Skip to content

Week 4: Biodiversity Data Acquisition and Cleaning

Core Questions

  • What can go wrong when using biodiversity occurrence records?
  • How should taxonomic, coordinate, temporal, and metadata uncertainty be documented?
  • How can bad spatial data produce convincing but wrong analyses?

Concepts

  • GBIF and biodiversity databases.
  • Occurrence records.
  • Taxonomic uncertainty.
  • Coordinate uncertainty.
  • Duplicate observations.
  • Sampling bias.
  • Metadata.

Python Tools

  • pygbif.
  • requests.
  • pandas.
  • GeoPandas.

Applied Lab

Students download occurrence data for a species or taxonomic group and build a cleaning pipeline that handles missing coordinates, duplicate records, impossible coordinates, obvious spatial outliers, coordinate uncertainty, and duplicate specimens or observations.

Deliverable

Submit a data cleaning report with raw and filtered maps, a table of filtering decisions, and a short explanation of remaining uncertainty.