Big Data and Cloud Computing Reveal Global Squirrel Hotspots and Coldspots
-
Used machine learning, open access big data, and cloud computing to create a "Super SDM" model predicting global distribution hotspots and coldspots for over 300 squirrel species
-
Compiled and used 665,529 squirrel occurrence records from GBIF and 132 environmental predictor variables
-
Ran models on powerful Oracle Cloud Infrastructure over 7 days due to large data volumes
-
Identified major squirrel hotspots in North America, Europe, Southeast Asia, and Africa; major coldspots in Sahara Desert, South America rainforests, Siberia
-
This workflow with abundant data and cloud computing sets a new standard for species distribution modeling, overcoming old limitations of data and computing power