GBIF quickstart — species occurrences over a bounding box¶

Fetch a year of bird occurrences over a small bounding box through the EarthLens facade, then map the occurrence points.

GBIF is a vector backend: download() returns a pyramids FeatureCollection (a geopandas.GeoDataFrame subclass) of Point geometries in EPSG:4326. Search is anonymous, so no credentials are needed.

Install. GBIF search needs the pygbif SDK: pip install earthlens[gbif]. The query is paginated (300 records/page); max_records caps how many are pulled.

Setup¶

Import matplotlib for the map and the unified EarthLens entry point.

In [ ]:

import matplotlib.pyplot as plt

from earthlens import EarthLens

import matplotlib.pyplot as plt from earthlens import EarthLens

Request parameters¶

A small London bounding box and a one-year window. variables=["birds"] resolves to GBIF backbone taxonKey 212; you can also pass a raw key ([212]) or a name lookup (["taxon:Panthera leo"]).

In [ ]:

bbox_lat = [51.4, 51.6]   # [lat_min, lat_max]
bbox_lon = [-0.2, 0.0]    # [lon_min, lon_max]
start, end = "2020-01-01", "2020-12-31"

bbox_lat = [51.4, 51.6] # [lat_min, lat_max] bbox_lon = [-0.2, 0.0] # [lon_min, lon_max] start, end = "2020-01-01", "2020-12-31"

Download¶

One call returns the occurrence FeatureCollection and writes a GeoParquet under path.

In [ ]:

fc = EarthLens(
    data_source="gbif",
    variables=["birds"],
    start=start,
    end=end,
    lat_lim=bbox_lat,
    lon_lim=bbox_lon,
    path="out/gbif",
    max_records=2000,
).download()

print(f"{len(fc)} occurrences")
fc[["scientificName", "eventDate", "license"]].head()

fc = EarthLens( data_source="gbif", variables=["birds"], start=start, end=end, lat_lim=bbox_lat, lon_lim=bbox_lon, path="out/gbif", max_records=2000, ).download() print(f"{len(fc)} occurrences") fc[["scientificName", "eventDate", "license"]].head()

Map the occurrences¶

Plot the points; the geometry column is ready for any geopandas workflow.

In [ ]:

ax = fc.plot(markersize=4, color="darkgreen")
ax.set_title("GBIF bird occurrences — London, 2020")
ax.set_xlabel("longitude")
ax.set_ylabel("latitude")
plt.show()

ax = fc.plot(markersize=4, color="darkgreen") ax.set_title("GBIF bird occurrences — London, 2020") ax.set_xlabel("longitude") ax.set_ylabel("latitude") plt.show()

Licensing¶

GBIF records carry per-dataset licenses (CC0 / CC-BY / CC-BY-NC). When any record is CC-BY-NC, the backend raises a LicenseWarning — honour the non-commercial obligation before redistributing.