Cross-source hypothesis
Species Without Addresses
Which observations may point to habitat outside recognized green spaces?
Find biodiversity observations outside mapped parks and compare them with vegetation imagery to create a human review queue.
Why this pairing matters
The non-obvious connection
Species observations describe where people looked; park maps describe where the city recognizes green space. Their disagreement can surface overlooked habitat or data errors.
The demonstration
One visible reveal
Recent observations glow outside mapped green polygons, then split into possible overlooked habitat, map omission, duplicate, or noisy observation for review.
Three-hour boundary
The smallest credible version
One district, a bounded species set, recent observations, park polygons, and a review queue. No ecological importance score.
Claim boundary
What the data cannot establish
Observer effort is highly biased, platforms can duplicate records, and sensitive species coordinates may need to be obscured.
Underlying evidence
Open every connected source
Community science
iNaturalist
iNaturalist contributors and platformInspect fields and samples →Biodiversity aggregation
GBIF occurrences
Global Biodiversity Information Facility and publishing institutionsInspect fields and samples →Community geospatial data
OSM social infrastructure
OpenStreetMap contributorsInspect fields and samples →Earth observation
MODIS NDVI/EVI
NASA LP DAAC at USGS EROS; hosted and processed by Microsoft Planetary ComputerInspect fields and samples →Shared ingredients