gbif-biodiversity

What problem does it solve?

It turns noisy, biased GBIF occurrence records into a cleaned, spatially thinned dataset and then produces species distribution model predictions.

Core Features & Use Cases

• GBIF occurrence retrieval: Download species occurrence records via the GBIF API using pygbif, including basic filtering (e.g., coordinates, year range, optional country).
• Spatial thinning for sampling bias: Reduce clustering by retaining at most one record per specified latitude/longitude grid cell using GeoPandas and spatial logic.
• End-to-end SDM fitting and prediction: Fit MaxEnt-style SDMs with elapid and create suitability rasters; optionally build BRT models using scikit-learn and then stack/raster-threshold for richness mapping.
• Use Case: You have records for a species (e.g., a mammal or bird) across a region and want a defensible suitability map plus a richness map across multiple species for ecological analysis.

Quick Start

Use the gbif-biodiversity skill to download GBIF occurrences for a target species, spatially thin them, and fit an SDM to output a suitability GeoTIFF.