Using data from 10,000 taxi rides and the Google Maps API, students at Columbia University's Graduate School of Architecture, Planning, and Preservation created this mesmerizing animation of the transit arteries of New York City. The project is by Tom McKeogh, Eliza Montgomery and Juan F Saldarriaga.

Saldarriaga describes the project on his website: 

Taxi! is an analytical model that maps the trip data for 10,000 taxi rides over the course of 24 hours. Geographic location data for the origin and destination of each ride is combined with waypoint data collected from the Google Maps API in order to generate a geographically accurate representation of the trip. We used data from taxi rides originating or ending in the neighborhoods of Lincoln center or Bryant Park. The visualization recreates a ‘breathing’ map of Manhattan based on the migration of vehicles across the city over a period of 24 hours, displaying periods of intensity, density and decreased activity.

In another data visualization, he maps origins and destinations of taxi rides across New York over the course of a day:

For more information about Taxi!, visit http://www.juanfrans.com/taxi. 

Via Laughing Squid.