig into each of these in a little bit more detail.

Data Management Plan (DMP)

A Data Management Plan (DMP) allows us to document what the data are, how they are collected, stored, accessed, kept secure, and who is responsible for doing this. In our case, this is a GNS Science document, so its normally not publicly available, but is something project funders (in our case GeoNet programme funders) might ask to see. There are many online resources about DMPs if you need to write your own.

Dataset Catalogue Entry

GNS Science maintains a catalogue of datasets that allows interested people to explore and find our data. A dataset catalogue entry is an outward facing summary of a dataset that is all about making it discoverable and providing access information. Most importantly, the entries can be read by computer systems that track datasets, so that agencies like data.govt.nz can automatically list our datasets and provide links to them.

We’ve talked in blogs about metadata for data, and you can think of a Dataset Catalogue Entry as metadata for a whole dataset.

Digital Object Identifier (DOI)

A Digital Object Identifier (DOI) is basically what its name suggests, a unique string of characters used to identify a digital object, such as an academic publication (e.g. Twenty years of volcano data at GeoNet—collection, custodianship, and evolution of open data on New Zealand’s volcanoes) or a GeoNet dataset (e.g. GeoNet Aotearoa New Zealand Earthquake Catalogue). A DOI looks a bit like a website URL, but critically starts with “doi.org”, e.g. https://doi.org/10.21420/0S8P-TZ38. The idea is that a DOI lives forever, even beyond the life of a dataset it points to. In our case, a DOI points to the Dataset Catalogue Entry for the dataset. Click on https://doi.org/10.21420/0S8P-TZ38 and see where it takes you.

You might wonder why the New Zealand Earthquake Catalogue needs a unique, persistent DOI when you can simply access it through the Quake Search application on our website. But what happens if the URL of that web page changes? The DOI will still exist, and you’ll still be able to find information about the earthquake catalogue. Referencing a dataset by its DOI is more stable than just using a web page URL. The DOI is also a way authors of academic publications can cite a dataset, thereby giving the dataset and its collector due credit for their work.

There are different ways to assign a DOI to data. For the GeoNet datasets, we have a pragmatic approach and assign a DOI to different dataset types, for example the earthquake catalogue, or our collection of seismic waveform data. Some agencies chose to also assign a DOI to an individual station producing data, but we feel the dataset approach is sufficient for our needs at the moment.

Our DOIs are issued by GNS Science’s Data Management Partner who request them from DataCite Fabrica, one of a dozen agencies worldwide that can issue DOIs. Getting a DOI is typically the last step before we publish a new GNS Dataset Catalogue Entry.

GeoNet Data Policy Website

A major focus of the Data Policy page on the GeoNet website is how to cite data. For this we provide a list of GeoNet datasets and their DOIs. You’ll notice the numeric part of all our DOIs starts with “10.21420”, that signifies GNS Science. The last part is a unique, random string generated by DataCite Fabrica. Adding a new dataset to the Data Policy web page gives a way for our community to quickly find the correct DOI when citing GeoNet data and provides access to metadata about the dataset through its Dataset Catalogue entry, which is what our DOIs point to.

FAIR Assessment

The last step in the process is to trigger GeoNet’s Science Operations and Data team to assess a dataset for FAIR. It’s a pretty low-tech process where we add the name of the dataset to a spreadsheet we use for FAIR assessment. We last did an assessment in September 2024, so we might do it again later this year, and we sometimes need a reminder to include our new datasets!

Building Trust

One of GeoNet’s core data principles is that we are a trusted source of data. This means our processes and procedures for data are aligned with international best practices and standards, metadata are actively maintained, data attributes are well understood, and quality is assessed. We hope that a DMP, Dataset Catalogue entry, DOI, Data Policy web page entry, and FAIR assessment contribute to building that trust in GeoNet as a data provider.

That’s it for now

Processes and procedures will never be the most exciting thing to read about, but in GeoNet data, as in many things, these are essential to a job done well. Although we understand you might not be enthralled by what wraps around our datasets, we hope you’ll appreciate that it is there and the value it adds.

You can find our earlier blog posts through the News section on our web page just select the Data Blog filter before hitting the Search button. We welcome your feedback on our data blogs and if there are any GeoNet data topics you’d like us to talk about please let us know!

Ngā mihi nui.

Contact: info@geonet.org.nz