City Streets
Comparing street-network orientation, terrain, and city similarity with probability measures on ℝP².
City Streets asks how much a street network reveals about a city. It extends Geoff Boeing’s work on street-network orientation from compass bearings to three-dimensional directions.
Combining bearing and grade makes each network a probability measure on \(\mathbb{RP}^2\). Wasserstein and Jensen–Shannon distances compare these measures modulo vertical-axis rotation, so similarly built cities match regardless of grid orientation. The project also compares hilliness, walkability, on-street parking, and staircases. This release covers 133 study areas, including all five explicitly labeled New York City boroughs.
Open the interactive dashboard
The dashboard includes:
- A sortable table of key measures and 42 Compare metrics
- Per-city pages with the direction measure on the sphere, bearing roses, hill fans, street-character breakdowns, and a map of the analyzed urban core
- A distance heatmap and MDS embedding, showing which cities are built alike up to rotation
- A connected-component staircase census with a consistent 30 m minimum length