Crash course: mobilize your data to GBIF

Workshop given to Forest & Nature Lab on December 8th 2025.

Transcript

1. Stijn Cooleman 0000-0003-3029-0451 Sanne Govaert 0000-0002-8939-1305 Crash course: mobilize your data to GBIF Workshop for Forest and Nature Lab (Ghent University) 8 December 2025 Photo by Sanne Govaert
2. Overview • DwC-A: Understanding the core required terms. • Data transformation: Mapping data using R scripts. • Creating metadata • Best Practices: Using GitHub for version control and reproducibility. • Publishing
3. Sanne Govaert Research software engineer INBO o Open data o Machine observation of biologging data o Co-maintainer of Camtrap DP Stijn Cooleman Data publisher INBO o Open data o Belgian GBIF node staff o Facilitator of biodiversity data mobilisation
4. Open science lab for biodiversity At the Research Institute for Nature and Forest (INBO) in Belgium. We offer technical support to researchers in the projects we collaborate in. Our support is mainly focused on open data publication and research software development. https://oscibio.inbo.be/
5. Belgian Biodiversity Platform This science-policy interface provides services to the Belgian scientific community engaged in biodiversity research and to policy-makers as well as to practitioners. Within the data team, we offer science support services related to Biodiversity Informatics Tools and Standards, if you want to publish data or use data from GBIF. https://www.biodiversity.be/
6. What I did vs what I am doing now
7. Publishing biodiversity data
9. Global Biodiversity Information Facility GBIF - www.gbif.org
10. Global Biodiversity Information Facility GBIF - www.gbif.org
11. Species recorded at a specific place and time 3.5 billion occurrence records
12. Providing biodiversity evidence for research and policy
13. What do you get in return
14. Registered datasets get a DOI, are findable through GBIF website and API, and citations are tracked Discoverability
15. Data search Darwin Core standardization allows cross-dataset search through website and API
16. Reproducible downloads Any query can be downloaded, gets a DOI and there are clear citation guidelines
17. Distributed, active community supporting data publishers and users Community
18. Backbone taxonomy All data gets matched to a backbone taxonomy → unique ID, higher classification, synonymy resolution
19. LIFE RIPARIAS: invasive species early alert system based on GBIF data alert.riparias.be Infrastructure that can be build upon
20. Vlaams Biodiversiteitsportaal (VBP) https://natuurdata.inbo.be/
21. Species search example in VBP
22. Data paper
23. How to publish your data
24. Request endorsement to become a data publisher Standardize your data into Darwin Core Document your data with standardized metadatan (EML) Choose a license: CC0, CC-BY, CC-BY-NC Publish your data to an IPT Register your dataset to make it discoverable How to publish data to GBIF
25. Licenses CC0 (aka CC Zero) - Give up your copyright and put your works into the worldwide public domain CC BY - Credit must be given to the creator CC BY-NC - Credit must be given to the creator - Only noncommercial uses of the work are permitted.
26. Integrated Publishing Toolkit (IPT)
27. Data standards A data standard is a set of rules or guidelines that defines how data should be collected, stored, exchanged, and interpreted.
28. TDWG - www.tdwg.org Biodiversity Information Standards
29. Darwin Core Archive
30. Darwin Core
31. TDWG maintained standard to express biodiversity information: glossary of terms - dwc.tdwg.org Darwin Core
32. Most popular way to package biodiversity data: • data as CSV files (core + extensions) • metadata as XML Darwin Core Archive
33. https://rs.gbif.org/extensions.html
34. DwC structures data into key classes to ensure global interoperability. • 1. Occurrence: ◦ Focus: Evidence of an organism being in a specific place at a specific time. ◦ Example Terms: occurrenceID, basisOfRecord, individualCount. • 2. Taxon: ◦ Focus: Scientific identification and classification. ◦ Example Terms: scientificName, kingdom, genus. • 3. Event: ◦ Focus: Information about the sampling or collection process. ◦ Example Terms: eventID, eventDate, samplingProtocol.
35. Ecological Metadata Language (EML)
36. EML https://eml.ecoinformatics.org/ https://ipt.gbif.org/manual/en/ipt/latest/gbif-metadata -profile https://docs.ropensci.org/EML/
37. Hands-on Workshop
38. Mapping exercise - Map demo data to Darwin Core and EML Goal - So you are able to map your own data How? - Make use of an existing .Rmd script to create a valid Darwin Core Archive Workshop
39. Download the whole repo and open the R project: https://github.com/inbo/fornalab-datasets The script is at /template/src/dwc_mapping.Rmd Useful links • https://rs.gbif.org/extensions.html • https://dwc.tdwg.org/terms/ Links for workshop
40. Demo dataset https://formica.ugent.be/ Vegetation surveys within plots of 3x3 m2 Hierarchical structure: • 9 regions • 3 elevations • 3 forest types (transects) • 5 plots within a transect
41. A solution https://github.com/inbo/fornalab-datasets/blob/main/datasets/fornalab-formica-wp1 -vegetation/src/dwc_mapping.Rmd
42. Next steps
43. Publish data to GBIF 1) Define the scope of your dataset. 1 study = 1 dataset 2) Do a quality check of your data 3) Map your data to Darwin Core and EML, ideally on Github a) Make sure your records have unique, stable ID’s b) Choose a license c) Do a quality check of your transformed data 4) Ask INBO to upload your data to the INBO IPT, so it can be published to GBIF
44. Sanne Govaert sanne.govaert@inbo.be Stijn Cooleman stijn.cooleman@inbo.be Slides available at: https://bit.ly/mobilize-to-gbif Photo by Sanne Govaert Thank you!